Williams System 9 - 11

Click to go back to the Williams solid state repair guides index.

1 Introduction

Williams System 9 Board Set from Space Shuttle

Williams System 11 (nothing) Board Set from High Speed

Williams System 11A Board Set from Pinbot

Williams System 11b Board Set from Jokerz! (Note that power supply is a prototype)

System 11b Board Set from Transporter the Rescue

System 11c Board Set from Rollergames

The Williams system 7 boardset was replaced in 1984 with the system 9 boardset, and then again in 1985 with the system 11 boardset. Combining the driver board, sound board, and cpu directly onto one board eliminated several design deficiencies of the earlier 3-7 boardsets; mainly the 40 pin interconnector, and extra wiring harness inter-board connectors. Larger ROMs could be fitted directly onto the boards allowing for more complex rulesets and sounds.

The System 9 and System 11 board sets are not 100% compatible. With a fair amount of effort, a System 11 board could be made compatible with a System 9, but it is not worth the extended effort to change the jumpers. It's far easier to just modify the rom images so that they are plug and play when installed in a System 11 CPU. You can find them already on the ipdb. The primary differences between the System 9 and System 11 board sets are the sound system and rom space. (add more info later)

As time went on, the System 11 board set underwent some changes too. The power supply, sound board, and CPU / driver board have minor changes. More apparent was the addition of circuit boards. Starting with the 11a board set, a more complex sound board was added (System 11 sound boards are strictly background sounds only). Later in the System 11 series, sound functions were partially (11B) or fully removed (11C) from the MPU, and eventually were solely handled by a dedicated sound board. Game-specific interconnect boards and auxiliary power driver boards were introduced in System 11B games. The interconnect boards acted as a pass-through for switch, lamp, flasher, and solenoid signal lines and circuits. It also allowed circuits with different voltages to share the same power source by using current limiting resistors to drop voltages to the appropriate levels. All general illumination fuses were moved to this board. The auxiliary power driver board has the AC relay, which finally centralized the location of it. Likewise, several fuses, beefier TIP36C transistors, and bridge rectifiers are centralized on the auxiliary power driver board.

2 Games

2.1 System 9

Title	Date of Release	Model #	Sound	Other Boards	Transformer	Notes
Strike Zone	1984	1916	unknown			This is a shuffle alley which uses some pinball boards
Star Light	06-1984	530	no speech		5610-09563-00	only a few were made with System 9 boards inside
Space Shuttle	12-1984	535	C-10716 (Speech only)		5610-09563-00 and 5610-10355-00
Sorcerer	03-1985	532	C-10716 (Speech only)		5610-09563-00 and 5610-10355-00
Comet	06-1985	540	C-10716 (Speech only)		5610-09563-00 and 5610-10355-00

2.2 System 11

Title	Date of Release	Model #	Sound	Other Boards	Transformer	Notes
Alley Cats	10-1985	1918	Unknown			This is a shuffle alley which uses some pinball boards
High Speed	01-1986	541	C-11030 (Background sound only)		5610-10897-00
Grand Lizard	04-1986	523	C-11030 (Background sound only)		5610-10897-00
Road Kings	07-1986	542	D-11197-542 (Background music only)		5610-10897-00

2.3 System 11A

Title	Date of Release	Model #	Sound	Other Boards	Transformer	Notes
Pinbot	10-1986	549	D-11297-549 (Background Speech & Sound)		5610-10897-00
Tic Tac Strike	11-1986	1919	Unknown			This is a shuffle alley which uses some pinball boards
Millionaire	01-1987	555	D-11298-555 (Background Speech & Sound)		5610-10897-00
F-14 Tomcat	03-1987	554	D-11298-554 (Used on 1st 5,000 games - Background Speech & Sound) / D-11581-554		5610-12136-00
Fire!	08-1987	556	D-11581-556		5610-12136-00
Fire! Champagne Edition	09-1987	556	D-11581-556

2.4 System 11B

Title	Date of Release	Model #	Sound	Power Supply	Aux Power Driver	Master Interconnect	Other Boards	Transformer	Notes
Big Guns	10-1987	557	D-11581-557	D-8345-557	D-11583-557	C-11762	Uses D-11814 Flash Lamp Resistor Board	5610-12136-00	Uses special solenoid switches
Space Station	12-1987	552	D-11581-552	D-8345-557	D-11813-552	C-11762-552	Uses D-11814-552 Flash Lamp Resistor Board	5610-12136-00	Uses special solenoid switches
Gold Mine	01-1988	1920	None	D-8345-1914	None	None			This is a shuffle alley which uses some pinball boards
Cyclone	02-1988	564	D-11581-564	D-8345-557	D-11813-574	C-11762-564	Uses D-11814-564 Flash Lamp Resistor Board	5610-12136-00	Does not use special solenoid switches
Banzai Run	05-1988	566	D-11581-566	D-8345-566	D-12247-566	D-12112		5610-12136-00	Uses special solenoid switches
Swords of Fury	06-1988	559	D-11581-559	D-8345-557	D-12247-559	D-12313-559		5610-12136-00	Does not use special solenoid switches
Taxi	08-1988	553	D-11581-553	D-12246-553	D-12247-559	D-12313-553		5610-12136-00	Uses special solenoid switches
Top Dawg	11-1988	1921	None	D-12246	None	D-12313-1921			This is a shuffle alley which uses some pinball boards
Jokerz!	12-1988	567	D-12338-567	D-12246	D-12247-559	D-12313-567		5610-12136-00	Does not use special solenoid switches
Earthshaker	02-1989	568	D-11581-568	D-12246-568	D-12247-559	D-12313-568		5610-12136-00	Does not use special solenoid switches
Black Knight 2000	04-1989	563	D-11581-585	D-12246	D-12247-563	D-12313-563		5610-12136-00	Uses special solenoid switches
Transporter the Rescue	04-1989	2008	D-11581-2008	D-12246	D-12247-2008	D-12313-2008		5610-12136-00	Does not use special solenoid switches
Police Force	08-1989	573	D-11581-573	D-12246	D-12247-566	D-12313-573		5610-12136-00	Does not use special solenoid switches
Shuffle Inn	10-1989	1922	None (assumed)	D-12246 (assumed)	None	D-12313-1922 (assumed)			This is a shuffle alley which uses some pinball boards
Elvira and the Party Monsters	10-1989	2011	D-11581-2011	D-12246	D-12247-2011	D-12313-2011		5610-12136-00	Does not use special solenoid switches
Bad Cats	11-1989	575	D-11581-575	D-12246	D-12247-566	D-12313-575		5610-12136-00	Does not use special solenoid switches
Mousin' Around!	12-1989	2009	D-11581-2009	D-12246	D-12247-2009	D-12313-2209		5610-12136-00	Does not use special solenoid switches
Whirlwind	01-1990	574	D-11581-574	D-12246	D-12247-574	D-12313-574	Uses a C-13287 Sound Overlay Solenoid Board to control 5 extra devices	5610-12136-00	Does not use special solenoid switches

2.5 System 11C

Title	Date of Release	Model #	Sound	Master Interconnect	Other Boards	Transformer	Notes
The Bally Game Show	04-1990	2003	D-11581-2003 w/ 512K ROMS	D-12313-2003	Uses a C-13286 Sound Overlay Lamp Board	5610-12136-00	Does not use special solenoid switches
Pool Sharks	06-1990	2014	D-11581-2014 w/ 512K ROMS	D-12313-2014		5610-12136-00	Does not use special solenoid switches
Rollergames	06-1990	576	D-11581-576 w/ 512K ROMS	D-12313-576		5610-12136-00	Does not use special solenoid switches
Diner	09-1990	571	D-11581-571 w/ 512K ROMS	D-12313-571		5610-12136-00	Does not use special solenoid switches
Radical!	09-1990	2015	D-11581-2015 w/ 512K ROMS	D-12313-2015		5610-12136-00	Does not use special solenoid switches
Dr. Dude	11-1990	2016	D-11581-2016 w/ 512K ROMS	D-12313-2016		5610-12136-00	Does not use special solenoid switches
Riverboat Gambler	11-1990	50007	D-11581-50007 w/ 512K ROMS (assumed)	D-12313-50007		5610-12136-00	Does not use special solenoid switches
Bugs Bunny's Birthday Ball	01-1991	20009	D-11581-20009 w/ 512K ROMS	D-12313-20009		5610-12136-00	Does not use special solenoid switches

Game date of release and model numbers provided by the Internet Pinball Database - http://www.ipdb.org

3 Recommended Documentation

3.1 Manuals & Schematics

Schematics for each game are essential in tracing down connections to lamps, switches, and solenoids. The owner's game manual is a handy resource to have for general game operation, game settings & diagnostics, switch & lamp matrices, fuse lists, circuit boards and parts, playfield assemblies, some basic schematics, and wiring diagrams.

Be aware that although rare, some factory documentation contains errors in schematics, wiring, switch/lamp matrices, and solenoid assignments.

3.1.1 Documentation Errors

• Error on the MPU I/O signals for U21 / U22 of High Speed.
• Flipper Circuit Schematics Error in the manuals for High Speed, Grand Lizard, and Road Kings.

3.2 Theory of Operation & Schematics

Details for board and cabinet schematics System 9-11 games can be found in the game manual or parts catalogs. Some copies can be found on Planetary Pinball or on IPDB.

Manufacturer	Guide	Version	Cover	Source	Games	Notes
Williams	Pinball Troubleshooting & Reference Manual	69-8955 (April 21, 1986)		PDF	Comet, High Speed, Sorcerer, Space Shuttle, Star Light	System 9 games. Also includes Strike Zone shuffle alley game.

3.3 Schematics

• D-11581 Sound Board

3.4 Parts Catalogs

Parts catalogs can also be useful, which include part numbers (helpful for purchasing parts online), exploded views of assemblies (helpful to see how the assemblies are put together), and board layouts & parts lists, and diagrams for controlled lamp, solenoid, and rubber locations. Parts catalogs also had supplementary updates to add and replace pages as newer games were released after the initial release of the catalogs.

Online copies of Bally & Williams parts catalogs can be found on Planetary Pinball.

Manufacturer	Catalog	Version	Cover	Source	Games	Notes
Williams	Blue Parts Catalog (1986-1989)	16-9064-B		Viewer PDF	Banzai Run, Big Guns, Black Knight 2000, Comet, Cyclone, Earthshaker!, F-14 Tomcat, Fire!, Grand Lizard, High Speed, Jokerz!, Millionaire, Pin Bot, Police Force, Road Kings, Sorcerer, Space Shuttle, Space Station, Swords of Fury, Taxi
Williams	Red Parts Catalog (1989-1992)	16-9190-B		Viewer PDF	Bad Cats, Diner, Dracula, Fish Tales, Funhouse, The Getaway, Hurricane, The Machine, Riverboat Gambler, Rollergames, Slugfest, Terminator 2, Whirlwind, White Water
Bally/Midway	Yellow Parts Catalog (1988-1992)	16-9147-B		Viewer PDF	Addams Family, Atlantis, Black Rose, Bugs Bunny, Dr. Dude, Dr. Who, Elvira, Game Show, Gilligan's Island, Harley Davidson, Mousin' Around, Party Zone, Pool Sharks, Radical!, Transporter, Truck Stop

3.5 Service Bulletins

Supplementary service bulletin books were released, which included service bulletins that detailed any problems or issues with games that were discovered after they were released.

Manufacturer	Version	Cover	Source	Games	Notes
Williams	Service Bulletin Book, 1989		PDF	Cyclone, Bonzai Run, Taxi, Earthshaker, Truck Stop, Transporter
Williams	Service Bulletin Book, 1990		PDF	Diner, Police Force, Pool Sharks, Rollergames
Williams	Service Bulletin Book, 1991		PDF	Dr. Dude, Jokerz, Riverboat Gambler	This book includes some WPC games

4 Technical Info

System 9 eliminated a huge issue with the earlier system 3 through 7 boardsets, the 40 pin interconnect between the MPU and Driver boards. All circuitry previously separated into MPU, driver, and sound boards was combined into a single MPU board. Satellite boards were still used for displays, speech, and solenoid expansion.

Starting with Big Guns, Williams introduced another weak link into their system, the interconnect board. Initial implementations were used to disperse the general illumination between the playfield and backbox. The connectors on interconnect boards were prone to overheating and failure just like prior implementations. At the same time, flash lamp resistor boards were being used. These were created to collocate all of the flashlamp current limiting resistors, and eliminate under playfield flashlamp resistor boards, which had proven themselves problematic in the high heat, high vibration environment under the playfield. Resistors often desoldered themselves and/or vibrated so much that the connection fractured. Flash lamp resistor boards were only used on three games: Big Guns, Space Station, and Cyclone. In addition to the interconnect and the flash lamp resistor board, the auxiliary power driver board made its appearance. The auxiliary power driver board housed all of the high current transistors (TIP36c), solenoid and flasher fuses (including discrete flipper fuses), the AC select relay, and the solenoid power bridge rectifiers. The auxiliary power driver board did go through some minor changes, but it was functionally equivalent through the last System 11C game.

Starting with Banzai Run, a more standardized interconnect board was created. This interconnect board merged the general illumination interconnect with the flasher driver board. Switch and lamp circuits passed directly through this board as a matter of manufacturing convenience. Later interconnect boards added 3 opto couplers to indicate when the flippers and AC relay were activated (the same circuitry as used in the flipper power section of the WPC-089 -1 Power/Driver board).

One of the largest advantages of the system 11 board set is its ability to switch one set of coil driver transistors between 2 sets of coils/flashers. This is accomplished via a relay on a separate board called the A/C select relay (the A/C select relay was eventually moved to the auxiliary power driver board starting with Big Guns). The theory is that coils that are less frequently activated will be on the A side of the relay, and that flashers will be on the C side of the relay. Most of the time during gameplay, the C side is active, enabling the flashers. If the driver transistor for the A/C select relay itself shorts, the relay defaults to the coil side, allowing the game to continue to operate in a reasonable fashion (usually the A side coils are the ball shooter lane, drop target resets, VUK - basically coils that don't need to be able to operate 100% of the time unlike a slingshot coil or pop bumper coil). Data East's board set was essentially a copy of the system 11 board set, but Data East programmed it the opposite way, so that the flasher side of the A/C select relay is active by default. Data East games, when the A/C relay doesn't operate, just sits there and flashes lamps instead of playing.

System 9 and early System 11 games do share a disadvantage with the earlier 3-7 board sets, they still utilize special solenoid circuitry. The pops and slings on system 9/11 games do not activate via a switch which is seen by the MPU's program which fires the coil. Instead, the switches directly fire their associated coils via logic gates on the main board. The main disadvantage with this system is that the solenoids fire continuously as long as their activation switch is closed. A locked on sling or pop will burn out components quickly. It is recommended to add a 1, 1.5, or 2 amp inline fuse to each coil on a directly fired system 11 game. Usually just the pops and slings are direct fired coils. However, other coils were designated to the special solenoids.

Later System 11 games did not use the direct fire solenoid switches. Switches on the switch matrix were used instead. To easily identify games which do not use direct fire solenoid switches, IJ18 on the MPU board will not have a harness connection attached to it. Likewise, the slingshots and pop bumpers will no longer have a set of scoring switches.

One way that System 9 games differ from the previous 3-7 board set is that the special solenoids are not controlled by the CPU at all. In other words, the special solenoids are not pulsed during solenoid test, as the physical board architecture does allow for this.

The software in system 9-11 games continued with the same machine architecture from system 6 and 7, except for the special solenoids as mentioned above. More ram became available, up to 2k, but no game used over 800 bytes. This is likely due to the lower cost of 6264 RAM than 6116 RAM as was used in early System 9-11 games more than any other reason. Additional diagnostics in the form of switch error checking appeared in system 11. Any switch not activated over 30 games (16 games for flipper switches) would cause an error to be called out at power on. The software was able to compensate for broken switches by designating another switch to perform the same function as the broken switch.

4.1 MPU Driver Boards

System 9 MPU

These high frequency filter capacitors were added at the factory, and are normal. Their purpose is the snub high frequency noise in the Advance and Auto Up/Down coin door button circuits.

System 9 MPU Williams used a single version of the System 9 MPU in all System 9 games. A few minor modifications to the board may have been made, but it's essentially the same board for the entire 3 game run of Space Shuttle, Sorcerer, and Comet. The picture at left shows an early board that featured two inductors (copper wound coils that "smooth" DC power) at L1 and L2, adjacent to connector 1J17. Later revisions of the board used "zero ohm" jumpers instead, presumably to reduce cost.

Williams System 10 MPU component side. Image courtesy of Brian Johnson.

Williams System 10 MPU solder side. Image courtesy of Brian Johnson.

Williams released a single game using an MPU board designated System 10. The game "4-in-1" (IPDB link) was released in 1985 and is more like Pachinko than pinball.

From the Internet Pinball Database:

"This game is known in Williams documentation as "4-IN-1" but that name never appears on the game itself. A total of 100 games were produced. Each came with four interchangeable playfield overlays for the themes of Baseball, Poker, Football, and Soccer. The operator could change the overlay and Instruction Card, re-option switches, and have a new game in the same cabinet."

System 11 games used one of four different revisions of the System 11 MPU. The first version of the System 11 MPU was completely self-contained, able to make game sounds on the board. A simple background sound board was used to add some game sounds. Between the baseline revision of the System 11 MPU and the last revision of the MPU (System 11C) incrementally more of the sound processing and sound circuitry was removed from the board (not stuffed). None of the System 11 shuffle alley games use a separate sound board.

Early System 11 (nothing) MPU

Later System 11 (nothing) MPU. Note extensive alkaline corrosion abatement, NVRAM install, and use of modern reset generator. Image courtesy of Chris Hibler

System 11 (nothing) MPU
This picture shows the early System 11 MPU. A later revision of this MPU, pictured at right, has an additional jumper W16 located between the 7 segment diagnostic display and the relay, and a 4.7K ohm resistor between the 7447 and the resistor DIP. The early PCB on the left has a trace routing error to the sound ribbon cable connector. If you use it with a later system 11A sound card you will get no sound because the sound card doesn't get reset properly. If you wonder why your High Speed doesn't beep when turning it on this one is also the culprit.

6.2V Zener Diodes added to the back of the System 11 (nothing) MPU. These zeners were added to the layout of the PCB in subsequent revisions of the board (11A, etc).

6.2V Zener Diodes (ZR3-ZR8) incorporated into the board layout and stuffed at the factory on a later board revision. Note: in this image, ZR7 and ZR8 have been replaced

The picture at left shows six 6.2V zener diodes added to the solder side of a System 11 (nothing) MPU. These zeners were left off the factory silkscreen of the board. Later revisions of the System 11 MPU added them to the component side of the board. Their purpose is to protect the special solenoid switch inputs. Either a ZPY6,2/ZD6,2/1N4735A (1 watt) or ZF5,2/1N5234B (1/2 watt) can be used. This mod is purely optional. In reality the diode doesn't help much if you short the switch input with the solenoid voltage. In this case the diode and the 7402 are usually damaged and need to be replaced.
Additionally, if the System 11 (nothing) MPU is used in a System 11A game or later, pins 6 and 7 of the adjacent connector (J18) must be tied to the ground plane at the edge of the board. These pins provide the ground for the special solenoids. Without the modification they do not work. Some early System 11A pins might not need the mod. If unsure just check if a cable is connected to pin 6 or 7 in your game.

A rare prototype System 11a MPU

Blanking, 5VDC, and Diagnostic LEDs factory installed in a "transition" vintage System 11a MPU. The prior version 7447 (U48) and resistor DIP (U47) has been replaced with a 7407, discrete resistors and LEDs.

A prototype System 11a MPU with original label

System 11a Prototype MPU
This MPU was only used in Pin Bot. It's actually a System 11 (nothing) MPU PCB layout, with the 7 segment LED display replaced at the factory with 3 LEDs as used in all later System 11 MPUs (11A/B/C). The latest trace revision of this board also has the ground mod (connect 1J18, pins 6 and 7 to board ground) allowing it to work in any 11A/B/C game without modification. According to a friend it was referred to by the name of 11a instead of 11A by distributors.

A late production System 11A MPU.

System 11A MPU
The System 11A MPU can be used in all System 11 games. If used in one of the four first System 11 games (Alley Cats, High Speed, Grand Lizard, and Road Kings), connector 1J15 and possibly the amplifier section must be stuffed. System 11A, 11B, and 11C games don't use the amplifier section of the MPU board. Sound amplification is accomplished on the sound board. On early production system 11A boards, the amplifier section was fully stuffed, just like the original System 11 board. Later in the production of System 11A boards, the amplifier section was not stuffed, just like an 11B board. But unlike System 11B boards, the System 11A board still has the same switch section layout as the original System 11 MPU, making it easy to identify.

Note: the sound mixing section of the System 11A MPU is stuffed with different resistor and capacitor values than that of the System 11B MPU. This may cause some sound quality differences as has been noted with EarthShaker.

System 11B MPU. Note that a large portion of the sound circuitry, including the TDA2002 amp, was not stuffed in late System 11A boards or any System 11B MPU boards.

System 11B MPU
The System 11B MPU can be used in all System 11 games too. If used in one of the four first System 11 games (Alley Cats, High Speed, Grand Lizard, and Road Kings), connector 1J15 and all of the sound amplification components must be added to the board. System 11A, 11B, and 11C games don't use the amplifier section of the MPU board. Sound amplification is accomplished on the sound board.

Note: the sound mixing section of the System 11A MPU is stuffed with different resistor and capacitor values than that of the System 11B MPU. This may cause some sound quality differences as has been noted with EarthShaker.

System 11C MPU. Note that none of the sound circuitry, including the 6802 processor and 6821 PIA, is stuffed in System 11C MPU boards. All sounds were generated by the sound board.

System 11C MPU
The System 11C MPU can be used only in System 11C games since none of the sound processing circuitry was stuffed. It can be used in games prior to 11C, if the sound section is populated. However, some of the components needed are rare and / or costly, and there are a fair amount to install. If there is no other choice but to use this board, it is possible.

The RottenDog "System 9211" Aftermarket MPU

The RottenDog "MPU004" Aftermarket MPU with sound section overlay

RottenDog MPU9211 and MPU004 Aftermarket MPUs
The RottenDog aftermarket MPU9211 (shown at left) was designed to support all Williams System 9 - 11 games. How well the board works across that great breadth of games is somewhat questionable.

The CVSD used on this board is a CMX639D4 produced by CML Microcircuits. It replaces the 55516 (or 55536) CVSD used at U3 on original System 9 - 11 MPU boards.

The RottenDog aftermarket MPU004 (shown at right) was an earlier design thought to support all Data East and System 11 games. Again, how well the board accomplishes this for Williams games is questionable. A significant source of confusion on this board (like all RottenDog boards with DIP switches) is that the DIP switches are labeled opposite of traditional labeling. That is, set all switches to OFF for Williams games, ON for Data East games.

The RottenDog "System 9211" Aftermarket MPU, later revision

A later revision of the RottenDog 9211 MPU included a second trim POT to adjust the balance between 1408 DAC sounds and 55536 CVSD sounds.

4.1.1 Lamp Matrix

Picture of the lamp column portion of the lamp matrix of a System 9 MPU board. Note that the TIP42 drive transistors were removed from this board.

The lamp matrix is made up of two parts - the lamp columns (drives) and lamp rows (returns). In most cases, the lamp row consists of eight 2N5060 over-current shunt SCR's and eight TIP122 drive transistors, while the lamp column consists of eight 2N6427 predrive transistors (WMS sometimes used MPS-A14 instead) and eight TIP42 or TIP42C transistors. However, there are some instances, as shown in the pic to the left, where the predrive transistors for the lamp columns are 2N6548s. It is unknown why Williams chose to use these transistors on occasion.

The over-current shunt circuit on the lamp row (returns) works by sensing the voltage drop across the 0.4 ohm, 3 Watt resistor on emitter of the TIP122 driver transistor. If a short circuit is present in the lamp row, the short circuit current will result in a higher than normal voltage drop across the emitter resistor. This voltage will trigger the gate of the 2N5060 SCR, causing it to turn on and shunt the base drive to the TIP122 transistor. This will turn off the driver transistor and prevent damage due to excessive current. Note that an RC circuit consisting of a 0.1uF capacitor and 1K ohm resistor is present on the gate of the SCR. This prevents a false trigger of the over-current shunt circuit due to an initial peak current from cold lamp filaments.

4.1.2 Switch Matrix

4.1.2.1 System 9 Switch Matrix Explained

The System 9 Switch Matrix circuitry, as found in the Sorcerer manual.

General Description

At regular intervals, say 1 every msec or so, the 6808 microprocessor scans the switch matrix. This is done by pulling one switch column to ground, then sensing the switch row return state. Each row on which ground is detected means that the switch at the intersection of the column grounded and the detected row ground, is closed.

Hardware Detail

The 6808 writes the column data to the address location occupied by the 6821 at U15 on the board. This causes the 6821 output registers (PB0 – PB7) to propagate the data to a 74LS244 octal buffer at U45. The 74LS244 mirrors that same data at it's output pins. The data pattern on these output pins will have all bits set to logic “1” except one bit, which will be set to logic zero, or electrical ground. The ground is applied to the base of a 2N3904 transistor (which had been held high by a 4.7K SIP pull-up resistor SR8) where the ground is applied to the switch column (called switch matrix drives on the schematic).

Closed switches connected to that column will then cause the normally high (via yet another pull up resistor, SR16) gate at a 74LS08 quad NAND gate (U51 and U52) to be pulled to ground. This NAND gate enables the circuitry to detect a switch closure returned via either connector 1J9 or 1J10.

The 6808 then “reads” the row returns by reading from the address space occupied by the same 6821 at U15 used to pulse the column strobes earlier. This time it reads 6821 registers PA0 – PA7.

Again, the processor knows which column it drew to ground, and it knows which rows were caused to be pulled to ground via closed switches attached to the column. From this information, the closed switches can be identified. Genius. Bania: “I'm tellin' ya Jerry, this switch matrix stuff is gold Jerry, it's gold!”.

The switch diode isolates the switch in the matrix. Without this isolation diode, multiple simultaneous switch closures will cause false returns and confuse the processor. A switch diode that is open will always prevent a switch from registering, in the same way that a broken wire or failed open switch would.

4.1.3 Jumper Settings and Locations for System 9 & 11 MPU Boards

Below are charts of the jumpers on System 9 & 11 boards. Rather than listing the boards by their platform (System 11A/B/C), each chart is listed by the numbers stenciled on the circuit boards.

5764-10749-00 REV. C (System 9)

Typically installed jumpers are W2, W5, W6, W7, W9, & W12.

Jumper	Location	Purpose	Notes
W1 / W2	Below U12	Microprocessor Internal RAM Enable. W1 IN, W2 OUT when using 6802, reverse when using 6808	U12 RAM can be removed with 6802 and W1 installed.
W3
W4
W5	Left of U49	Country jumper. Remove for German factory settings, add for US factory settings
W6	Below U5	Along with W7, used to send +12V to opto rows	Typically installed, but not used
W7	Below U5	Along with W6, used to send +12V to opto rows	Typically installed, but not used
W8	Below connector IJ9	Ties resistor network SR15 to +12v
W9	Below connector IJ9	Ties resistor network SR15 to +5v
W10	Below U50	Installed - sounds from CPU board can be played without speech board attached	Typically not installed, except Strike Zone
W11	Below U37	Used when U19 game ROM 2 is used
W12	Below U36	Used to increase U20 game ROM 1 size from 2764 to 27128

5764-10881-00 REV. - (Typically System 11)

Typically installed jumpers are W1, W2, W4, W5, & W7.

Jumper	Location	Purpose	Notes
W1	Below U15 CPU chip	Ties crystal to clock input (pin 39) of U15 CPU
W2 / W3	To the right of U25 RAM chip	Used to increase the ROM size
W4	Below U31 (to the right of U25)	Used to increase the ROM size
W5	Below U39 & U38 (PIA chip)	Used to apply +5v to SR14 resistor network
W6	Below U30 & U27	Used to apply +12v to SR14 resistor network
W7	Below U48	Country jumper. Remove for German text and settings, add for English text and settings
DS1	Left of U13		Marked as DS1 on board (lower of 2 jumpers) - leave out
DS1	Left of U13		Marked as DS1 on board (upper of 2 jumpers) - leave out
DS2	Left of U13		Marked as DS2 on board (lower of 2 jumpers) - keep installed
DS2	Left of U13		Marked as DS2 on board (upper of 2 jumpers) - keep installed

5764-10881-00 REV. B (Typically System 11 - Referred to as Prototype System 11a in text above)

Typically installed jumpers are W1, W2, W4, W5, W7, W8, W11, W12, W13, W14, W16, W17 & W18.

Jumper	Location	Purpose	Notes
W1	Below U15 CPU chip	Ties crystal to clock input (pin 39) of U15 CPU
W2 / W3	To the right of U25 RAM chip	Used to increase the ROM size
W4	Below U31 (to the right of U25)	Used to increase the ROM size
W5	Below U39 & U38 (PIA chip)	Used to apply +5v to SR14 resistor network
W6	Below U30 & U27	Used to apply +12v to SR14 resistor network
W7	Below diagnostic LEDs	Country jumper. Remove for German text and settings, add for English text and settings
W8 / W9	Below U24 (sound CPU chip)	Used to choose what to use as /RST for U24 sound CPU
W10 / W11	To the right of 3 radial caps in reset section	Choice of +18v (W10) or +12v (W11) for sound amplifier
W12	A resistor below U10 PIA	Pull down resistor for E1 (pin 1) and E2 (pin 19) of U11 & U13 address line buffer chips
W13	A resistor below U16 data line buffer chip	Pull down resistor for /E (pin 19) of U16 data line buffer chip
W14 / W15	Below U29	1ms IRQ signal (W14 in) or 2ms IRQ signal (W15 in)
W16	Below U27	Blanking signal to input "D" of U48 LED decoder chip	not present on first pcb revision
DS1	Left of U13		Marked as DS1 on board (lower of 2 jumpers) - leave out
DS1	Left of U13		Marked as DS1 on board (upper of 2 jumpers) - leave out
DS2	Left of U13		Marked as DS2 on board (lower of 2 jumpers) - keep installed
DS2	Left of U13		Marked as DS2 on board (upper of 2 jumpers) - keep installed

5764-12091-00 REV. - (Typically System 11A)

Typically installed jumpers are W1, W2, W4, W5, W7, W8, W11, W12, W13, W14, W16, W17 & W18.

Jumper	Location	Purpose	Notes
W1	Below U15 CPU chip	Ties crystal to clock input (pin 39) of U15 CPU
W2 / W3	To the right of U25 RAM chip	Used to increase the ROM size
W4	Below U31 (to the right of U25)	Used to increase the ROM size
W5	Below U39 & U38 (PIA chip)	Used to apply +5v to SR14 resistor network
W6	Below U30 & U27	Used to apply +12v to SR14 resistor network
W7	Below diagnostic LEDs	Country jumper. Remove for German text and settings, add for English text and settings
W8 / W9	Below U24 (sound CPU chip)	Used to choose what to use as /RST for U24 sound CPU
W10 / W11	To the right of 3 radial caps in reset section	Choice of +18v (W10) or +12v (W11) for sound amplifier
W12	A resistor below U10 PIA	Pull down resistor for E1 (pin 1) and E2 (pin 19) of U11 & U13 address line buffer chips
W13	A resistor below U16 data line buffer chip	Pull down resistor for /E (pin 19) of U16 data line buffer chip
W14 / W15	Below U29	1ms IRQ signal (W14 in) or 2ms IRQ signal (W15 in)
W16	Below U27	Ties XTAL (pin 38) of U15 CPU chip to ground	Only cut when using Hitachi 6808 or 6802 MPU at U15
W17	Below U23	Ties XTAL (pin 38) of U24 sound CPU chip to ground	Only cut when using Hitachi 6808 or 6802 MPU at U24
DS1	Left of U13		Marked as DS1 on board (lower of 2 jumpers) - leave out
DS1	Left of U13		Marked as DS1 on board (upper of 2 jumpers) - leave out
DS2	Left of U13		Marked as DS2 on board (lower of 2 jumpers) - keep installed
DS2	Left of U13		Marked as DS2 on board (upper of 2 jumpers) - keep installed

5764-12206-00 REV. A (Typically System 11B & 11C)

Typically installed jumpers are W1, W2, W4, W5, W7, W8, W11, W14, W16, W17 & W19.

Jumper	Location	Purpose	Notes
W1	Below U15 CPU chip	Ties crystal to clock input (pin 39) of U15 CPU
W2 / W3	To the right of U25 RAM chip	Used to increase the ROM size
W4	Below U31 (to the right of U25)	Used to increase the ROM size
W5	Above U25 RAM chip	Used with 2K x 8 6116 RAM at U25
W6	Above U25 RAM chip	Used with 8K x 8 6264 RAM at U25	only 2K will be used from the factory
W7	Below diagnostic LEDs	Remove for German text, add for English text	This jumper is not used in games with 16 digit score displays. Instead, install the U26 and U27 EPROMs for country and game.
W8 / W9	Below U24 (sound CPU chip)	Used to choose what to use as /RST for U24 sound CPU
W10 / W11	To the right of 3 radial caps in reset section	Choice of +18v (W10) or +12v (W11) for sound amplifier
W12	A resistor below U10 PIA	Pull down resistor for E1 (pin 1) and E2 (pin 19) of U11 & U13 address line buffer chips
W13	A resistor below U16 data line buffer chip	Pull down resistor for /E (pin 19) of U16 data line buffer chip
W14 / W15	Below U29	1ms IRQ signal (W14 in) or 2ms IRQ signal (W15 in)
W16	Below U27	Ties XTAL (pin 38) of U15 CPU chip to ground	Only cut when using Hitachi 6808 or 6802 MPU at U15
W17	To the left of U24	Ties XTAL (pin 38) of U24 sound CPU chip to ground	Only cut when using Hitachi 6808 or 6802 MPU at U24
W18	Below U23	Used with 8K x 8 6264 RAM at U23	only 2K will be used.
W19	Below U23	Used with 2K x 8 6116 RAM at U23
DS1	Left of U13		Marked as DS1 on board (lower of 2 jumpers) - leave out
DS1	Left of U13		Marked as DS1 on board (upper of 2 jumpers) - leave out
DS2	Left of U13		Marked as DS2 on board (lower of 2 jumpers) - keep installed up to 11B
DS2	Left of U13		Marked as DS2 on board (upper of 2 jumpers) - keep installed up to 11B

4.1.4 Combining the System 9 Sound ROM Images

Pinside user, "Dumbass", has created replacement System 9 MPU boards. One of the new things that he implemented in the board design is a daughterboard that can be mounted to the main MPU, which handles the speech duties. Therefore, the speech board with its sometimes problematic ribbon cable no longer has to be used. However, the sound and speech chip code now has to be combined together on a 27256 PROM, which is used at position U49.

The DOS/Windows "copy" commands to accomplish this for the System 9 games, are shown below (space formatting added for clarity). The file pad.732 is a special file and is not part of the released ROM images. It is a 4K (4096 Bytes) binary file filled with 0xff. Create it using your favorite hex editor, or download the file here: File:Pad.zip. Below is how the code should be concatenated for it to work properly. It is very similar to what is done for the System 9 code to be modified for a System 11 MPU.

Space Shuttle
Checksum = 0000h

copy /b pad.732     + spch_u5.732 + spch_u6.732 + spch_u4.732 + cpu_u49.128 = cpu_u49.256

Sorcerer
Checksum = E67Fh

copy /b spch_u7.732 + spch_u5.732 + spch_u6.732 + spch_u4.732 + cpu_u49.128 = cpu_u49.256

Comet
Checksum = F9CCh

copy /b spch_u7.732 + spch_u5.732 + spch_u6.732 + spch_u4.732 + cpu_u49.128 = cpu_u49.256

4.1.5 Converting the System 9 ROM Images to work in a System 11 MPU

It is easy to create System 11 compatible ROM images from System 9 ROMs.
The DOS/Windows "copy" commands to accomplish this for the System 9 games, are shown below (space formatting added for clarity).

The file pad.732 is a special file and is not part of the released ROM images. It is a 4K (4096 Bytes) binary file filled with 0xff. Create it using your favorite hex editor, or download the file here: File:Pad.zip.

The System 11 CPU works out of the box with a 27128 EPROM at location U27. You may choose to double the ROM image, creating a 27256 image so that its size is consistent with the other System 11 U27 ROMs.

Remove jumper W8 and install jumper W9 on the System 11 CPU to enable sound to work properly. System 11 controls the sound reset via software.

4.1.5.1 Space Shuttle

copy cpu_u49.128 cpu_u21.128
copy /b pad.732     + spch_u5.732 + spch_u6.732 + spch_u4.732 cpu_u22.128
copy /b cpu_u20.128 + cpu_u20.128 cpu_u27.256 <=== This command doubles the 128K image into a 256K image

4.1.5.2 Sorcerer

copy cpu_u49.128 cpu_u21.128
copy /b cpu_u21.128 + cpu_u21.128 cpu_u21.256
copy /b spch_u7.732 + spch_u5.732 + spch_u6.732 + spch_u4.732 cpu_u22.128
copy /b cpu_u22.128 + cpu_u22.128 cpu_u22.256
copy /b pad.732     + cpu_u19.732 + cpu_u20.764 cpu_u20.128
copy /b cpu_u20.128 + cpu_u20.128 cpu_u27.256
copy /b pad.732     + cpu_u19.l2.bin + cpu_u20.l2.bin cpu_u20.l2.128
copy /b cpu_u20.l2.128 + cpu_u20.l2.128 cpu_u27.l2.256 <=== This command doubles the 128K image into a 256K image
Sorcerer is the only System 9 pinball game using 2 ROMs which make things complicated for repair. You can use the cpu_u20.128 image in a system 9 CPU jumpered for Space Shuttle or Comet. If you install W12 and remove W11 the CPU will be jumpered for a 27128 EPROM.

4.1.5.3 Comet

copy cpu_u49.128 cpu_u21.128
copy /b spch_u7.732 + spch_u5.732 + spch_u6.732 + spch_u4.732 cpu_u22.128
copy /b cpu_u20.128 + cpu_u20.128 cpu_u27.256 /b <=== This command doubles the 128K image into a 256K image

4.2 Sound Boards

System 9 Speech Board Using 2532 PROMs (from Space Shuttle)

System 9 Speech Board Using 2732 PROMs (from Comet)

The speech board used in the three System 9 games, Space Shuttle, Sorcerer and Comet is essentially the same as the speech board used in System 6/7 games. However the speech board was enhanced during the Space Shuttle run. Four jumpers were added to accomodate either 2532 or 2732 EPROMs. When burning EPROMs, make certain which style of ROM is used.

System 9 Speech Board - Left unit is from early Space Shuttle, right is from later Space Shuttle

Only the early Space Shuttles use 2532 EPROMs from the factory. Later Shuttles, Sorcerer and Comet use 2732 EPROMs. The later speech board is backward compatible. If using such a board, either 2532s or 2732s can be used for the first two System 9 games (or System 6/7 games), provided that the jumpers are set for the appropriate EPROM.

In addition to the speech board generating speech sounds, it mixes the sound and speech signals via the potentiometer on the speech board.

Background Sound board as found in High Speed

D-11197 Background Sound board as found in Road Kings

D-11297 Sound board as found in Pinbot, fully populated

D-11298 Sound board as found in Pinbot. This version only has U4 populated (U19 is not). Uncertain why U19 is needed. Note that the CVSD "lane" of the board is not stuffed.

D-11581 Sound board, not completely populated

D-11581 Sound board, completely populated

D-11581 Sound Board, Completely Populated, with W10/W11 Jumpers. W10/W11 enable use of 512K sound ROMs as found in all System 11C games like Game Show, Dr. Dude, Pool Sharks, Rollergames, Diner, Radical, and Bugs Bunny

D-12238 Sound Board - specific to Jokerz! only

D-12238 Sound Board - specific to Jokerz! only - Connector PinOut

Use hyperlink at right to open D-11581 sound board schematics

Parts missing from partially populated D-11581 sound boards

Link to: Schematics for the D-11581 sound board.
Note: The parts shown at right represent the delta from a partially populated D-11581 to a fully populated D-11581. *The value of R13 will vary between 4.9Kohms (or 4.99Kohms) and 10Kohms. If R14 and R15 are 10Kohms, then R13 should be 4.9Kohms (or 4.99Kohms). If R14 and R15 are 20Kohms, then R13 should be 10Kohms.

4.3 Power Supply

Pre-Taxi Power/Regulator Board. This power supply has had the header pins replaced and the 5V filter cap at C10 replaced. Note that this board has the 20A fuse for the GI still installed. This indicates that it is an early System 7 board. The fuse and it's holder should be replaced by a wire in all System 9 and 11 games

WMS Power Supply - Revision B (Pinbot) Note that the top fuse holder just below the large 15000uF capacitor is broken on the left side

These power supplies use obsolete parts at Q1 (SDS-201) and Q3 (SDS-202). The PCB print was used from Black Knight through Swords of Fury, with slight modifications. Up until and including Fire!, Q1 and Q3 were the obsolete SDS-201 and SDS-202. Starting with some Fire!, the board uses an MJE15030 at Q1 and an MJE15031 at Q3.

Part numbers printed on an SDS-201 at Q1 may be "6557" or "9057S". Part numbers printed on an SDS-202 at Q3 may be "MDS60" or "9058S"

A closeup of the corrected trace routing error on a Fire! power supply board.

WMS Power Supply (High Speed) Schematics.

There was a trace routing error on the original Fire! power supply pcbs. WMS cut the trace and used a jumper to correct the error. This error was corrected in the PCB layout at the start of "Big Guns" game production.

WMS Power Supply - Revision E (Cyclone)

With Big Guns and games after, the solenoid and lamp power power sections were no longer stuffed because they were relocated to the Aux Power board. The GI relay is also missing on these boards because discrete GI relay boards were placed under the playfield and/or behind the backbox lamp insert. By doing this, specific GI strings could be turned on/off versus "all or none" of the general illumination. This later board has the fuses only for display and 5V power. In a pinch, the unstuffed parts could be installed.

Starting with Big Guns, 2 100mA (1/10A) SB fuses are installed just to the left of the power supply in the backbox. These fuses protect the +100V and -100V circuits against shorts. Not all games used 1/10A slo-blo fuses. Some games used 1/8ASB fuses instead. This may have been due to 1/8ASB fuses being more readily available.

The schematics for this power supply also list part numbers MPSD52 and MPSD02. The equivalent part number for an MPSD52 is a 2N5401. The equivalent part number for an MPSD02 is a 2N5551.

Taxi and After Power/Regulator Board

Schematics for Taxi and After Power/Regulator Board

This power supply was used from Taxi thru the remainder of the System 11 game production. Note that Taxi used an early revision of the board. The placement of some of the resistors and zener diodes in the HV section of the board is different than later board.

4.4 50V Flipper Power Supply Board

50VDC power supply as used in System 11 (nothing) and 11A games (C-9939).

System 11 (nothing) and System 11A games all used the 50V power supply board as pictured at left. This simple board takes 48VAC from the transformer secondary and uses a 35A/400V bridge rectifier to produce about 50VDC. The voltage is smoothed somewhat by a 100uf electrolytic capacitor. The transformer secondary is protected from a short in the bridge rectifier by either a 4ASB or a 5ASB fuse. Be sure to consult the manual for your game for the proper fuse value. Never overfuse a circuit.

4.5 Auxiliary Power Driver Board

At least two different revisions of the Aux Power Board were used in System 11B/11C games, pictured below.

Auxiliary Power Board from Space Station (D-11813-552)

Auxiliary Power Board from Earthshaker (D-12247-568)

4.6 Interconnect Boards

Backbox Interconnect Board from Space Station

Backbox Interconnect Board from Earthshaker

The interconnect board was essentially designed for ease of manufacturing and assembly. When Williams ultimately settled on using the same unit in all games, (give or take some parts here or there), modularity was apparent too. There was no longer a need to build complex, custom wiring harnesses between the CPU / driver board, backbox insert, and playfield. The interconnect board remedied this issue.

Backbox Interconnect Board from Banzai Run

Although the first four System 11B games, which include Big Guns, Space Station, and Cyclone use some form of an interconnect board, it wasn't until Swords of Fury when the interconnect board was used as a standard for the remaining System 11 games. The interconnect for Banzai Run (shown on the right) is completely different from anything else used by any game.

The interconnect board used in Swords of Fury uses a different part number, D-12185-559, compared to later System 11 games that used part number D-12313-x (x denotes the game's model number). They are not interchangable. Swords of Fury and Taxi had lane change switches still mounted on the cabinet. With later games lane change switches were removed and two optocouplers at position U2 and U3 (connected to the switch returns) were added to the board. The circuit is essentially the same as was used in early WPC games and enabled elimination of problematic physical lane change switches from the game. Starting with Jokerz, the third coupler at position U1 is used by the game software to check if the A/C relay on the aux power board is working. U1 transmits information back to the MPU as switch 2 on the switch matrix, provided that the A/C has energized.

The early versions of the interconnect board were located on the left side of the backbox and positioned vertically. The more common interconnect boards were located in the backbox positioned horizontally, just below the CPU board. The later board passes the following circuits:

• lamp general illumination circuits
• flipper power and ground
• lamp columns and rows
• switch columns and rows
• flash lamp power

What differentiates the board from game to game are the resistor values used for the flash lamp power circuits. In some cases, some of the resistors are left unpopulated and / or a zero ohm jumper is installed.

4.7 Master Display Driver Board

4.7.1 System 9 (C-8363)

C-8363 Early System 9 Master Display Driver Board

This is the same master display that is used on all System 6A and System 7 games and is used on most System 9 games.

4.7.2 System 9 (D-10749)

D-10749 Late System 9 Master Display Driver Board

This master display board was introduced during the System 9 run and is mostly found in some of the later Comets that were manufactured. The major difference between this master display board (D-10749) and the previous master display board (C-8363) is the D-10749 board replaced the card-edge connectors used on the C-8363 with pin-type connectors.

Any System 9 machine that is wired with card-edge connectors to use a C-8363 master display board can be converted to use a D-10749 master display board by re-wiring to use pin-type connectors.

The following table shows the mapping of the connectors & pins from a C-8363 to a D-10749 to aid in this conversion:

D-10749 Connector-Pin	Function	C-8363 Connector-Pin	Notes
J8-1	Strobe 16	J5-14
J8-2	Strobe 15	J5-16
J8-3	Strobe 14	J5-17
J8-4	Strobe 13	J5-13
J8-5	Strobe 12	J5-12
J8-6	Strobe 11	J5-15
J8-7	KEY
J8-8	Strobe 10	J5-2
J8-9	Strobe 9	J5-4
J9-1	Strobe 8	J5-3
J9-2	Strobe 7	J5-11
J9-3	Strobe 6	J5-10
J9-4	Strobe 5	J5-9
J9-5	Strobe 4	J5-8
J9-6	Strobe 3	J5-7
J9-7	Strobe 2	J5-6
J9-8	KEY
J9-9	Strobe 1	J5-5
J10-1	D1	J7-4
J10-2	C1	J7-2
J10-3	B1	J7-1
J10-4	A1	J7-5
J10-5	D2	J7-8
J10-6	KEY
J10-7	C2	J7-7
J10-8	B2	J7-6
J10-9	A2	J7-9
J10-10	Comma 1 & 2	J5-18
J10-11	Comma 3 & 4	J5-1
J10-12	Blanking	J7-3
J11-1	GROUND	J6-5
J11-2	KEY
J11-3	-100 VDC	J6-6
J11-4	+100 VDC	J6-2
J11-5	N.C.	J6-4
J11-6	+5 VDC	J6-3
	-100 VDC	J6-1	N.C. on D-10749

4.7.3 System 11 (Discrete Displays High Speed to Pin Bot and Banzai Run)

System 11 Master Display Driver Board

System 11 Master Display Driver Board with UDN7180 "daughter cards". Image courtesy of Aaron Gross.

The image at right shows "daughter cards" that replace UDN7180 chips, which were beginning to be become expensive and harder to obtain during the production run. The daughter cards use discrete transistors, much the same way as one version of the older System 3-6 Master Display Panel did. U9, U12, U13, and U14, originally UDN7180s, have been replaced on this board.

4.7.4 System 11A (Millionaire / Early F-14)

System 11A Millionaire Display Driver Board

This PCB actually came out of an early F-14 where they were recycled. Millionaire had it fully populated with additional lamp covers. The mounting brackets for the later D-11610 display changed so if you get a replacement for your F-14 first check which one is used in your game.

4.7.5 System 11A/B (F-14 to Swords of Fury)

System 11A/B 4 7 Digit Segmented Displays Driver Board

4.7.6 System 11B (Taxi and Police Force)

System 11B Display Driver Board D-12232-2 for 2 16 digit segmented displays + one external connector for a 7 digit display

Taxi and Police Force uses a D-12232-2 master display with J4 ribbon connector stuffed. If it's missing from the game, any other dual 16 digit alphanumeric can be used to make it work with Taxi and Police Force. You would need to add U3 (6118), resistors R10, 11, 12, 13, 15, 16, 17, 18 (10K) and connector J4. Jumper W1 should be added to connect the two center segments g' and m' on the lower display to form one large center segment g'. Jumper W2 should be added to light the comma on the jackpot display. Remove resistors R54, 56, 57, 60, 62, 64, 65, 67. The removal is these resistors makes the lower display essential a numeric display.

4.7.7 System 11B (all other WMS 11B/C games)

System 11B/C Display Driver Board D-12232-1 for 2 16 digit displays

4.7.8 System 11C (Riverboat Gamber)

System 11C Display Driver Board D-12232-3 for 2 16 digit displays and 2 connectors for 7 digit jackpot displays

Riverboat Gambler uses a D-12232-3 master display with all ribbon connectors stuffed. If it's missing from the game, any other dual 16 digit alphanumeric can be used to make it work with Riverboat Gambler. You would need to add U3 and U4 (both a 6118), resistors R10, 11, 12, 13, 15, 16, 17, 18 (10K) and connectors J4 and J6. Jumper W1 should be added to connect the two center segments g' and m' on the lower display to form one large center segment g'. Jumper W2 should be added to light the comma on the jackpot display. Remove resistors R54, 56, 57, 60, 62, 64, 65, 67. The removal is these resistors makes the lower display essential a numeric display. This display can also be used in Police Force and Taxi.

4.7.9 System 11B/C (all Bally 11B/C games)

WMS Bally left side 16 digit Master Display Driver Board

WMS Bally right side 16 digit Master Display Driver Board

4.8 Other Boards

4.8.1 Sound Overlay Board (C-13287)

Williams "C-13287 Sound Overlay Solenoid Board" as found in WhirlWind.

This relatively simple board was used on Whirlwind only to accommodate 5 additional solenoids (flashers or motors). The board logically sits between the System 11 MPU and the sound board, intercepting data normally sent directly to the sound board, and acting on that data only when specific codes are commanded. The board also relays data not specifically intended for it, on to the sound board.

4.8.2 High Current Driver Boards

Williams C-12493 High Current Driver Board from Earthshaker

The high current driver board was used on Earthshaker to drive the shaker motor. It was used on several other Bally/Williams WPC games while designated different part numbers.

4.8.3 Space Station Opto Position Board (C-11872)

Williams C-11872 Space Station Opto Positioner PC Board Assembly

This board is unique to Space Station. The "blobs" at Q1 and Q2 (both 2N3906s) in the picture at left are clear silicone caulk used to dampen vibration on the board.

4.8.4 Flash Lamp Resistor Board (Remotely Located)

Flash Lamp Resistor Board (from Pinbot)

The earlier design of flash lamp resistor board was mounted under the playfield within in the vicinity of the flash lamp circuit(s) that were powered by it. Each board had a maximum of 4 power resistors (2 per flash lamp circuit). The higher resistance resistor is 330 ohm 7 watt, while the lower resistance resistor is 5 ohm 10 watt. The purpose of the larger value resistor is to keep the flash lamp powered with minimal voltage. This keeps the flash lamp filament "warm", which puts less of a strain on the flash lamp. In turn, it is supposed to extend the life of the flash lamp. The smaller value resistor is used as a current limiting resistor which allows the solenoid voltage to be supplied to power the flash lamp. By powering the flash lamps with solenoid voltage, it eliminates the need for an additional, dedicated secondary transformer winding.

Note: If LED flash lamps are installed instead of incandescent lamps, the 330 ohm warming resistor must be removed from the circuit. If it's not removed the lamp will remain on, albeit not at full intensity.

4.8.5 Flash Lamp Resistor Board (Centrally Located)

• Flash Lamp Resistor Boards
• 

  D-11814-557(Big Guns)
• 

  D-11814-564 (Cyclone)
• 

  D-11814-552(Space Station)

The flash lamp resistor boards were used on early System 11B games prior to the use of the "standardized" interconnect board. The board is located in a very awkward location. It is screwed to the back of the lower cabinet. Big Guns, Space Station, and Cyclone all have this board in this location.

The resistors on the board "eat current" (limit current) since the 13V #89 flash lamps are wired to coil voltage (about 38VDC). When implementing this board, Williams chose to change the wire colors as the circuit passes through the board. Combined with the generally poor Williams System 11 documentation, this makes it tough to track down problems with non-working flash lamps. On the plus side, the wire wound cement resistors on the board do not fail often, probably due to the location of the board (resistor boards located under the playfield have high resistor failure rates due to heat and vibration). However, they obviously can fail. If there is power at the flash lamp, but grounding the flash lamp transistor does not light the lamp, the associated drive transistor may have failed.

The flash lamp resistor boards were abandoned after Space Station. Flash lamp circuits and resistors became integrated into the "standardized" interconnect board.

4.8.6 A/C Relay

A/C relay board (early design ie. F-14, Pinbot)

A/C relay placed on aux power driver board starting with 11B games (later design)

Starting with Road Kings, System 11 games make use of what is referred to as an A/C relay. The purpose of the A/C relay was to drive twice the amount of coils and flashers by using half the amount of drive transistors. This process is commonly referred to as "multiplexing". Typically, if the A/C relay was at rest, the "A side" solenoids would engage. When the A/C relay was powered, the "C side" flashers would engage. However, there were sometimes solenoids multiplexed with other solenoids.

For example, High Speed, which does not use an A/C relay, uses 8 discrete drive transistors to drive solenoids / flashers 1-8. Pinbot, which does use an A/C relay, uses the same 8 discrete transistors to drive solenoids / flashers. However, a total of 16 solenoids and flashers (A side and C side) are driven. By employing the use of an A/C relay, the System 11 MPU / driver board could be used for many more years without a "drastic" redesign.

Initially, the A/C relay was located on a small circuit board under the playfield for System 11/11A games. Likewise, this A/C relay board did not have Molex header connections. Wires were directly soldered to the circuit board. Once the Auxiliary Power Driver Board was introduced with the System 11B platform (starting with Big Guns), the A/C relay was moved to this board for all subsequent System 11B/11C games.

In the example of the A/C relay shown in the pic on the left, the only way to discern that the A/C relay is the highlighted relay board is due to the wire colors soldered to the board.

4.8.7 Diode Board for Multiplexed Solenoids

Diode board as found mounted to the bottom of the playfield (Pinbot)

Diode board from Pinbot

Diodes from diode board moved to aux power driver board starting with 11B games

4.8.8 Motor Control Board (D-12045)

Williams "D-12045" as found in Cyclone, Jokerz!, Riverboat Gambler and Diner.

This board is used in 4 Williams Games

• Cyclone ... for the random award wheel
• Jokerz! ... for draw poker (the wheel has various cards pictured on it),
• Riverboat Gambler ... for the backbox roulette wheel and
• Diner ... for the clock hands count time.

The board uses a single U-opto to detect the wheel at the "home" position. The MPU implements an "open loop control system" stepping the motor to the desired position while assuming that each step was successful. The game flags an issue with the PCB/motor assembly when the software thinks that the wheel should be at the home position but the opto beam hasn't been broken by the opto interrupter on the back of the wheel.

4.9 Flipper Coils

Serial wound flipper coils as seen on earlier (pre F-14 Tomcat) System 11 / 11A games (pic from Pinbot)

Wiring schematic for serial wound flipper coils

Early System 11 and 11A games used serial wound coils. These coils only have a single diode on them.

Placeholder for Parallel wound flipper coils as seen on later (post F-14 Tomcat) System 11A/B/C games (pic from xxxx)

Wiring schematic for parallel wound flipper coils

Later games (starting with F-14 Tomcat) used parallel wound coils, which have 2 diodes on them.

System 11 games use different parallel wound flipper coils dependent on flipper placement and application. Below is a chart of the flipper coils used, their wrapper color, and strength.

System 11 Flipper Coils
Part #	Color	Strength	Notes
FL-11753	Yellow	Weakest
FL-11722	Green	Weaker	24-600 series wound similar strength
FL-11630	Red	Standard	23-600 series wound similar strength
FL-11629	Blue	Strongest

4.10 Accessing Bookkeeping, Settings, and Diagnostic Modes

Example System 11 Coin Door Buttons. The Advance button is missing it's black cap, and yes...that is a bit of soft drink spooge on the bracket.

The coin door buttons for system 9 and 11 games are shown at right. The function of the three buttons is (from left to right in the picture)...

• Reset High Score to Date
• Auto Up/Manual Down
• Advance

Function of the Reset High Score to Date is self explanatory.

The Auto Up/Manual Down (AUMD) button is interpreted by the game software in multiple ways. Starting from attract mode, if the AUMD switch is up, pressing the advance button causes the game to enter audits and adjustments mode. If the AUMD switch is down, pressing the advance button causes the game to enter diagnostic mode. With the AUMD button in the up position during many of the diagnostic tests, the next step in the test will be automatically entered. For example, display test will display each digit on the display in turn. Coil test will advance through each coil. A particular step of the test may be paused on by pressing the AUMD button down.

The Advance switch moves the diagnostic, audit, or adjustment to the next step (or previous step if the AUMD button is down, except in diagnostic mode).

When in adjustments mode, the credit button is used to zero, turn off, or turn on features.
Note: One of the adjustment settings (adjustment 69 on High Speed for example) can be used to clear all of the game audits.

At the end of all diagnostic tests, the game will enter audits next, and then adjustments. To exit diagnostics, audits, and / or adjustments, either advance until the last adjustment, or simply turn the machine off.

4.11 Transformer Power Selection Connector

• System 11 Power Selection Jumpers
• 

  Typical power selection connection for a System 11 transformer - 115VAC (from Space Station)
• 

  European power selection connection for a System 11 transformer - 220VAC.
  1 - Black, 4 - White, 3/6 - Orange.
  Image courtesy of Chuck Sanderson

5 Problems and Solutions

5.1 Sporadic Problems in General - System 9

Cracked Header Joints and Headers cut too deeply by factory on System 9 MPU

Just like the previous Williams platforms, the System 9 platform is prone to cracked solder joints on the header pins. Likewise, Williams' board manufacturers had some tendencies to cut the header pins too short on the solder side. By cutting the pins too short, the cuts partially go through the solder meniscus. Cutting through the solder meniscus makes a solder joint less reliable, and it may prematurely fail over time.

By having cracked solder joints and / or cut solder meniscus, sporadic issues may occur. To resolve a cracked solder joint issue, the best approach is to completely remove the old solder, and add new solder to the joint. Due to dirt and contaminants, simply adding solder to the existing joint is not sufficient. If the header pins are cut too short, an attempt to remove the solder and new solder can be made. However, in some instances the only resolution is to replace the header pin connector. It is recommended to use a high quality replacement header.

5.2 Power Problems

5.2.1 Bridge Rectifier Fuses

Fuse holders with 8 amp slo-blo fuses added to one of the AC inputs to each of the bridge rectifiers. (Pinbot)

The updated schematic (from High Speed) showing placement of additional transformer secondary fusing for solenoid and controlled lamp bridge rectifiers.

A design flaw carried over from the earlier systems was the lack of fuses between the transformer secondary and the bridge rectifiers used for solenoid and controlled lamp power. In theory, if either of these bridges short, the main power fuse in the game should blow. But that is not always the case. If the primary power fuse fails to blow, the wiring between the transformer secondary and the bridge becomes the fuse. The resulting melted wiring is not pretty!

For games before Big Guns, interrupt one of the AC input lines and install a fuse holder with an 8 amp fuse installed (Williams opted to use an 8A normal blo, while Data East chose to install 8A slo-blo fuses). For games after Big Guns, an 8A normal blo fuse is already present between the transformer secondary and the controlled lamp bridge. As for the solenoid bridge, the addition of the auxiliary power driver board added fusing to protect this circuit. The 25V and 50V solenoid bridges were added to this board. Both bridges are fused with 4A slo-blo fuses.

A nice solution for this problem in the Inkochnito Bridge Board. Surf on over to http://www.inkochnito.nl. Click on the Bridge Board image for more information.

5.2.2 Alternate to the 12 position wafer connection

The severely burned wafer connector at Data East equivalent of J1 here has been replaced with parts currently available. Pins 5 and 8 are not populated as those connections are not used.

The 12 position "wafer" connector (CN1 or J1 for Williams games), can be replaced with currently available parts, shown at left. The original "wafers" are difficult to find and pricey. Part number for the alternate connector is Molex 03-09-2121 or 03-09-2122.

5.2.3 Rottendog Power Supply Missing 5v Failure

If an aftermarket Rottendog WDP011C power supply is missing 5v, and other common causes of a missing voltage have been ruled out, then it's likely that the 5v regulator at U191 has failed. An available replacement is the Texas Instruments PTR08060 5v regulator from Mouser or Digikey.

5.2.4 Rottendog Power Supply - Big Guns & Cyclone GI Comes On, Game Doesn't Boot

If using the now obsolete Rottendog WDP011A power supply for Williams Cyclone or Big Guns, the +12v jumper may have to be moved from the bottom jumper location to the top. If a Rottendog equipped Big Guns or Cyclone turns on the GI lights, but the game doesn't boot, (only the 5v light on the CPU is lit), this is most likely the reason.

5.3 System 11 MPU Jumpers

Normally, jumpers on the MPU board do not require modification. The default configuration works for all System 11 game ROMs. However, there are 5 exceptions as follows...

W7 - Language Select Jumper

W7 is used only in games up to Cyclone. Games with 16 digit alphanumeric displays do not use this jumper. If installed, the game automatically reverts to English after losing the content of the RAM. If a different language is preferred, the "adjustments" menu provides an option to change the default language. Only a few early System 11 game ROMs contained multiple languages. For these games, if W7 is removed, the game generally defaults to German text.

Later System 11 games used language specific ROMs at U26 and U27.

W16 and W17 - Hitachi CPU select

A Hitachi processor, installed at U15, with W16 cut (lower right of pic)

The W16 jumper, on original System 11 boards, is in a different location and used for backwards compatibility with system 9 boards. There is no need to change it.

On System 11A and later boards:

• W16 is connected to pin 38 (XTAL) of the 6802/6808 at U15, the CPU responsible for most game operation.
• W17 is connected to pin 38 (XTAL) of the 6802/6808 at U24, the CPU for the sound subsystem.

These two jumpers are used to indicate the manufacturer of the 6802/6808 microprocessor since the clock circuit of Hitachi microprocessors is different from the clock circuit of other manufacturers.

If U15 is a Hitachi microprocessor, remove or cut W16. If U24 is a Hitachi microprocessor, remove or cut W17. Otherwise, W16 and W17 should remain installed.

A Hitachi microprocessor can be used in System 9 and System 11 (nothing) MPU boards by simply lifting pin 38 out of the socket. Cutting the pin is not recommended since there may be need for that IC in another board.

W5/W6 and W18/W19 - RAM Type Select

System 11B and 11C MPU board provide the ability to use either a 6116 or a 6264 RAM at locations U25 and U23 (sound subsystem).

For U25...

• install jumper W5, remove W6 to use a 6116.
• Remove jumper W5, install W6 to use a 6264.

For U23...

• install jumper W19, remove W18 to use a 6116.
• Remove jumper W19, install W18 to use a 6264.

5.4 MPU boot issues

5.4.1 Steps for Repairing a Dead MPU

A scope trace of the 4Mhz crystal input at pin 39 of the microprocessor (blue), and "E", the external clock at pin 37 of the microprocessor (amber)

Following is an ordered, disciplined process for reanimating a dead MPU board.

1. Inspect for Alkaline Corrosion. Abate all corrosion properly before continuing.
2. Test for valid 5/12/-12VDC power to the board.
3. Verify correct jumpers are installed (or not installed) for Hitachi 6802/8 processors. More info here.
4. Note and address any error messages displayed upon boot (if any).
5. Validate the ROM images.
6. Test socketed chips in a known working board or chip tester.
7. Verify a proper (4Mhz) clock signal to the 6802/8 processor at pin 39.
8. Verify a proper (1Mhz) external clock signal at pin 37 of the 6802/8.
9. Verify the reset circuitry is providing a valid "low and then high" reset signal at pin 40 of the 6802/8. Note that replacing the original reset circuitry with a modern reset generator will lengthen the time until the signal transitions to high. This is normal, and helps with debugging as the signal can be seen much easier.
10. Use a logic probe to test for activity on the IRQ (pin 4), Address and Data pins of the 6802/8.
11. Use a logic probe to ensure that the bus buffer chips at U11 and U13 (address, R/W, and external clock), and U16 (data) are correctly mirroring the input signals to the output signals.
12. Install Leon's test ROM to ensure that each 6821 PIA is receiving the correct chip select and data to produce valid outputs at pins 2 through 17. More information regarding use of Leon's test ROM can be found here.

5.4.2 Normal Game Boot Behavior

The 7-Segment diagnostic display on a System 9 MPU

Normal MPU boot behavior for System 9 is to show a "0" in the diagnostic display on the MPU. The "0" will light during boot up, and continue to stay lit, as long as the game is on.

The 7-Segment diagnostic display on a System 11(nothing) MPU

Normal MPU boot behavior for games with a 7-segment display on the MPU is to show a "0" in the display. The "0" does not go away. The "0" will continue to be displayed under normal boot conditions.

+5 VDC, Diagnostics, and Blanking LEDs as found on a System 11A, 11B, and 11C MPUs

Normal MPU boot behavior for games with 3 LEDs instead of the 7-segment display on the MPU is for the "+5 VDC" LED to light first and stay on, then a split second later, the blanking LED will light and stay on. At the same time that the blanking LED lights, the diagnostic LED will begin to blink at a fairly rapid pace with a 50% duty cycle (equal durations of the LED being on then off). The blanking LED will continue to remain on.

+5 VDC, Diagnostics, and Blanking LEDs driven by game software expecting a 7-segment display as in High Speed

If a System 11 MPU with 3 LEDs is installed in a game whose original board contained the 7-segment LED diagnostic display first introduced with the System 7 board set, the status of the 3 LEDs won't help much. The game software is attempting to display a "0" (assuming your game boots correctly) on a 7-segment display. Since the board contains just the three LEDs, the MPU circuitry merely lights all three LEDs as shown in the picture at left.

5.4.3 MPU Boot Error Codes

MPUs used in System 9 and 11(nothing) games display a code on a 7-segment display if an error is detected during boot. MPUs used in System 11A, 11B, and 11C games use the Diagnostic LED to blink a "codes".

5.4.3.1 System 9 Games

The following table lists the error codes displayed on the System 9 MPU 7-segment display.

Code	Description
0	Normal game boot with no problems detected. The game should be in attract mode.
1	CPU board locked up. Possible cause is memory protect circuit and U18 CMOS RAM "stuck bits".  That is, an actual failure of the CMOS RAM chip or the 6808 processor's ability to communicate error free with the RAM.
2	U20 Game ROM 1 faulty
3	U20 Game ROM 1 faulty
4	U19 Game ROM 2 faulty
5	Blanking signal stuck, coin door closed, memory protect circuit faulty, or the U18 CMOS RAM faulty
7	System failure. Occasionally, the following components can contribute to this problem: U21 (4MhZ crystal), components in the IRQ circuit, broken leads or failed C9 (22uF) in the reset section, or loosely seated ICs on the CPU board. Note that boot code "7" is the hardware default and characteristic of the CPU "reset" signal not complying with the proper low-to-high spec at pin 40.
Other or no indication	U20 Game ROM 1 faulty

If the game is locked up, there may be a bad power supply. The 5 volt logic supply is the first one to check. Then, check the 12 volt supply, which is used to reset the CPU chip. Test for both DC (correct voltages) and AC (too much ripple on the DC circuit).

5.4.3.2 System 11(nothing) Games

The following table lists the error codes displayed on the System 11(nothing) MPU 7-segment display.

Note: These error codes assume that either High Speed, Grand Lizard, or Road Kings ROMs are installed since those were the only three games shipped with this version of the MPU board containing the 7-segment display. Later System 11 MPU boards replace the 7-segment display with 3 discrete LEDs that indicate 5V presence, diagnostics execution, and blanking status. While the System 11(nothing) MPU board works fine with other/later game ROMs, the 7-segment display with later game ROMs doesn't provide useful information other than some confidence that the software is running correctly. With later game ROMs installed and the MPU booting properly, the 7-segment display will appear to alternate rapidly between "0" and "7" (as fast as the "Diagnostics" LED would be flashing in a System 11A through 11C MPU board).

Code	Description
0	Normal game boot with no problems detected. The game should be in attract mode.
1	CPU board locked up. Possible cause is memory protect circuit and U25 CMOS RAM "stuck bits".  That is, an actual failure of the CMOS RAM chip or the 6802 processors ability to communicate error free with the RAM.
2	U27 Game ROM checksum failure
3	U26 Game ROM checksum failure
4	not used
5	Blanking signal "stuck", or the coin door is closed, or the memory protect circuit is faulty, or the CMOS RAM at U25 has failed.
Other or no indication	General system failure. Check the 5VDC power supply as well as the integrity of game ROM 2 at U26.

A zero displayed during Memory Chip Test (using the CPU board switch SW2) indicates that the blanking circuit is NOT functioning properly.

An eight displayed during memory chip test (using the CPU board switch SW2) indicates that the blanking circuit is functioning properly.

5.4.3.3 System 11A, 11B, 11C games

The following table lists the number of blinks of the DIAGNOSTIC LED, the error message that might be displayed, and the explanation (Source: Williams System 11A game manual. with embellishment).
A video demonstrating an MPU indicating a U10 failure (8 blinks) can be found here.

Code	Message	Notes
0	no message	Normal game boot with no problems detected. The game should be in attract mode.
1	U25 RAM FAILURE	The RAM at U25 could not be used properly. This means that the RAM failed a read/write test at power up.  The game will remain in the current state and no other tests are performed until the game is turned off, then back on.
2	MEM. PROT. FAILURE	This message means that either:    a) the Coin Door may be shut,    b) the Memory Protect Switch may be stuck in the ON position,    c) the memory protect logic is protecting the memory, or    d) a U25 RAM failure is occurring.
3	U51 PIA FAILURE	The PIA at U51 has failed.
4	U38 PIA FAILURE	The PIA at U38 has failed.
5	U41 PIA FAILURE	The PIA at U41 has failed.
6	U42 PIA FAILURE	The PIA at U42 has failed.
7	U54 PIA FAILURE	The PIA at U54 has failed.
8	U10 PIA FAILURE	The PIA at U10 has failed.
9	IRQ FAILURE	The interrupt request line is not working normally. It may be missing, stuck, too fast, or too slow.
10	U27 ROM FAILURE	U27 failed checksum (test 11 is skipped).
11	U26 ROM FAILURE	U26 failed checksum.

All of the above failures (except #9) may mean that the device has failed but it may also mean that the 6802 microprocessor is unable to communicate adequately or properly with the device. For example, if the CPU is attempting to checksum the game ROM at U27, but one of the 8 data lines leading to U27 has been severed, the wrong data will be received by the microprocessor and the checksum test will fail even though the ROM is perfectly fine.

An U25 RAM failure can occur without actual 6116 RAM failure. One of the most common reasons why U25 RAM failure occurs is lack of power to the RAM chip. If the D1, 1N5817, diode shorts open, logic power will no longer have a pathway to the U25 RAM. However, if the D2, 1N4148, blocking diode is fine, and the battery holder and batteries are fine, power will be retained by the U25 RAM. Although, the batteries will die prematurely if D1 has shorted. Another reason why U25 is no longer powered by the +5VDC logic power is due to a severed trace either caused by a blunt object or battery corrosion.

An interesting side effect to the U25 RAM failure blink code is that the blanking LED will blink 10 times after the single blink from the diagnostic LED.

5.4.3.4 U51, U27, and possibly other boot message causes

This OEM System 11 MPU socket is tired, worn out, and no longer provided solid contact with the ROM chip legs.

System 11 MPUs booting with error codes, blinks, or messages on the display may need new sockets for the game ROM(s). In the picture at left, it's easy to see light right through the socket holes (note the brown background vs the white background). The metal female connections have fatigued over time and are no longer providing a solid mechanical connection to the ROM chip legs. Removal and replacement with a quality dual wipe or machine pin socket is advised.

This situation can cause a wide variety of boot problems including complete failure to boot, "U27 ROM failure" message, "U51 failure" message, etc.

Of course, the root cause could be due to many other reasons. The fairly complicated chip select logic outlined on sheet 4 of the System 11 MPU schematics could be the cause. The best way to attack the problem is with a logic probe, ensuring that the various chip select gates are yielding the intended logic states.

5.4.3.5 Five "Knocks"

Sometimes when turning a system 11 game on, the knocker will fire 5 times, causing most folks to wonder, "what the hey"?. All that knocking is designed to call the operators attention to a potential switch issue. Like the "credit dot" in later Williams games, including System 11 games that support the dot character, these 5 knocks mean that the game hasn't detected the closure of a particular switch (or switches) in a certain number of games. A message will be displayed indicating which switch or switches to examine.

Usually, this is an indication of a failed switch. Most switches get closed during the normal course of game play, long before the game's software counts up to the limit of "games played without closing each switch" and issues this indication.

To address this issue, use the coin door diagnostic panel buttons to enter switch edge test. Manually close the switch. If a switch closure is indicated, either normal game play failed to close the switch (you're not a very good player), or the switch is dirty. Closing a switch with your finger is different than closing a switch with a ball. Your finger generally pushes more firmly on the switch contacts. Clean the switch by dragging an old business card or piece of paper between the switch contacts while pinching the switch blades together with your fingers.

Caution: do not attempt to adjust switches with anything metal while the game is turned on. You will eventually short coil or lamp power to the switch matrix, damaging the MPUs switch matrix circuitry.

If closing the switch manually doesn't register in switch edge test, see the "switch problems" section of this Wiki.

5.4.3.6 System 11 "Adjust Failure" and "Factory Setting"

The typical "Adjust Failure" message on a System 11 game, coin door closed.

The "Adjust Failure" message seen in the picture at left occurs when the game software can't make sense of the information contained in the battery backed RAM at U25. This could be because the RAM has failed, but this is unlikely. The vastly more common reason is that battery backup power has been interrupted to the games battery backed RAM because either the batteries are not installed, dead, or dying. Hopefully, they haven't made a mess of the board by leaking alkaline. See below. "Adjust Failure" is displayed under these conditions when the coin door is closed.

Note that on some games (not certain if all) which use two 16-digit alpha-numeric displays, the message displayed will actually read "AJSTMENT FAILURE".

The typical "Factory Setting" message on a System 11 game, coin door open.

If the coin door is open under these conditions, the message "Factory Setting" is displayed.

5.4.3.7 Testing for Battery Backup Power at the RAM (U25)

Game off. Meter set to DC volts. Black lead on game ground. Red lead on RAM pin 24.

Testing for power at the RAM chip is easy.

1. Set DMM to DC volts (or on less than 20V if not using an auto-ranging meter)
2. Game off
3. Black lead on backbox ground
4. Red lead on pin 24 of the RAM as shown at right

Three AA batteries normally provide 4.5VDC. If you are reading some voltage, but not quite 4.5VDC, the batteries need to be replaced...and don't forget to move them off of the MPU board while you are at it.

Other causes for low or no battery backup power to the RAM chip are:

1. Failed diode D2. This blocking diode normally prevents the 5VDC source from "charging" the alkaline batteries. D2 is a 1N4148.
2. Failed diode D1. This blocking diode normally prevents the 5VDC battery backup power from powering parts of the board other than the RAM chip. In this case, the batteries will deplete rapidly. D1 is a 1N5817.

5.4.4 Cross Connecting Black and White Backbox Connectors

The infamous System 7/9 connectors that can be cross connected. Yikes!

In one of the biggest mistakes in pinball manufacturing, Williams used two identical 24 pin .062 connectors to connect cabinet and backbox wiring. The connectors are identical with the exception that one is black and one is white. During manufacturing, the wrong color connector was sometimes used. Bottom line: do not simply mate black to black and white to white. Instead, ensure that the wire colors going into one side of the connector are the same as the wire colors coming out the other side. This is the only foolproof method of avoiding the catastrophe of cross connecting circuits, causing certain damage to the MPU board.

For System 9 MPUs cross-connected, the damage can include, but is certainly not limited to, U41, U9, the game ROMs, and one or more 6821s.

For System 11 MPUs cross-connected, the damage can include, but is certainly not limited to, U11, U31, U34, U35, U36, U44, the game ROMs, and possibly the 6821s.

A YouTube video demonstrating the repair of a System 9 MPU that was damaged in this way can be found here.

An excellent disciplined approach to repairing a System 11 MPU damaged in this way can be found on PinSide here.

5.4.5 Relocating the battery from the System 9 MPU board

Yikes! What a pity...

Relocating the 3 AA batteries used to retain settings across power cycles to a location far away from the MPU board is always a good idea. Leaky alkaline batteries are the #1 killer of pinball boards. Sometimes the battery terminals don't look corroded, but the metal rivet which contacts the battery are actually missing.

If "04 00" in the credit/match display is seen when the game is turned on, the game is in audit mode versus attract mode. Below are several reasons why the game has defaulted to audit mode.

• The batteries have failed and need replacing.
• The battery voltage is not reaching the U18 (5517-2) RAM. Check pin 24 of U18 for approximately 4.3v with the power on and 3.9v with the power off. Lack of battery backup power could also be due to an open D3 (1N4148) blocking diode. This diode is used to keep the CPU's logic power from charging the batteries.
• Blocking diode D2 (1N5817) has shorted, and the batteries are trying to run the MPU board when the game is off.
• There are other problems, such as a faulty 5517-2 RAM.

Simply removing the batteries will not allow a game to boot directly into "attract mode" when switched on. It also will not retain the settings such as the number of balls per game, the free play setting (obtained by setting maximum credits to 0), or high scores. However, a System 9 game will boot, and is still playable with the lack of batteries. To complete the boot sequence into attract mode, open the coin door, switch the game off, and then quickly back on. The game should leave audit mode, and go into attract mode. Credits would need to be added from the coin door, and if necessary, settings would need to be changed before starting a game.

The best option is to remotely locate the battery holder somewhere below all the other boards. This ensures that even if the remotely located batteries leak, they won't leak onto (or even drip onto) components of the MPU board. Use good quality alkaline batteries, mark the date of replacement with a Sharpie, and replace the batteries annually.

Adding a connector between the battery pack and the MPU board is a good idea. You can easily remove the battery pack from the board. Plus, if the batteries are forgotten, and do leak, the MPU board will not have to be removed to add another battery pack. A 3 x AA battery holder is the typical recommended replacement. If only a 4 x AA battery holder is available, a jumper can be soldered in the first battery position. Likewise, a diode can be placed in this position instead. This will prevent the batteries from being charged and 'cooked' by the game if blocking diode D3 on the MPU board fails. Keep in mind that adding a secondary diode to this circuit will decrease the voltage passing to the RAM memory by .5 to .7 volts (the typical voltage drop across a diode) if D3 is still good. Install a 1n4001 or 1N4004 diode in the position closest to the last + terminal (where the Red Wire exits). The banded side of the diode must be pointing in the direction of current flow, which is towards the (+) terminal marking on the MPU board, and away from the battery pack.

Williams System 9 MPU Board

On the System 9 MPU, solder the battery cables: Ground (Black Wire) to the Bottom Left pad and Positive (Red Wire) to the Top Right.

After adding a remote battery pack, and while the board is still out of the game, it is a good practice to measure the battery pack's voltage at the (+) and (-) pads of the MPU board. All battery packs are pretty cheaply made, and failures "out of the box" are somewhat common. Checking to make certain the battery pack is functioning before reinstalling the MPU board in the game will save you some headaches.

D2 and D3 Diodes Highlighted on a Williams System 9 MPU Board

Since the MPU board is already out, another good practice is to check the D3 blocking diode. An open blocking diode will not allow the battery pack voltage to pass through to the non-volatile memory, and the newly installed battery pack will be ineffective. Conversely, a shorted blocking diode will allow the board's +5vdc logic power bus to pass through to the battery pack. This in turn, will charge the batteries, while the game is turned on. Alkaline batteries do not like being charged. They will heat up, and fail prematurely, (rather quickly). In worse cases, the new batteries can even leak or explode if charged. Testing the D3 diode is quick and easy, and worth the trouble checking it out. When in doubt, replace the D3 diode with a 1N4148, or add a secondary 1N4004 to the battery pack. Once again, if a secondary diode is added, it will decrease the voltage passing to the 5517-2 RAM memory, if D3 is still good.

5.4.6 Relocating the battery from the System 11 MPU board

A remote battery holder installed on a System 11 MPU. The purple wire is positive.

5.4.7 Installing NVRAM instead of batteries

RAMTRON DIP NVRAM installed on a System 11B board (Cyclone)

SMT adapted NVRAM installed on a System 11B board (Transporter The Rescue)

Like most other pinball mpu boards, the battery-backed RAM can be replaced with a non-volatile memory RAM (NVRAM). Replacement RAM that can replace existing 6116 or a 6264 SRAM will be needed. DIP versions are becoming scarce, but there are other solutions available that use surface mount equivalents on a small circuit board that plug into the RAM's socket. Unfortunately, U25 is soldered in on System 11 boards from the factory, and would need to be removed to replace with NVRAM.

If installing any version of 6264 NVRAM on an 11B or 11C board, jumper W5 (located just above U25) needs to be moved to jumper W6. This jumper has a completely different purpose on 11 and 11A boards. 6116 NVRAM will have to be used with 11 and 11A boards, or else modification with wire wrap will be necessary. If 6116 NVRAM is used on 11B or 11C boards, jumper W5 does not have to be chaged.

There are several vendors who sell surface mount NVRAM on an adapter board. The adapter in the picture shown on the right was purchased from Pinitech.

It has been documented that some variants of DIP NVRAM may need to have diode D1 replaced with a jumper or jumped around the diode, in order for the game to boot successfully. D1, a 1n5817 diode, has a very low forward voltage drop (about 0.2 volts) vs. the normal 0.4-0.6 volts most other diodes have. If replacing U25 with the memory RAM, some memory RAM will not unlock and allow writes until the voltage is 4.8 volts or so. The 1n5817 D1 diode is just enough to prevent some memory RAMs from allowing writes. To solve this, solder a jumper wire around the D1 terminals, or remove D1 entirely and replace with a jumper. Surface mounted NVRAM on an adapter board is not susceptible to this problem.

5.4.8 Installing a Memory Capacitor Instead of Batteries

Memory Capacitor Added to a System 11B MPU Board. The circled areas show an added jumper for the cap's negative terminal to ground and a jumper added in place of the blocking diode.

An alternate method of adding a SuperCap. This style cap has two leads on the bottom of the cap. The negative lead of the cap is connected to battery ground via the red jumper wire. The necessary jumper across D2 is circled.

Another alternative is to install a memory capacitor. In essence, a memory capacitor is similar to a rechargeable battery, although, the likelihood of a memory capacitor leaking is greatly reduced compared to a rechargeable battery. When the game is turned on, it is charging the capacitor. When the game is turned off (this is where the memory cap slightly differs than a rechargeable battery) the memory capacitor slowly loses its charge over time. Therefore, it is imperative that the game periodically be turned on to allow the capacitor to charge up to its full capacity again. If a game will not be turned on for long lengths of time, a memory capacitor may not be the best solution.

When installing a memory cap, two things will have to be done. It will be necessary to add a jumper to tie the negative lead of the cap to ground or positive lead to the non-banded side of the blocking diode. The picture to the left shows a jumper added to tie the negative lead of the cap to ground on a System 11B MPU board. Installation on System 11, 11A, and 11C boards is similarly performed. Secondly, the 1N4148 blocking diode (D2) will have to be removed, and replaced by current limiting resistor. A value from 100 ohms to 270 ohms will work. Once the capacitor is installed, and the board installed back in the game, it is a good idea to leave the game on for 15-20 minutes to allow the game to initially charge the capacitor to full capacity. After that, turning the game on monthly for about 10 minutes to allow the cap to recharge is a good idea.
Note: Both photos show a jumper installed in place of the blocking diode. It is recommended that a limiting resistor be installed instead.

Note: It has been determined that memory capacitors are not the most effective way to retain memory on System 11 games, if the game will not be turned on for long lengths of time.

5.4.9 Repairing Alkaline Corrosion

Closeup of the delicate traces under the battery holder of a System 11 MPU. Luckily, these cleaned up nicely.

Closeup of traces under the battery holder. Corrosion (darkened spots) has invaded beneath the solder mask (green coating) and is beginning to eat away at the copper traces.

As can be seen at left, the traces on System 11 MPUs are very small. Not much alkaline damage is required to eat completely through these delicate traces. Unfortunately, there are a great many of these traces under the battery holder, and in a very vulnerable position to alkaline damage.

Also, to the left of the battery holder is the reset section of the MPU. That section is, on balance, much easier to repair than the delicate traces under the battery holder and at the 6821 PIAs immediately under the battery holder.

The advice often dispensed to deal with alkaline corrosion is to scrub the affected area with an acid such as vinegar, rinse with water, and then rinse again with isopropyl alcohol. This might work for light surface alkaline corrosion.

Unfortunately, often the alkaline will eat into traces beneath the solder mask or beneath chips. This dictates desoldering affected chips to remove the corrosion and repair traces. The alkaline interacts chemically with the solder making it very difficult to heat. Desoldering from the (presumably undamaged) back side of the board is much easier. Scuffing the alkaline damaged solder with a wire brush to remove the surface layer of damage is also helpful.

Once all damaged components have been removed, sand the affected area carefully with your choice of fine grit sandpaper. 220 grit works well. The objective is to remove the corrosion completely without destroying additional traces. All corroded areas must be addressed.

Next, rebuild traces as necessary. Corrosion tends to sever traces where the trace meets the solder pad of a part. Strong magnification and bright lighting helps. Either tin or conformal coat the bare copper and then reinstall parts.

Another closeup of the delicate traces near the battery holder of a System 11 MPU. These traces are too far gone for a repair to be considered reliable.

A 9 pin, 4.7K bussed resistor network replacing SRC5. Pin 1 is on the right

A closeup of alkaline damage that has crept under the solder mask (the darker traces). Eventually, this corrosion will eat completely through the trace, making the board unreliable.

The OEM resistor/capacitor network.

At the top of the board are a few "SRC" parts. These are 10 pin resistor/capacitor SIPs. The OEM part connects each driver line to a 4.7Kohm pull-up resistor connected to 5VDC. This guarantees that the signal is held at logic high until the PIA grounds the signal. The OEM part also connects each driver line to a high frequency noise filter capacitor (470pf). These parts are no longer available. A completely acceptable substitute is a 9 pin bussed, 8 resistor network (4.7Kohms) at positions SRC1 - SRC5 and SRC7 - SRC9. This part provides the pull-up resistor. In practice, the high frequency filter capacitor does not impact the displays.

The part is installed as shown in the picture at right, with pin 1 of the part connected to the right side which is labeled pin 1 in the board silkscreen. A 10 pin (9 bussed resistors) part may be used ONLY if the left most pin is clipped and not connected to ground.

An image of the original OEM part is shown at right.

Another example of alkaline corrosion can be seen in the picture on the left. This damage can be cleaned up, retinned, and the board placed back in service.

5.4.10 Blanking Signal Stuck Low or Pulsing

The blanking signal is a failsafe interlock with critical game functions including the display, sounds, lamp matrix, and solenoids. It prevents operation of each of these game features to protect the game circuitry in the event of a system failure. For instance, if the MPU is not running, and the lamp matrix is stuck on at a particular row or column, those parts will heat up quickly, and at the very least damage that circuitry. Coils remaining locked on could cause even larger damage to the game.

Each of the game functions listed above require the blanking signal to be high (not pulsing but always high) to operate. As an example, the solenoid driver circuitry logically "ands" the blanking signal with solenoid commands using a 7408 AND gates (at U17, U18, U19 and U20).

Lack of blanking, with the blanking LED never lighting, means that the game program is not running. It either never started running, locked up due to an invalid ROM image or a problem with the data, address, control, or chip select signals on the MPU, or detected a problem with the RAM memory.

C58, a timing element for the 555 timer. Image Courtesy of Chris Hibler.

In rare cases, C58 can fail. C58 holds the input to the 555 timer high between pulses from the display circuitry. This causes the blanking signal to pulse. A pulsing blanking signal can manifest in missing digits on the display, some lamp columns not being turned on, and possibly other issues. C58 is a 1uf tantalum capacitor. If the blanking signal is pulsing, replace C58.

Williams System 11 CPU Blanking Circuit. Image Courtesy of Chris Hibler.

The blanking circuit explained

1. A properly executing CPU will send data to the display 6821 PIA at U51 (pin 4) at regular intervals (every 2.8 msec).
2. This action causes PA2 at pin 4 of the 6821 to pulse high periodically. The DE schematics call this signal "TG". It is merely called "blanking" in WMS schematics.
3. The signal is inverted by the 7404 at U36, in at pin 13, out at pin 12.
4. The coupling capacitor at C57 can be ignored. These ceramic caps rarely fail.
5. The signal is applied to the base (pin 2) of a 2N4403 at Q50 and to pin 2 of the 555 timer at 1C.
6. The 555 timer is configured as a "missing pulse detector" meaning that if it does not detect the pulse that originated upstream at the 6821 PIA at U51, the output of the 555 timer at pin 3 (and 7) will go low, which invokes blanking and protection of the system circuitry.

For the longer, more detailed explanation, Stern service bulletin 75 provides the "Blanking Circuit Theory of Operation" for Data East MPUs which are identical to Williams MPUs with respect to the blanking circuit.

When operating properly, the pins of the 555 timer will behave as follows (pin number, logic signal)

1. low
2. pulsing at the same rate as the 6821 PIA at U51, pin 4
3. high
4. high
5. pulsing very rapidly. i.e. "ticking"
6. high
7. high

If Leon's System 11 test ROM runs correctly on an MPU, this guarantees that all data, address, and control signals are working correctly. If blanking doesn't go high with a game ROM installed, ensure that the ROM images are valid. This is by far the most probable once Leon's ROM runs correctly.

5.4.11 An Alternative to Rebuilding the Reset Circuit / Using a Reset Generator (System 9)

Reset section of a System 9 MPU board.

The image above shows the single jumper and the exact minimum parts required when using a reset generator on a System 9 MPU. The three electrolytic caps (all 100uf/25V), two inductors, and 5 ceramic caps (all .001uf labeled "102") are part of the 5, 12, and -12 volt power filtering circuitry. The D2, 1N5817 diode isn't needed if an NVRAM is used instead of a SuperCap or remote batteries. However, a jumper must be installed at D2 to power the NVRAM.

The reset section of the System 9 MPU is almost identical to the System 11 reset section. As such, a modification similar to the mod outlined below can be performed on System 9 boards too.

1. Remove resistors R6, R8, and R12. Really, all but R6 can simply be clipped off. One through hole at R6 will be used later.
2. Remove transistor Q4
3. Install a reset generator (MCP120-460GI/TO) where Q4 was removed. Orient the reset generator the same way that Q4 was oriented, matching the PCB silkscreen.
4. Add a jumper from the upper leg of where R6 was removed (oriented as shown in the picture, upside-down from game installation) to the via that is about 1/4" southeast of the R6 connection.

That's it. Most of the rest of the reset circuit can be removed, but this provides no advantage since those parts are already "out of circuit" with this modification. After this modification is performed, the MPU board will no longer need the +12v to successfully boot. However, +12v is still necessary to power the sound amplifier on the board.

Another image of the bare minimum part requirement when using a modern reset generator. The three electrolytic caps are all 100uf/25V. Note that the right most electrolytic cap is oriented backwards from the other two. The 5 ceramic caps are all .001uf and are labeled "102". These are sometimes called 1000pf ceramic caps.

The MCP130 reset generator variant, with an internal 4.7K ohm resistor can be used too, but it isn't necessary. This is because the resistor at R170 provides an external 4.7K ohm pull up for the /RESET line.

Note that the reset generator has a longer reset delay than the OEM circuitry. At boot, the 7 segment LED will seem to show a "7" for a fraction of a second. It will turn to a "0" and remain "0" after that, assuming that the board is booting properly.

Factory installed "zero ohm jumpers" where inductors are normally installed on a System 9 MPU board. Image courtesy of John Wart Jr.

Factory installed "zero ohm jumper" where an inductor is normally installed on a System 9 MPU board. Image courtesy of John Wart Jr.

Note that some System 9 MPUs shipped from the factory with zero ohm jumpers installed instead of inductors (see images left and right). When repairing alkaline corrosion, it's completely acceptable to replaced damaged inductors with jumpers.

5.4.12 An Alternative to Rebuilding the Reset Circuit / Using a Reset Generator (All versions of System 11)

Reset section of a System 11A MPU board. Components to be removed are circled.

Reset section with reset generator and jumper installed.

There is an alternative to the stock reset circuit on all versions of the System 11 MPU boards. A reset generator can be installed. Since the reset section of all versions of the System 11 MPU board is susceptible to battery damage, this is a very easy, and minimalistic way to rebuild the reset section. To do so, 4 components will need to be removed and a jumper installed. The following work will have to be done.

1. Remove the Q39 transistor.
2. Remove the following resistors - (from top to bottom as installed on the board) R58, R60 (these two are just below the Q39 transistor), and R69.
3. Add reset generator MCP120-460GI/TO (also used on Data East MPUs) where Q39 was. Orient the reset generator in the same manner Q39 was oriented (pin 1 of reset generator will install in the bottom through hole). In other words, the "flat" side of the reset generator will face to the right.
4. Add a jumper between the right pad of R58 and the right pad of R60 (both resistors have been removed).

This a very simple, and in some cases, cheaper alternative to repairing the existing reset circuit. After this modification is performed, the MPU board will no longer need the +12v to successfully boot. However, +12v is still necessary for the sound section of the MPU on most boards, except the 11B and 11C variants. Likewise, any of the other components in the reset circuit can be removed if desired. Please consult the System 11 MPU board BOM and schematics to determine exactly what components can be removed. The MCP120-460GI/TO reset generator can be purchased from Great Plains Electronics. The MCP130 variant with an internal 4.7K ohm resistor can be used, but it is not necessary. This is because resistor network SR19 pin 3 is an external 4.7K ohm resistor used to pull up the /RESET line.

Below is a prime example of where a reset generator is very handy. The amount of components replaced are reduced from 15-20 to 1. Note that because this was an 11B board (also true for 11C boards), there is no sound amplifier on the board. Hence, the +12v is not needed at all, and none of the parts related to the +12v were repopulated. If this were an 11(nothing) or an 11A board, the +12V inductor (can be replaced by a simple jumper), the W10/W11 jumper, and associated capacitors (C27, C28, and C29) would need to be populated.

• A Prime Candidate for Using a Reset Generator (11B MPU Board Shown)
• 

  Battery alkali leeched into the reset section.

• 

  Components removed, affected traces and pads cleaned and treated.

• 

  Components repopulated. The image above shows the single jumper and the exact minimum parts required when using a reset generator.

Another image showing the single jumper (string tied to it) and the exact minimum parts required when using a reset generator (System 11B/11C). System 11 (nothing) and System 11A must retain a few more parts.

In the above gallery pics, note that some components in the reset area had to be reinstalled. These components are in the area of the reset section, but are not part of the reset circuitry. They are actually related to the game RAM, enabling CMOS power to reach the RAM and to control pin 18, the chip enable pin of the RAM. Without these parts, the MPU will not successfully boot. Instead, there will be a single blink code from the diagnostic LED indicating a U25 RAM error (a "1" will display on boards with the 7-segment display). If the game display is connected to the MPU board, a "U25 RAM ERROR" message will be displayed.

• D1 - 1N5817 diode
• Q40 - 2N3904 transistor (a 2N4401 was used in this case)
• R64 - 1 Kohm 1/4 watt resistor
• R65 - 10 Kohm 1/4 watt resistor
• R68 - 4.7 Kohm 1/4 watt resistor

And, once again, for a System 11 (nothing) or a System 11A MPU, both of which have onboard amplifiers in the upper left corner and require 12VDC power, be sure to retain or replace the following parts:

• L2 - 4.7uH inductor (or a simple jumper)
• C27 - .001uf/50V ceramic capacitor
• C28 - .001uf/50V ceramic capacitor
• C29 - 100uf/25V electrolytic capacitor

5.4.13 Connecting a logic probe to the MPU Board

Logic Probe connected to System 9 MPU Board

Logic Probe connected to System 11 (11A) MPU Board

Of course, +5v and ground can always be acquired at any one of the bypass capacitors, which are located at most every chip, and simply labeled as "B".

5.4.14 Using a PC Power Supply For Bench Testing MPU Board

System 9 MPU connected to a PC power supply

System 11 MPU connected to a PC power supply

The 1J17 power connection is essentially the same for System 9 and all variants of the System 11 MPU boards. The ground connection can be attached to pins 1 through 3. The +5vdc connection can be connected to pins 4 through 6. -12vdc gets connected to pin 8 and +12vdc gets connected to pin 9.

5.5 Game resets

System 9/11 games are far more tolerant of low line conditions vs. the newer WPC games. Some things to check if the game is resetting are the usual culprits for this type of thing: connectors, filter capacitors, slam switches, bad chip sockets, etc. It is good preventative maintenance to replace the +5 volt filter capacitor on the power supply with a new one; most of these are 25+ years old and might be getting to the point of wearing out. Certainly replace them if you are getting resetting on your system 11 game.

Most system 11 games give an indication that they've been slam tilted; if you're getting a "game reset" but you get a noise and/or a message on the displays beforehand, it is probably a slam switch issue vs. a true reset issue. Check all the slam switches in the game (usually the coin door is the main one, and also the ball roll tilt if present).

5.6 Solenoid problems

5.6.1 A/C solenoid/flasher problems

Williams System 11 Auxiliary Power/Driver Board, showing the A/C select relay, the clear "cube" at the top, center of the picture.

The A/C select relay is a dual "make/break" (DPDT, double pole, double throw) relay that supplies power to either the "A-Side" or the "C-Side". It is located on the Auxiliary Power/Driver Board for games that use the board which was introduced with Big Guns and used on all subsequent games. For games prior to Big Guns, like F-14 Tomcat, the relay is probably located under the playfield. There are exceptions. High Speed and Grand Lizard do not use an A/C select relay. Road Kings uses one for a handful of solenoids, but this game did not follow the 1-8 solenoid convention which games after it adopted.

There are a couple ways that the A-Side / C-Side (A/C after this) relay circuit can fail. If the driving transistor (Q8 for Big Guns and later, Q7 for games prior to Big Guns, and located on the MPU board) shorts, the A/C select relay will be constantly energized, resulting in flashers firing when coils should have fired. If the driving transistor never switches the A/C select relay on, coils will fire when flashers should have fired.

Testing this behavior is easy. Use the diagnostic buttons to select solenoid test, and watch carefully to see what fires. The display will show the coil/flasher under test. The first 8 drive circuits tested will have an "A-Side" and a "C-Side". That is, circuits will fire in the following order: 1A, 1C, 2A, 2C, ... 8A, 8C. If the A/C select circuit is not working correctly, the same coil or flasher will fire twice in a row, once for the A side and once for the C side.

The A/C select relay is generally solenoid 12 in coil test. Placing the "Auto Advance" button in the up position, to repeatedly test solenoid 12, should allow you to hear the A/C relay clicking on and off.

If all of "one side" of the A/C select circuit is not working, the fuse supplying the voltage to that side may be blown. F2 (DC side of BR1) and F8 (AC side of BR1) fuse the 25V circuit. F4 (DC side of BR2) and F7 (AC side of BR2) fuse the 50V circuit. Another suspect is fractured solder joints on the A/C select relay. Reflowing these solder joints is simple. The A/C select relay is relatively heavy and it's weight puts stress on the solder connection at the printed circuit board.

Williams System 11 Auxiliary Power/Driver Board A/C Select Relay showing normal, working switch contacts.

It is rare for the A/C relay itself to fail. A possible source of trouble are the miniature contacts on the relay wipers. These are high power, high current, contacts that can be filed with a small point file to restore proper current carrying capacity. Under no circumstances should you ever use contact cleaner on these contacts, as this could cause a spark or fire to occur. Be careful to not mangle the contact wipers in an attempt to adjust them. Adjust with care. 99% of the time, these switch contacts need no maintenance at all.

Williams System 11 Auxiliary Power/Driver Board A/C Select Relay showing "welded" switch contacts.

Pictured at left is a disassembled A/C select relay with both the A Side and the C Side connected at the center contact pads. It's rare for one of these relays to "weld" itself to both sides. In this particular instance, the game would fire both the A-side coil and the C-side flasher at the same time.

Other failure points are the connectors and headers connecting the aux driver board to the main cpu/driver board. Fractured solder joints can affect the proper operation of the relay. It is good practice to replace, or at minimum, reflow the header pins, and to replace connector pins in their housings with new pins. System 11 games use .156 connectors in most locations.

And finally, the solder joints holding the A/C select relay to the board may fracture and cause loss of connectivity.

On a related note, if the flashers never illuminate during game or test, but the A/C relay physically engages, likewise the flasher lamps are good, and power is present at the lamps, suspect potential failure of the U1 (4N25) U1 optocoupler on the interconnect board. Starting with Jokerz, the U1 optocoupler on the D-12313 interconnect board was used to identify whether or not the A/C relay has engaged. Under normal operation, if the A/C relay has engaged, the voltage on the "C" side passes back to the optocoupler. In turn, there is a switch on the switch matrix, typically switch 2, which receives the signal from the U1 optocoupler. Note that U1 is on some previous games like Taxi, but it was not used.

5.6.2 Special solenoid problems with RottenDog MPU board

RottenDog MPU with a red switch bank for selecting either Williams or Data East special solenoid implementation.

These DIP switches on this RottenDog are configured for a Williams game, even though they do not seem to match the PCB silkscreen.

The RottenDog aftermarket MPU board can accommodate both Williams and Data East special solenoid implementations. This is accomplished with an 8 DIP switch bank, pictured at left. If set to the wrong game system, special solenoids will not operate properly in your game. Note: The DIP switch positions should be set the opposite of what you might think, based on the PCB silkscreen.

5.6.3 Special solenoid problems

A typical special solenoid circuit

Special solenoids on system 11 games are similar to the earlier variants used on all previous Williams' board sets. During gameplay/test modes the primary switches on the special solenoids (usually the pops and slings) are active, allowing actuation of those switches to fire the solenoids. This is done by grounding an input to a 7402 logic chip which in turn pre-drives a 2n4401 pre-driver transistor which drives a TIP122 Darlington transistor to fire the solenoid.

During solenoid test mode only, the secondary path to the 7402 is utilized instead to fire the solenoids. Since there are 2 paths to fire the solenoids, it is possible for a special solenoid to work in game/test mode, but fail the solenoid test due to a failed PIA or 7402 chip, and vice-versa.

Note: The System 9 circuit board architecture provides no physical way to control the special solenoids other than via the primary switch associated with the special solenoid device. Therefore, System 9 special solenoids are not activated during solenoid test.

If a special solenoid actuation switch locks on, the solenoid itself will lock on as well, (hopefully) blowing the associated fuse before blowing all or some of the TIP122/2N4401/7402 circuit. An inline fuse holder can be added to all special solenoids with a 1 amp fuse installed to help prevent this situation.

There is a 22uF capacitor and a 100 ohm resistor mounted across the special solenoid switches. The use of the capacitor and resistor creates what is called an RC circuit. The RC circuit is used to filter noise from the switch signal as well as to ensure a minimum pulse length for the solenoid activation. If using a polarized capacitor the positive terminal goes to the tie point with one end of the resistor attached. Should a special solenoid lock on, and the switch leaves are properly gapped, the issue may be a shorted switch capacitor or resistor.

Most system 11B games do not use the special solenoids in the same way. Special solenoid switches were added to the switch matrix instead, keeping all or some of the solenoid pulses under CPU control. It is easy to identify whether a game has CPU controlled special solenoids by inspecting the activation switches on the special solenoids. If there are both a primary and a secondary switch installed, the game operates via the direct-fire special solenoid setup. (Note that a secondary switch in the case of slingshots does not mean the normal 2nd switch all slings shots employ. Rather, it refers to a switch that only activates when the sling activates. It is installed near the pivot point of the sling arm underneath the playfield.) Also the secondary switches go to CPU connector J18. If nothing is connected to J18, then the special solenoids are CPU controlled. In games like BK2K only the kickers are CPU controlled while the pop bumpers still use the secondary switches. There was no rhyme or reason (no definitive start / stop point) which System 11B titles used special solenoid secondary switches, and which did not. Please refer to the game list above to determine which did / did not use the secondary switches.

6.2V Zener Diodes (ZR3-ZR8) that "protect" the board logic from shorts between the sense switches and coil power. Note: in this image, ZR7 and ZR8 have been replaced

The usual cause of special solenoid circuit failure is a stuck switch that locks the associated solenoid on, causing the TIP-122 transistor to fail and the coil to get so hot that the winding insulation melts (causing a reduction of coil resistance, eventually approaching zero ohms). The zener diode (pictured, left) in the circuit may fail shorted as well. This will cause the associated coil to lock on even after repair of the other damaged circuitry since the zener is connected directly to ground. The shorted zener diode causes the 7402/2N4401/TIP122 circuit to react identically to a switch closure. Gack!

Note: If your game doesn't use the special solenoid switch inputs as in most System 11B games and all System 11C games (i.e. no connection at 1J18), then these zeners serve no purpose and may be clipped off.

5.6.4 Unusual Solenoid Problems

A pinched wire (white/brown) that causes a tough to find solenoid problem.

The System 11 method of securing boards sometimes causes odd problems. The securing screws must be removed completely to remove a board vs the improved WPC method of sliding the board off the screw head via a "keyhole".

When securing the boards, it's important to ensure that wires aren't pinched behind the board. If a wire is pinched severely enough, the insulation will be displaced and the conductor will contact the backbox ground plain, which can cause a wide range of odd problems.

The wire in the picture shown at left happened to be the ground path for coil #1, the outhole kicker on Whirlwind. When the game was powered on, the coil would lock on until the blanking circuit became active. Pressing the diagnostic/audit coin door button would also result in the coil locking on. This was actually the A/C relay flipping to the coil side for diagnostics and providing power to the coils. The outhole coil had a path to ground via the pinched wire and would lock on. A game could be started and the coil appeared to work properly until any of the 1 through 8 coils were activated (when the A/C select relay would energize again). At that time, the outhole coil would fire at the same time as the other 1 through 8 coil.

This one was a real head scratcher!

5.7 Flash Lamp Problems

System 11 (High Speed) warming board. Note that flashlamp power is no longer routed through this board. The orange wire connects to two flash lamps wired in series which are then connected directly to flashlamp power. The two black wires are ground. The gray-blue wire is the path to ground via a drive transistor.

System 9/11 flash lamps are driven by the same kind of circuitry that drives solenoids. Flashers are generally wired two in series (like old Christmas lights). If one lamp fails, the neither lamp will light. They are wired in series so that they split the DC voltage across two lamps, preventing them from burning out quickly.

To make the flashlamps "agile" (i.e. turn on rapidly) and to extend the life of the flashlamps, a "warming" circuit was employed which connects two series flashlamps to coil power via a 330 ohm 1 watt resistor. This causes the lamp filament to turn on very slightly, and to "warm up", ready for the power to find ground via a 1 ohm (5 ohm in System 11) resistor.

Over the years, these games have undergone "maintenance" to replace parts on the resistor board and have sometimes also been rewired, attaching the wires to different solder lugs on the flashlamp warming board.

Note: Warming boards are not shown in any of the System 11 documentation.

Note: If LED flash lamps are installed instead of incandescent lamps, the 330 ohm warming resistor must be removed from the circuit. If it's not removed the lamp will remain on, albeit not at full intensity.

5.8 Lamp problems

This may seem painfully obvious, but with any lamp problem, where all lamps in the same circuit are not lighting, first see if the associated fuse is blown.

5.8.1 General Illumination Issues

As a rule of thumb, general illumination (GI) lamp issues with System 9 or 11 games are dependent on the era of the game. The reason for this is GI circuits passed through different boards over the years. And, the majority of GI issues are typically due to the circuit board. Not so much the circuit board itself, but the junction between the header pin connections on the circuit board and the connector pins located in the female housings.

It is extremely common for the header pins and / or the female connections related to the GI circuit to heat up, and in worse cases char or burn. When some of these games were in operation earning money, they could potentially be turned on for 8, 12, or more hours every day. If the connectors were compromised at some time in the game's life, the issues become compounded.

The problem is a vicious cycle. If either side of the connector heats up, it creates more resistance. If it creates more resistance, it draws more current. If it draws more current, it heats up. This cycle continues until the connector blows the circuit's fuse (provided the proper fuse is installed and not by-passed), the connectors are so bad that the circuit no longer has continuity, or the connectors are replaced with new ones. It is recommended that both the male header pin connector be replaced and the associated pin connectors which connect to it. Replacing only one side of the connector is only asking for repeat problems down the road. Header pins rated for 7 amp are highly recommended as a replacement. If the 7A headers are used, the likelihood of the connector failing due to heat is greatly reduced. Likewise, the use of Molex Trifurcon connector pins are recommended.

Another prevalent problem is the failure of the relay board or the relay board connectors. GI relay boards are typically found under the playfield and on the back of the lamp insert panel. The job of the relay board is to switch the GI lamps on/off for special lighting effects. Over many years of operation, the solder joints on the connector may fracture. Although it is more common for the male/female connectors to burn badly. Repinning these connectors, as is typical for GI connectors, is the prescribed fix.

5.8.1.1 Games From Space Shuttle to Fire!

Typical System 9/11 Power Supply with GI Drivetrain Notated

During this time period, general illumination entered the power supply [1 / 2], was parsed, and exited the power supply as 4 separate circuits [4]. Although there were 4 discrete GI circuits with 4 separate fuses [5], all 4 circuits were turned on/off via one relay [3] located on the power supply.

Note that the factory connection on the power supply for System 9 games differs than what is shown in the picture. Connections [1 / 2] do not exist on System 9 games, and only two wires are used to input the general illumination instead of four.

5.8.1.2 Games From Big Guns to Swords of Fury

The interconnect board from a Cyclone mounted to the left side of the backbox

With all the games mentioned in this particular section, there were some design growing pains. Williams was creating a preliminary interconnect board, however, it was very different from game to game at first. Therefore, GI issues may occur at the power supply connections, the interconnect board connections, and / or at one of the discrete GI relay boards used. Since single-sided circuit boards were used for the GI relay boards, cracked solder joints where the relay is soldered to the board is possible in addition to the header pins.

A shown at the pic on the left, the general illumination headers, connectors, and connector housings have been changed out due to fatigue / failure. This is a common failure point on games such as Big Guns, Space Station, and Cyclone.

5.8.1.3 Games From Taxi to Bugs Bunny's Birthday Ball

The games included during this era used a somewhat standardized interconnect board located below the CPU board. The GI circuits coming directly from the transformer enter the board on the right side at connector J6. The top 4 wires are yellow and the lower 4 wires are yellow with a white trace. In actuality, the GI wires are only 4 wires, instead of 8 wires. Once the upper 2 wires reach the board, each of them are daisy-chained through the insulation displacement connector (IDC). Equally, the bottom 2 wires are connected in the same fashion. This was presumedly done to reduce cost, and decrease the life expectancy of the connection as a whole.

Williams Interconnect Board with general illumination wire colors marked. Interconnect board is from a Black Knight 2000

The most typical header connections that fail on the interconnect board are J6 (the GI input and return) and J7 (the GI output and return lines to the backbox insert). The purpose of the picture to the left is to easily identify what GI circuits pass through what connections. This will aid in attempting to check the continuity of your rework, whether only replacing just one of the header connectors or both.

Please note that the wiring color codes for J6 in the pic are not the incoming wire colors. The wiring codes in the pic are illustrated this way to determine where the GI circuits disperse throughout the board. The game's schematics in the manual include this information too.

Keep in mind that discrete GI relay boards were still used during this time period. Dependent on the game, there may be one to three discrete GI relay boards used. Again, since single-sided circuit boards were used for the GI relay boards, cracked solder joints where the relay is soldered to the board is possible in addition to the header pins.

5.8.1.3.1 Repairing the General Illumination Input Connector / Wiring

General Illumination input connector with looped (butterfly) wiring

One method for repairing the GI input connector

From the factory, only 4 wires total were connected to the GI power and return input connector, which is located on the right hand side of the interconnect board. With only 4 discrete connections per wire from the transformer, 8 wires are needed to connect to the interconnect board. The top 4 wires are yellow and the lower 4 wires are yellow with a white trace. To increase 2 wire connections to 4 wire connections, the wires connect to the IDC connector via one connector pin, pass through that connector pin, and loop back to another connector pin. In doing this, the life expectancy of the connector and the header pins on the interconnect board are reduced (too much current for these connections to maintain over their lifespan). The end result is typically burnt connector pins, housings, headers, and in some cases, the interconnect board too.

To resolve the burnt connection issue, the header pins on the interconnect board should be replaced with new ones. Preferably 7A higher current rated header pins. As for the harness side, the housing should be replaced too, and Molex Trifurcon crimp pins used within the housing. To compensate for the extra wires needed, each of the input wires will need a "Y" junction soldered in place. To do so, take a small piece of wire approximately 4" - 5" long. In the middle of the small piece of wire, carefully strip the insulation, leaving about 1/4" of exposed wire. Place a piece of heatshrink on the single incoming wire. Then, bend the small piece of wire at its center where the insulation is stripped, until both ends leads are nearly parallel with one another. Take the incoming wire, and solder it to the exposed wire in the center. The end result will look like a rudimentary "Y". Pull the heatshrink over the junction, and that's it. This needs to be done for the remaining 3 wires.

In addition to the wire modifications, you can also do a board modification. The J6 connector on the right side can be burned to a crisp. Often lifting pads under the printheader. To repair this you can add a jumper wire to the top 4 traces, connecting pin 1-2-3-4 together. This can also be done for the bottom 2 traces, connecting pin 6-8 together, but you are still missing pins 5 and 7 to be connected to 6 and 8. Turn the board over and solder a jumper over pins 5 thru 8. The same can be done for pins 1 thru 4. Now you have the best possible connection.

• 

  Dr Dude Interconnect Board with Modification

• 

  Dr Dude Interconnect Board modification details (front)

• 

  Dr Dude Interconnect Board modification details (back)

5.8.2 Controlled Lamp Issues

5.8.2.1 Lamp Columns or Rows from n to 8 Inoperative

A broken SIP resistor, causing lamp matrix columns 5, 6, 7 and 8 to fail.

The same broken SIP resistor after removal.

5.8.2.2 Lamp Sockets

Failed lamp sockets are not very common with System 9-11c games. If a specific lamp will not illuminate, spinning the lamp socket canister or solder tab can resolve the issue in some cases. One thing to keep in mind is that all controlled lamp sockets have a diode soldered to two of the lamp socket tabs. Due to this, the legs of the diode can get pressed against other leads on the lamp socket resulting in an electrical short. The end result may be a lamp which shines more brightly than other lamps in the same column or row, or a whole lamp column or row will illuminate dimly when they shouldn't be lit at all. The resolution is to inspect all associated lamps in the column or row, and look for solder tabs or diode legs touching where they should not be.

5.8.2.3 Lamp Boards

Starting with Pinbot, Williams introduced lamp boards. Lamp boards were used in areas where lamp inserts were positioned too closely together, and it did not make logical sense (or it was not physically possible) to use individual lamp sockets. With subsequent games, lamp boards were used whenever lamp inserts were "nested" in groups. This trend was presumedly done to save on assembly and production costs. One benefit of lamp boards is that bulb replacemnt became easier in most cases. A simple turn of the twist lamp socket (WMS # 24-8767), and the 555 bulb could be removed.

Early direct solder lamp board as seen on Pinbot - the large color matrix below the top and right colored targets

The first generations of lamp boards had the lamp column and row wires soldered directly to the board. On one hand, this is beneficial, because there is one less mechanical connection used in the lamp circuits. On the other hand, it makes it more difficult to remove the lamp boards completely from the game to clean and / or service. On the plus side, very little goes wrong with the direct wired lamp boards.

"Divots" in the connection pad of a "twist in" lamp.

Two things that can fail are individual lamp circuit diodes or the solder pads can develop "divots", where the twist sockets make physical contact with the board. To overcome the divots, simply add a little solder, and carefully reflow the solder pads. Make certain not to apply too much heat...just enough to evenly flow the solder.

The solder on these header pins has been reflowed nicely.

With Fire!, Williams started using lamp boards with .156" header connections for the lamp column and row lines. This allowed for easy removal and service of the lamp boards. However, it introduced some new problems. Because the lamp boards used were single-sided, (this was presumably a cost analysis decision), the .156" header pins can develop cold solder joints and crack due to vibration. The problem can easily be rectified by removing the old solder from the header pins, and applying fresh, new solder. This style of lamp board is susceptible to the same issues (bad diodes and divots) as mentioned in the direct wired lamp boards section above.

5.8.2.4 Adjacent Columns or Rows Lighting Simultaneously

Column drive transistor tabs touching, shorted together. Here, lamp columns 3 and 4 would light at the same time.

A problem seen fairly often is the tabs of either column drives or row returns touching, and therefore shorted and causing more than one column (in the picture at left) or row to light at the same time. If you have this issue, it's easy to quickly examine the column drive and row return transistors.

5.8.2.5 Lamp Matrix Row and Column Testing

The CPU logic for the lamp matrix can be tested by connecting a spare lamp using a jumper wire. The following sections show the separate procedures for testing the lamp matrix columns and rows. The example is on a Sys11A PinBot, but applies to all Sys11 CPU boards.

Testing the lamp matrix columns:

Jumper connection for testing lamp matrix columns.

Use the following procedure to test the TIP42 transistors that drive the lamp matrix columns. Note that a diode is not needed for these tests since it's function is to prevent interaction between the lamps in the matrix. In this test we are only connecting a single lamp at a time.

1. Remove the backglass and open the insert to get access to CPU board connectors 1J6 (row) and 1J7 (column).
2. Unplug connectors 1J6 and 1J7 (lower right corner of CPU board)
3. Turn the game on and go to the "All Lamps" test in the Test/Diagnostic Menu. This is done by opening the coin door an pressing: MANUAL-DOWN, ADVANCE, AUTO-UP, ADVANCE x 2
4. Clip one end of the test jumper to 1J6 pin 1, the rightmost pin on the connector
5. Touch the other end of the jumper to 1J7 pin 1, the rightmost pin on the connector
6. The test lamp should flash to indicate that the column driver is working.
7. Repeat the test for the pins 2 through 9 on 1J7. There is no pin 5 as it is the key.

If a column doesn't light or is stuck on, reference the lamp matrix table in the manual to identify the transistor to test.

The following table shows the lamp number and driving transistor for each of the column pins.

Pin	Wire Colors	Lamp number	Transistor number
1J7-1	Yel-Brn	1	Q66
1J7-2	Yel-Red	9	Q64
1J7-3	Yel-Orn	17	Q62
1J7-4	Yel-Blk	25	Q60
1J7-5	Key Pin
1J7-6	Yel-Brn	33	Q58
1J7-7	Yel-Blu	41	Q56
1J7-8	Yel-Vio	49	Q54
1J7-9	Yel-Gry	57	Q52

Testing the lamp matrix rows:

Jumper connection for testing lamp matrix rows.

Use the following procedure to test the TIP102/122 transistors that drive the lamp matrix rows.

1. Remove the backglass and open the insert to get access to CPU board connectors 1J6 (row) and 1J7 (column).
2. Unplug connectors 1J6 and 1J7 (lower right corner of CPU board)
3. Turn the game on and go to the "All Lamps" test in the Test/Diagnostic Menu. This is done by opening the coin door an pressing: MANUAL-DOWN, ADVANCE, AUTO-UP, ADVANCE x 2
4. Clip one end of the test jumper to 1J7 pin 1, the rightmost pin on the connector
5. Touch the other end of the jumper to 1J6 pin 1, the rightmost pin on the connector
6. The test lamp should flash to indicate that the column driver is working.
7. Repeat the test for the pins 2 through 9 on 1J6. There is no pin 4 as it is the key.

If a row doesn't light or is stuck on, reference the lamp matrix table in the manual to identify the transistor to test. The following table shows the lamp number and driving transistor for each of the row pins.

Pin	Wire Colors	Lamp number	Transistor number
1J6-1	Red-Brn	1	Q80
1J6-2	Red-Blk	2	Q81
1J6-3	Red-Orn	3	Q82
1J6-4	Key Pin
1J6-5	Red-Yel	4	Q83
1J6-6	Red-Grn	5	Q84
1J6-7	Red-Blu	6	Q85
1J6-8	Red-Vio	7	Q86
1J6-9	Red-Gry	8	Q87

5.8.2.6 Using MOSFETs in the Lamp Matrix Circuitry

Procedure by Marco Albus
The original Williams (and Data East) lamp matrix circuitry dissipates a lot of heat through the 27 ohm current limiting resistors. Often, this part of the circuit board is badly heat damaged. Using modern MOSFETs (IRF9530, IRF9Z34N or FQP17P06), the heat (and power consumption) can be reduced substantially.

Once MOSFETs are substituted for the TIPs in the lamp driver circuit, it is possible to replace the 27 ohm resistors with simple jumpers or even leave the 27 ohm resistors in place if they are in good shape as the low resistance will have a negligible effect on the operation of the MOSFET. However, the problem with either of these is that the MOSFETs are driven by 18 VDC Source to Gate voltage which is dangerously close to the absolute maximum rating of the IRF9530 and IRF9Z34 of 20 VDC (this also holds true for Data East MPUs). The FQP17P06 has a higher Vgs rating (25VDC) than either the IRF9530 or IRF9Z34N - but you are still driving the part at over 70% of its absolute maximum. Typical uses for these MOSFETs drive them by about 10 VDC Source to Gate voltage (or -10V for P-Channel MOSFET's used here).

Typical "Zero Ohm Jumper" implemenation (left), "Voltage Divider" implementation on right. Image courtesy of Marco Albus.

The schematics shown at left compare the widely used "Zero Ohm Jumper" implemenation with the "Voltage Divider" implementation that uses two 1K, 1/4 watt resistors to divide the source 18 VDC in half. Doing so drives the MOSFET with a 9 VDC Vgs.

5.8.2.6.1 MOSFETs Installed in a System 9 MPU Board

1K resistors and IRF9Z34N FETs installed on the component side of the board.

1K resistors installed on the solder side of the board.

Parts Required

• 8 P-Channel MOSFETs (IRF9530, IRF9Z34N or FQP17P06)
• 16 1K Ohm 1/4 Watt resistors

Procedure

• Remove the TIP-42 transistors Q23-30
• Remove the 27 Ohm resistors R101, 103, 105, 107, 109, 111, 113, 115
• Remove the 2.2K ohm resistor network SR5
• In place of the 27 Ohm resistors, install 1K Ohm resistors
• In place of the TIP 42 transistors, install the P-Channel MOSFETs, oriented the same as the TIP-42s were oriented
• On the solder side of the board, install eight 1K Ohm resistors between the MOSFET gate and the 18VDC source.

5.8.2.6.2 MOSFETs Installed in a System 11 MPU Board

1K resistors installed on the component side of the board.

1K resistors installed on the solder side of the board.

Parts Required

• 8 P-Channel MOSFETs (IRF9530, IRF9Z34N or FQP17P06)
• 16 1K Ohm 1/4 Watt resistors

Procedure

• Remove the TIP-42 transistors Q52, 54, 56, 58, 60, 62, 64, 66
• Remove the 27 Ohm resistors R82-89
• Remove the 2.2K ohm resistor network SR16
• In place of the 27 Ohm resistors, install 1K Ohm resistors
• In place of the TIP 42 transistors, install the P-Channel MOSFETs, oriented the same as the TIP-42s were oriented
• On the solder side of the board, install eight 1K Ohm resistors between the MOSFET gate and the 18VDC source.

A very clean alternate implementation with resistors on the component side. Image courtesy of PinSider DumbAss.

Shown at left is a very clean installation of all parts on the component side of the board. Note that the 1K resistors are situated such that the "body" of the resistor passes between the MOSFETs instead of the legs, thereby eliminating the possibility of a short.

5.9 Switch problems

5.9.1 Diode and Switch Matrix Wiring Orientation on a Microswitch

Typical wiring for any Microswitch

First and foremost, all switches located in the switch matrix MUST have a diode. Secondly, the diode has to be installed in the correct orientation. If a diode is installed "backwards", the end result my be that all of the switches in that particular row get triggered or even stranger issues occur. Finally, the diode legs must not be connected to any of the other wires or terminals of the switch, or the switch will funk up the switch matrix.

Here is the correct orientation for the wiring and the diode on a microswitch. The white wire w/ trace and non-banded side of diode is connected to NC terminal. The banded side of diode is connected to common terminal. Finally, the green wire w/ trace is connected to NO terminal.

5.9.2 Switch Columns or Rows from n to 8 Inoperative

A broken SIP resistor, causing switch matrix columns 5, 6, 7 and 8 to fail.

5.9.3 Switch Matrix Row and Column Testing

The CPU logic for the switch matrix can be tested by simulating switch closures using a jumper wire on both System 9 and System 11 (all revision levels) MPUs.

Testing the System 9 switch matrix, but jumpering from pins on 1J8 to pins on 1J10.

As seen at left, testing the switch matrix rows and columns on a System 9 MPU is identical to testing on a System 11 MPU except for which connectors to use. The switch columns on the System 9 MPU at 1J8 are numbered right to left. The switch rows at 1J10 are numbered left to right. 1J9 is available (sometimes not stuffed) for opto switch processing, although this feature may never been used.

The System 9 switch matrix operates on 5VDC. The switch columns at 1J8 are "strobed" by being pulled to ground successively by the 6821 PIA. The switch columns are normally pulled high via a pull-up resistor. If the switch at the intersection of the column and row is closed, when the column is strobed, the row will be pulled low too. In this way, the processor knows which column it strobed, and which rows return that strobe.

The following sections show the separate procedures for testing the system 11 switch matrix columns and rows. The example is on a Sys11A PinBot, but applies to all Sys11 CPU boards.

Testing the switch matrix columns:

Jumper connection for testing switch matrix columns. This picture shows a diode which is not necessary for this test.

1. Remove the backglass and open the insert to get access to CPU board connectors 1J8 (column) and 1J10 (row).
2. Turn the game on and go to the "Switch Edges" test in the Test/Diagnostic Menu. This is done by opening the coin door an pressing: MANUAL-DOWN, ADVANCE, AUTO-UP, ADVANCE x 6
3. Unplug connectors 1J8 and 1J10
4. Clip one end of the test jumper to 1J10 pin 9, the leftmost pin on the connector
5. Touch the other end of the jumper to 1J8 pin 1, the rightmost pin on the connector
6. The display should report that switch 1 was actuated. The test may report the switch name, refer to the switch matrix table in the manual to correlate the name to the switch number.
7. Move the jumper to 1J8 pin 2 and check the reported switch by comparing to row 1 in the switch matrix table
8. Continue to test the rest of the pins on 1J8. There is no pin 6 as it is the key.

The following table shows the switch number that should be reported for each of the column pins.

Pin	Wire Colors	Switch number
1J8-1	Grn-Brn	1
1J8-2	Grn-Red	9
1J8-3	Grn-Orn	17
1J8-4	Grn-Yel	25
1J8-5	Grn-Blk	33
1J8-6	Key Pin
1J8-7	Grn-Blu	41
1J8-8	Grn-Vio	49
1J8-9	Grn-Gry	57

Testing the switch matrix rows:

Jumper connection for testing switch matrix rows. This picture shows a diode which is not necessary for this test.

1. Remove the backglass and open the insert to get access to CPU board connectors 1J8 (column) and 1J10 (row).
2. Turn the game on and go to the "Switch Edges" test in the Test/Diagnostic Menu. This is done by opening the coin door an pressing: MANUAL-DOWN, ADVANCE, AUTO-UP, ADVANCE x 6
3. Unplug connectors 1J8 and 1J10
4. Clip one end of the test jumper to 1J8 pin 1, the rightmost pin on the connector
5. Touch the other end of the jumper to 1J10 pin 1, the rightmost pin on the connector
6. The display should report that switch 1 was actuated
7. Move the jumper to 1J10 pin 2 and check the reported switch by comparing to column 1 in the switch matrix table
8. Continue to test the rest of the pins on 1J10. There is no pin 4 as it is the key.

The following table shows the switch number that should be reported for each of the row pins.

Pin	Wire Colors	Switch number
1J10-9	Wht-Brn	1
1J10-8	Wht-Red	2
1J10-7	Wht-Orn	3
1J10-6	Wht-Yel	4
1J10-5	Wht-Blk	5
1J10-4	Key Pin
1J10-3	Wht-Blu	6
1J10-2	Wht-Vio	7
1J10-1	Wht-Gry	8

5.9.4 Shorted or non-working switch column

One or more failed switch columns can be caused by a failed column drive transistor (Q42 through Q49, all 2N4401s) or a failed 1K SIP resistor pack at SRC6.

The column drive transistor part was changed with the system 11B MPU. System 9 to 11A MPUs use 2N3904 transistors at location Q42 to Q49 with 1.5Kohm inline resistors (R71 through R78). These column drive transistors fail quite often as these MPUs age. These 2N3904 transistors can be replaced with 2N4401s.

When System 11B was introduced, the switch column drives were changed to 2N5550s with 470 ohm inline resistors. The more commonly found, and beefier 2N5551, is a good substitute for the 2N5550. Before replacing the transistor, be sure which kind the board uses. Some early system 11B MPUs were stuffed with 2N3904s despite the schematics indicating otherwise.

A broken SIP can also cause a switch column not to work. This usually happens with System 11 and 11A CPUs. With 11B WMS also changed the switch section layout and eliminated most of the SIPs. The likely candidates are SRC6 and SR11. In the case of SRC6, measure resistance between pin 1 and 2 to 9. If it's broken it's usually open after pin 3 or 4. This is a bussed 1K SIP with additional 470pfd capacitors connected to ground and practically NLA. Fortunately the caps are not needed because they are also mounted on the board so it can be replaced with a standard 9-8 1K SIP. Pin 10 which is the common ground for the caps is not used here. SR11 is a standard bussed 560 ohm 10-9 SIP. Here you measure resistance between pin 1 and pin 2 to 10. Be aware that two types of SIPs are used on the board: bussed and isolated. If you suspect a different SIP has failed check the schematics for the pin layout.

A broken resistor SIP.

Pictured at left is a broken resistor package similar to SRC6 but out of a System 11 master display. As can be seen, the part may look solid to the naked eye, but it's actually fractured.

5.9.5 Ground Row Shorts

Note the fracture in the SIP at SR11. The fracture became apparent only after desoldering the part.

A possible cause for one or more ground row shorts (indicated by the game software on the score display during switch test) is a fractured resistor SIP at SR11 (a 560 ohm 10-pin bussed resistor network). SR11 is a pull-up resistor that is connected between each switch row return and 5VDC. If the return is not pulled up, that return will look like a low signal continuously, which the MPU interprets as a row ground short. Visual inspection of the SIP might or might not identify the cause of the problem. A better test is to remove the switch row return connector (J10) and probe each pin with a logic probe. The logic probe should indicate "high" for every pin.

Specific to the board pictured at left, a logic probe indicated that J10 pins 1, 2, and 3 were all high but the remaining pins had no signal at all. The fracture can easily be seen between pins 5 and 6 of the SIP (pin 4 isn't used) meaning that pins 6 through 10 were not connected to 5VDC and therefore not pulling the switch row return to logic high. Hence, the MPU declared a row ground short for rows 1 through 5. Note that the actual fracture can rarely be seen prior to attempted removal of the resistor network from the MPU board.

5.9.6 4N25 Opto Couplers in the Switch Matrix

Three 4N25 Opto Couplers on a Williams System 11 Interconnect Board

System 11 games that include an Interconnect Board may use 4N25 Opto Couplers to detect power being applied to the flippers or to detect 28VDC being applied to the A/C relay (as in Rollergames). Pre-Fliptronic WPC games that use the -1 power/driver board also use 4N25 Opto Couplers.

These opto couplers are connected to the game switch matrix. Detecting these switches is most often used for lane change and for entering HSTD initials. If one of these functions no longer works, the 4N25 opto couplers should be considered suspect.

Fortunately, testing an opto coupler is simple using the diode test function provided by most DMMs. Set the DMM to diode test, compare results against those pictured below.

• Testing a 4N25 Opto Coupler
• 

  The internal schematic of the 4N25. Pins 1 and 2 connect to an internal diode (LED). Pins 4 thru 6 connect to a PNP photo transistor.
• 

  Testing the diode connected to pins 1 and 2
• 

  Testing part of the photo transistor. Black on pin 5, red on pin 6
• 

  Testing part of the photo transistor. Black on pin 4, red on pin 6

5.9.7 Opto Switches Not Registering with a Rottendog MPU / Driver Board Installed

Older Style Rottendog MPU Board

Modified Older Style Rottendog MPU Board

If the opto switches are not registering at all in a game, and a Rottendog MPU board is being used, suspect that it is an older style board. This board can easily be upgraded though. U2 will have to be switched from a 74HCT240 to a 74HC240 chip, and RA5 will have to be changed from a 9-pin 4.7K ohm bussed SIP resistor network to a 9-pin 560 ohm bussed SIP resistor network.

5.10 Display problems

System 9/11 High Voltage Section Repair

WARNING: This circuit uses high voltages. Don't continue, unless you are confident in your diagnostic abilities.

5.10.1 High Voltage Issues With the Power Supply

5.10.1.1 Check Voltages

If all displays are blank, your high voltage (HV) section may not be working. On the Power Supply Board, use a DMM set to DC volts with the - lead grounded, probe the following connector pins to determine if the HV section needs repair. If the display fuse, F1 is blowing, you should remove the applicable display connector (with power off) before testing the voltages. Note that these are "loaded" voltages. If you are testing on the bench, or without displays connected, your measurements may vary.

If you have Power Supply D-11883 or D-12246:

3J2 pin 1 = -100 volts DC
3J2 pin 3 = +100 volts DC

If you have Power Supply D-8345-xxx:

3J5 pin 3 = -100 volts DC
3J5 pin 4 = +100 volts DC

If the test points are more than about 5 volts out of spec, then your HV section may be malfunctioning (if you or a previous owner replaced Z2/Z4 with 1N4763A diodes to purposely reduce the display voltage, test readings in the 90's range would be normal).

Check the table below for a solution.

5.10.1.2 Troubleshooting table

Symptom	Possible Cause	Replacement
0V ON BOTH +100/-100 lines	Check F1	1/4 Amp SB
0V on +100V line	Open Diode D3	1N4004
	Open Q2	2N5401
	Shorted Zeners ZR1, ZR2	1N4730A and 1N4763A
	Open Q1	MJE340
	Open R1	39k ohm, at least 1 watt
0V on -100V line	Open Diode D4	1N4004
	Open Q4	2N5551
	Shorted Zeners ZR3, ZR4	1N4730A and 1N4763A
	Open Q3	MJE350
	Open R4	39k ohm, at least 1 watt
F1 1/4 Amp Fuse Blows	Bad Capacitor at C1 and C3	100uF,150V
	Shorted Display	Display Glass*
	Shorted UDN7180	UDN7180*
	Shorted UDN6118	UDN6118*
+118V on +100V line	Shorted Q1	MJE340
-118V on -100V line	Shorted Q2	MJE350
* located on display board

Displays Test With no Shorts
Once the displays are tested and shorts are eliminated, we can proceed with the HV section repair.

Replacing the components in the HV Section
Since there are not that many components, if you are having problems isolating the fault, a quick solution is to replace all the components in the HV section. If only one side (+ or -) is failing, it is possible to rebuild only the failing side. Check the parts section of the wiki to find suppliers, and get the replacement parts for one or both failing sides:

Part	+Side Part	-Side Part	+,-Location
Transistor	MJE340 (or MJE15030)	MJE350 (or MJE15031)	Q1,Q3
Transistor	2N5401	2N5551	Q2,Q4
Zener diode	1N4730A	1N4730A	Z1,Z3
Zener diode	1N4763A**	1N4763A**	Z2,Z4
Resistor	39k Ohm,1W	39k Ohm,1W	R1,R4
Resistor	680 Ohm,1/4W	680 Ohm,1/4W	R2,R5
Resistor	330k Ohm,1/2W	330k Ohm,1/2W	R3,R6
Capacitor	0.1uF,250V metal polyester	0.1uF,250V metal polyester	C2,C4
Capacitor	100uF,150V	100uF,150V	C1,C3
**This is a 91V Zener to reduce the voltage to prolong display life

A visual depiction of the parts in a System 11 high voltage section. Note that the legs of the MJE15030/MJE51031 are crossed as this power supply originally used OEM parts SDS-201 and SDS-202

Remove and replace HV components

• Clip the old components from the board (make sure you have new ones first).
• Use one of the desoldering methods to remove solder from the holes.
• Stuff board with new components.
• Check for correct orientation on transistors, diodes and the large capacitor if you replace it.
• Leave a little space under components for air flow.
• Bend leads on components so they won't fall out when board is inverted for soldering.
• Double check that all the correct parts are in the correct places and properly oriented.
• Solder the parts to the board
• Clip excess off leads

5.10.1.3 Note on installation of MJE340/MJE350 replacement transistors

The MJE340/MJE350 is the heat-sinked transistor. On most Williams boards, this transistor pin configuration is not the same as the original part. It will need to be installed vertically with heat sink attached to the transistor only. The transistor will sit at about a 45 degree angle so the legs can be lined up to fit in the correct holes.

Check your board and insure correct orientation before soldering in place. Late versions of the System 11 series boards were designed to use the pin configuration of the MJE340/MJE350 transistors.

Do not mount vertically if the power supply is designed to use the pin configurations of the newer transistors!

5.10.1.4 Note on installation of MJE15030/MJE15031 replacement transistors

High Voltage section with MJE15031 installed. Note the crossed legs on the left transistor.

Power supply with pads for both styles of HV transistors

The MJE15030/MJE15031 are rated for higher power and can be used instead of the MJE340/MJE350 (the MJE340/MJE350 are well within design specifications and are suitable replacements at 1/3 the cost of the MJE15030/MJE15031).

The MJE15030/MJE15031 pin configuration has the Base and Emitter reversed as compared to the MJE340/MJE350. In the image at left, the left transistor (MJE15031) shows one method of crossing the legs safely. The transistor's two left legs are shifted right on position. The third leg is then jumpered to the boards left through hole position. If you prefer not to cross legs, this transistor can also be mounted vertically, but this transistor would be facing the opposite direction as compared to the MJE340/MJE350 pictured above.

There are some power supplies which have pads for both the SDS-201/SDS-202 or the MJE15030/MJE15031 style of transistors. In the pic to the right, the MJE15030/MJE15031 transistors are installed from the factory, but there are solder pads just below for installation of the SDS-201/SDS-202 transistors. If installing ME15030/MJE15031, the upper pads are used, and no leg crossing is required.

Ready to test

To test the rebuilt power supply, return to the "Check Voltages" section of this guide.

Other Resources
The System 9-11 HV section of the PSU is very similar in design to the earlier System 3-7. So it will be worth reading through the Sys 3-7 PSU Problems section entitled +/-100v Display HV Section of PSU, for more detail. It also provides links to source complete HV rebuild kits which will normally cost under $10 shipped.

5.10.2 Display Panel Problems

5.10.2.1 Segment Identification

Alphanumeric and Numeric Segment Identification.

5.10.2.2 Display Will Not Light

Testing a high voltage plasma display with an inexpensive plasma ball

If power to the display is correct, and the display was moved to a known working display panel position (if possible), it is possible that the display has outgassed. A plasma ball can be used to determine if a display glass is outgassed.
A Youtube video demonstrating the use of an inexpensive plasma ball for this purpose can be found here.

5.10.2.3 Segments Light "In Sympathy" With Other Segments

In this image, segment "b" and segment "k" light simultaneously, even though only one segment should be lit. The last two characters illustrate the issue clearly. Image courtesy of PinSider Pinball50.

Internal shorts within a display glass can cause segments to light "in sympathy" with other segments. That is, segment "b" and segment "k", for example, will both light, when only one of the segments should be lit. Segment "b" will light whenever segment "k" is lit and segment "k" will light whenever segment "b" is lit. This is caused by an internal short within the glass. The short can be confirmed by removing the leads of the associated segments from the PCB and measuring resistance between them. The resistance should be infinite, or open.

In this case, there is no way to repair the glass and it should be replaced.

5.10.2.4 Commas stuck on (Players 1&2 or 3&4)

System 9: Check IC36 on the MPU (6821 PIA)

Pin 16 = Players 1 & 2 Pin 17 = Players 3 & 4

With the game in attract mode, check these pins with a logic probe. They should go low each time the displays turn off. if they stay high, the PIA is likely bad and will need to be replaced.

5.10.2.5 Display Fuse F1 Blows

Testing the UDN Chips
If the display fuse is blowing, you should check the display board for shorts before connecting a new or rebuilt power supply to the display board, as a shorted display or chip can damage a good power supply. Remove the display board from the system. The display characters are driven by the two types of UDN chips, the UDN7180 & UDN6118 (or UDN6184). Locate these chips (there are several) on the board and test them with your DMM set to diode check. Clip your + (RED) lead to the ground trace of the board. Probe the UDN chips as shown in the diagram. If any shorts are read in the tested pins, the display should not be connected to the power supply until the shorts are corrected.

Shorts Are Found in UDN Chip Test
If a short reading is found in the tested pins (don't test the pins labeled 'dont care'), the UDN chip should carefully be desoldered and removed from the board. Take care to preserve this chip, as they are nearly impossible to find and expensive to replace. Now install an IC socket in its place. Repeat the test with no chip installed. If the short is gone, then the UDN chip needs to be replaced. If the short remains, then the display glass needs to be replaced.

5.10.2.6 Replacing Display Glass

Replacement glass installed with glass nipple in proper location

Replacement glass installed after circuit board was drilled to allow for glass nipple in different location

All plasma displays have a finite life expectancy. Display glass replacements were and are offered in several configurations. The differences in these configurations are mainly the placement or lack of a display "nipple" on the back of the glass. The glass nipple is what can pose as somewhat of a problem, when replacing a display glass. Williams did place a rather large hole in the display circuit board. However, the placement of the glass nipple has changed. To overcome this issue, a hole can carefully be drilled through the circuit board to allow for the glass nipple. When drilling, drill successive holes, starting with a small drill bit first, rather than drilling one large hole. This will reduce the chances of the circuit board "splintering".

At one time, pinballpcb.com did offer replacement displays and / or just the replacement display circuit boards. Neither of these products are available anymore.

Replacement glass installed

Original display glass was secured to its circuit board with double-sided, high tack, foam tape. If this cannot be acquired, a simple method to secure an existing or new display glass is to use silicone caulk or RTV caulk. However, in following this method, the glass must be supported on the board until the caulk cures. Wrapping the glass with rubber bands is a great way to do this. Once the caulk is cured, just cut off the rubber bands.

5.10.3 Logic States of Display Output Signals for a Working MPU Board

These logic states were recorded from a System 11A board. This board was running Swords of Fury code in display test mode manually paused with all characters / digits on "8". This is for games with a 4 x 7 display board.

L=LOW, H=HIGH, P=PULSE

IJ3 (pin)	Logic State
1	L
2	P
3	P
4	P
5	P
6	None (Key)
7	P
8	P
9	P
10	L
11	L
12	H

IJ2 (pin)	Logic State
1	P
2	P
3	P
4	P
5	P
6	P
7	None (Key)
8	P
9	P

IJ1 (pin)	Logic State
1	P
2	P
3	P
4	P
5	P
6	P
7	P
8	None (Key)
9	P

IJ22 (pin)	Logic State	IJ22 (pin)	Logic State
2	L	1	L
4	L	3	L
6	L	5	L
8	L	7	L
10	L	9	L
12	L	11	L
14	L	13	P
16	L	15	L
18	L	17	L
20	P	19	P
22	P	21	P
24	P	23	P
26	P	25	L

5.11 Sound problems

5.11.1 System 9 Sound Problems

5.11.1.1 Jumper Settings for a newer System 9 Speech Board

Jumpers Circled on a Comet System 9 Speech Board

As mentioned above, the System 9 speech board is essentially the same as the System 6/7 speech board. However, the newer speech board has jumpers added to increase the flexibility of EPROMs used. Either 2532 or 2732 EPROMs can be used. If using 2532 EPROMs, install jumpers W2 and W4, and remove jumpers W1 and W3. If using 2732 EPROMs, install jumpers W1 and W3, and remove jumpers W2 and W4.

5.11.1.2 Missing Sound or Speech Calls

To start a sound test, simply press the SW2 momentary switch at the bottom of the CPU board located between connections 1J16 and 1J17. Once SW2 is pressed, the CPU will cycle through all of the sound / speech calls continuously. Each System 9 game has a chart located in the manual, which lists the speech and sound calls made during test. The benefit of the chart in the manual is that it lists the ROM where a particular call originates. If a particular sound / speech call is not heard, the associated ROM listed for that call may be bad.

Another result of speech calls missing is possibly a bad U2 or U3 on the speech board. Both of these chips are 1458 op-amps.

5.11.1.3 Isolating the CPU Sound from the Speech Board

Some of the sound / speech calls originate from the CPU board, while the remaining calls originate from the speech board. If after running a game through sound test by pressing SW2, and some sounds are missing or very low in volume, a good idea is to isolate the CPU board from the speech board. All of the sounds originated on the CPU board, leave the board via the 40 pin ribbon cable, and go to the speech board for mixing. Once mixed, all of the sounds and speech are then returned to the CPU board, amplified, and sent out to the speakers.

Much like the System 6/7 sound and speech boards, the sound can be isolated from the speech board for testing purposes. The System 6/7 Type 2 sound board uses a jumper at position W1. When jumper W1 is installed, the sound board can be diagnosed without the speech board installed. The equivalent with the System 9 CPU board is jumper W10. This jumper is typically not connected. However, when it is connected, the sounds originating from the CPU board can be heard.

Locate jumper W10 on the CPU board. W10 is located to the right of SW2 on the board. There should be two very small wires clipped on either side of W10, as if W10 was installed and removed. Connect an alligator test lead across the short leads of W10. If the short leads are not present, a short light gauge wire or wire wrap can be tack soldered to the solder pads of W10, or use a jumper wire to connect the lower legs of C32 and R76 which are just right of W10 (see pic, below). With a jumper installed at W10, some sounds should be heard when SW2 is pressed. Sound test, via the coin door diagnostic button is likely to provide sounds for all but sound 7, which is a voice call. If no sounds are heard, further troubleshooting must be performed.

• Adding W10 to a System 9 CPU Board
• 

  Locating jumper W10
• 

  Temporarily jumpering W10
• 

  Alternate for jumper W10

5.11.2 System 11 Sound Problems

System 11 games are known to "hum" a bit. The "hum" is caused by electromagnetic interference from the lamp matrix insinuating itself on the sound circuit. To minimize the interference, ensure all boards are screwed tightly to the backbox ground plain, with all screws installed. This improves grounding and will help minimize "hum".

Another cause of hum could be an inconsistent +5DC volts from the power supply. An indicator of this being the cause of hum would be the game occasionally resetting as well. Replacing caps C8 and C10 on the power supply may correct this issue.

Depending on the version of System 11 board, an 11C board or anything else (nothing, 11A, and 11B), sounds and music are handled with different specifics. With the first 3 generations of MPU boards, the sounds originate from the MPU board and the background music originates from the sound board. With the 11C board, both sounds and music are on the sound board, as the MPU no longer has a sound section populated on it.

5.11.2.1 Volume POT value

Some System 11 games shipped with a 5K linear volume POT. Others shipped with a 10K linear volume POT. Both POT values work fine and either can be used to replace a failed volume POT.

Some System 11 games shipped with a 50 ohm resistor soldered to the volume POT. It is thought that this resistor was included so that the volume could never be turned to zero. For route games, at least some sound was thought to increase game earning potential. This resistor can be removed or jumpered around to enable turning the sound volume completely off.

5.11.2.2 Boot Up Diagnostic Tones for the D-11297, D-11298, and D-11581 Sound Boards

The D-11297, D-11298, and the D-11581 sound boards all provide diagnostic tones ("bongs") at power on. A single "bong" means that the sound board is operating properly.

Lack of a bong indicates a problem with the sound board or it's connections. The possible causes for failure are numerous, including a failed amplifier, missing or incorrect connector placement, missing or noisy (AC ripple) power for a required voltage (-12VDC), failed speakers or the connection to the speakers, etc.

If the sound board is healthy enough to perform it's own self-diagnosis, it will attempt to indicate the problem by playing a specific number of bongs, as shown in the tables below.

D-11297 & D-11298 Sound Boards

"Bong" Count	Description
1	Normal sound board boot with no problems detected.
2	U3 RAM Problem
3	U4 Sound ROM Problem
4	U19 Sound ROM Problem

D-11581 Sound Board

"Bong" Count	Description
1	Normal sound board boot with no problems detected.
2	U5 RAM Problem
3	U4 Sound ROM Problem
4	U19 Sound ROM Problem
5	U20 Sound ROM Problem

5.11.2.3 Sound Diagnostics Via the MPU Board for Games Prior to 11C

Pressing the upper switch of the two switches on the side of the MPU board will test both the CVSD and DAC on the MPU board. The first sound is typically sound ROM (on the MPU) specific, and it is a test of the CVSD chip. The second sound is described as a "twing", and it is a test for the DAC. If no sounds are heard, refer to the potential issues listed above in the previous section. If one "ring" is heard, it denotes that the U23 RAM chip has failed. Two or four rings denotes that there is a problem associated with U21 ROM. Three or five rings denotes that there is a problem associated with U22 ROM.

5.11.2.4 Connecting the Speaker and Volume Connections for a System 11 MPU Board on the Bench

Speaker and Non-adjustable Volume Control Connected on a System 11 MPU Board

To test the sound on a System 11 MPU on the bench, a speaker and volume control must be connected to connectors 1J15 and 1J16 respectively.

The speaker should connect to pins 2 and 3 of 1J15.

In the adjacent pic, a 470Kohm resistor was used in lieu of a volume pot. Other resistor values can be used, but 470Kohm allows the sound to be set to a comfortable audible level. Alligator clips can be used, but if the intentions are to test sound boards in the future, a resistor can be crimped into a .156" housing as shown. This resistor/connector set up can be used on all Williams sound boards from System 3 through System 11.

If using alligator clip leads to power the board, use special care as to not send +12v down the -12v line (the pins are adjacent to one another). Sending +12v down the -12v bus will most definitely destroy the 1408 DAC chip (if populated).

5.11.2.5 Sound Is Very Faint Even at Maximum Volume

Two hard to find 55536 CVSD ICs.

While most techs would suspect either a TL082 or a TL084 amplifier (on a pre-DCS sound board) or a 1458 amplifier (on a System 11 MPU) as the cause of this issue, the 55516/55532/55536/55564 CVSD (Continuously Variable Slope Delta-Modulator) can also cause this issue.

Isolation to the CVSD is pretty simple.

• Place the game into sound test.
• Use an oscilloscope to probe pins 9 and 12. A scope must be used since the signals are too subtle for a logic probe to detect.
• Pin 9 is a clock signal.
• Pin 12 is a data signal.

When the game is making sounds that use the CVSD, both pin 9 and pin 12 should show "off nominal" signals. If the CVSD is working correctly, the output at pin 3 should be showing signal also. If it is not, then the CVSD has failed.

5.11.2.6 Some Sounds Lower Volume Than Others

The sound section of a System 11 MPU and the 2N3904 amplifying transistor.

Another sound anomaly manifests as some sounds being more faint than others is a failure in the 1408 DAC sound lane channel. In High Speed, for example, "Sound Test" exercises both the 1408 and 55536 sound lanes. Sounds 1, 3, and 5 are all made via the 1408 DAC lane. If those sounds are faint, the issue could be the 10uf capacitor (C8), the 2N3904 transistor (Q1) that is used as a pre-amplifier, the 3.3K resistor (R7), or the 1408 DAC. Eliminating the easy possibilities first, replace the 10uf capacitor. If that doesn't fix the issue, replace the 2N3904. The 2N3904 may "diode test" as failed, but not always. If the issue still remains, replace R7. Lastly, replace the 1408 DAC (U2). An equivalent part that may be used is the DAC0808.

A video, showing the problem and resolution can be viewed here.

The 1408 in the sound section of a System 11 MPU lost its top!

In another instance, Pin*Bot, all of the sounds were faint, except the last two in test. This ended up being a bad 1408 DAC. The last two sounds were unaffected, because the 55516 CVSD chip handled them. With a visual inspection, the 1408 in the pic could not be determined it was bad, until it was in the midst of being desoldered from the circuit board.

5.11.2.7 Voice callouts are much softer than music/sound

The sound mixing section on the System 11A and 11B MPU boards were populated with different resistor and capacitor value parts. If an 11A MPU is installed into an 11B game (F-14 into an Earthshaker for instance), it is possible that the voice callouts will be noticeably softer than the other music/sound. The following table details the difference between the two sound mixing sections (Thanks to Pinsider "Mneubey").

System 11 (nothing) sound section. Image courtesy of PinSider "DumbAss"

  Resistors

    R21 is 12K in 11B, 15K in 11A
    R22 is 22K in 11B, 27K in 11A
    R19 is 56K in 11B, 27K in 11A
    R16 is 180K in 11B, 56K in 11A
    R17 is 270K in 11B, 220K in 11A
    R15 is 180K in 11B, 43K in 11A
    R20 is 12K in 11B, 27K in 11A

  Capacitors

    C12 is 470pf in 11B, 1200pf in 11A
    C11 is 100pf in 11B, 180pf in 11A
    C10 is 470pf in 11B, 1800pf in 11A

System 11B sound section. Image courtesy of PinSider "DumbAss"

5.11.2.8 Sound Board D-11581 Mixer Section Resistor Values

"Mixer" section resistor values across D-11581 sound board games. Data courtesy of PinSider "DumbAss"

The sound mixing section on the System 11A through 11C sound boards were populated with different resistor values. The table shown at left details the differences between the sound mixing sections for each game.

Note that the value of R13 will change also. It should be 4.99K for "Fire!" through "Police Force", and 10K for all other games.

5.11.2.9 Missing Some Music or Garbled Music with a D-11581 Sound Board and 5 Bongs

A sound board booting with 5 bongs indicates a failure of the U20 ROM. While U20 could, in fact, have failed, 5 bongs could also mean that the software can't access U20. In one such case where 5 bongs were heard, during music test, tests 4 and 6 would play music but the music was garbled and/or incorrect. The root cause was W10/W11 jumper settings on the sound board that didn't match the ROM size. W10/W11 should be set to the left for 27256 ROMs; to the right for 27512 ROMs.

5.11.2.10 Modifying a pre-System 11C Sound Board to use 27512 ROMs (D-11581)

D-11581 sound board jumpers W2 and W3. Note that W3 is "always IN". The solder mask over the W3 jumper trace has been removed for clarity. Normally the trace is covered by a dashed white silkscreen line.

D-11581 sound board jumpers W2 and W3. W3 has been cut to enable use of 27512 ROMs.

Some revisions of the D-11581 sound board will not accept 27512 sound ROMs without modification. By the time that revision C was manufactured, jumpers W10 and W11 were added above U20 to enable use of larger ROMs. Prior to that, jumpers W2 and W3 were implemented, but were of little use since both jumpers tied A15 (pin 1) of each EPROM to +5VDC. Making the modification pictured at right, then installing jumper W2, enables the use of 27512 ROMs.

5.11.2.11 Jokerz specific sound issue

A special case of interference is present on the Jokerz game, which uses a unique stereo sound board. A deficiency in its design prevents all of the noise from being eliminated from the board. Details can be found in the original Williams service bulletin here: Jokerz service bulletin at IPDB.org

5.11.2.12 Whirlwind Doesn't play Sound 7 During Test

This particularly annoying issue appears to have no impact during gameplay. The only manifestation is that "Sound Test", sound 7 doesn't play. The normal sound is an explosion.

This issue is caused by old game ROM versions at U26 and U27. Revision PA1 doesn't send the command to the sound subsystem to play sound 7, even though sound 7 is displayed on the alphanumeric display. Updating U26/U27 to "LA3" as can be found at www.ipdb.org will correct this issue.

Correct Whirlwind sounds are demonstrated in a YouTube video found here.

5.11.2.13 Sound Board Theory of Operation (D-11581)

D-11581 Sound Board, Completely Populated and including jumpers W10 and W11

Major components of the D-11581 sound board are:

• 68B09E externally clocked CPU, which is also used in all WPC MPU and pre-DCS sound boards
• 68B21 Peripheral Interface Adapter, capable of running at 2Mhz
• Up to 3 ROMs. All boards can accommodate 27256 ROMs. Some later boards can be jumpered to accommodate 27512 ROMs.
• 6116 RAM (2K bytes). Some later boards can be jumpered to accommodate a 6264 RAM (8K bytes)
• YM2151 sound synthesizer paired with the YM3012 DAC (outputs channel 1 and channel 2)
• 55516 CVSD (continuously variable slope delta) audio encoder, which provides a 3rd audio channel
• 1408 DAC (digital to analog converter), which provides a 4th audio channel
• Various 74XX ICs that perform address decoding and chip selection logic
• 1458 pre-amplifiers and TDA2002 amplifiers

The D-11581 sound board shares a lot of board design and architecture of the pre-DCS sound board that was introduced with WPC games.

Power

Power requirements for the board are supplied via connector J3. The typical sound board power levels are required, 5VDC, 12VDC, -12VDC, and Ground. Three 100uf capacitors reduce noise on the power busses along with several small ceramic caps which filter high frequency noise.

Reset

The on board 68B09E is reset via the external reset signal provided by the MPU board, via connection J4, pin 18. If attempting to operate the board without the MPU board connected, the processor must be manually reset via momentarily grounding pin 37 of the 68B09E.

Clocking

The clocking signal for the board originates at the 8Mhz crystal oscillator. U16, a 74LS74 Dual D-Type FlipFlop divides the clock signal yielding both the "E" and "Q"clock signal. E and Q are phase shifted by one half clock cycle.

Chip Select Logic

Chip select is accomplished via the normal "memory mapped I/O" method, using U14 (74LS139) and U15 (74LS138).

• Chip select for all three ROMs (U4, U19, U20) is implemented by U14.
• Chip select for the RAM (U5), YM2151 (U3), and the 6821 (U12) is implemented by U15.

How sounds are produced

Sound select signals are provided via J4, a dual row .100 connector of 20 pins. These sound selects are held "normally high" via a 4.7K/470pf 10-pin bussed SIP pull-up resistor/capacitor (SR1) tied to 5VDC. These SIP resistor/capacitor parts are no longer available. A bussed 4.7K 9-pin resistor network can be used as a replacement. The capacitor feature of the original part filtered high frequency noise from the communication lines. In practice, noise doesn't impact the interface significantly enough to be of concern.

The MPU board will pull the appropriate sound select signals to ground via the 6821 PIA at Uxx. When the appropriate signals are pulled to ground, the PIA is programmed to interrupt the 68B09E processor via the FIRQ and NMI signals. The 68B09E then reads the sound select signals presented at the 6821s input port (PB0 through PB7) via the processor data bus (D0 through D7).

The processor then employs the 4 sound channels available to it, along with encoded sound files contained in the ROM images, to create the intended sounds.

5.11.3 Where Do The Sounds Come From?

The following sections can be used to ensure that all sounds are played by the Williams combined MPU and Sound Board System.

5.11.3.1 System 9 Games (alphabetical)

Game	Audio Demo Link
Comet	Audio
Sorcerer	Audio
Space Shuttle	Audio

5.11.3.2 System 11 Games (alphabetical)

Game	Audio Demo Link
Bad Cats	Audio
Bally Game Show	Audio
Banzai Run	Audio
Black Knight 2000	Audio
Cyclone	Audio
Diner	Audio
Earthshaker	Audio
F-14 Tomcat	Audio
Fire!	Audio
High Speed	Audio
PinBot	Audio
Pool Sharks	Audio
Rollergames	Audio
Space Station	Audio
Swords of Fury	Audio
Taxi	Audio
Whirlwind	Audio

5.12 Flipper problems

5.12.1 Flipper Circuit Schematics Error

Correct wiring schematic for serial wound flipper coils

Incorrect wiring schematic for serial wound flipper coils

Early System 11 and 11A games used serial wound coils. These coils only have a single diode on them. Some of the early System 11 manuals show incorrect wiring for the EOS switch. One of the EOS switch wires should be soldered to the center lug of the coil. However, the other wire should be soldered to the lug with the non-banded side (anode) of the diode, not the lug with the banded side (cathode) of the diode. Manuals which have this misinformation are High Speed, Grand Lizard, and Road Kings. Millionaire and Pinbot appear to be fine.

5.12.2 Flipper Electrical Problems

System 11 (nothing) and System 11A games derive 50VDC flipper power from the "50V Flipper Power Board". System 11B and 11C games derive 50VDC flipper power from the "Aux Power Supply" board that was introduced with these games.

System 9 and 11 flipper circuitry is identical to all prior Williams System 3-7 games. The design had not changed from even the old electro-mechanical flipper implementation when the End of Stroke switch was introduced.

Yes, Williams used serial wound coils from High Speed to Millionaire and then converted to parallel wound coils beginning with F-14 TomCat and throughout the rest of the System 11 run and into the WPC run. But the basic operation of System 9/11 coils remained "EM-like" whether serial or parallel wound.

There are only three switches that can affect the electrical operation of a System 11 flipper, including strength.

1. Switches inside the Flipper Enable relay (K1), located on the MPU board.
2. The cabinet flipper button switches.
3. An End-of-Stroke (EOS) switch associated with each flipper coil and integrated into the flipper mechanism assembly.

5.12.3 Series Wound Flipper Coil Operation

When the flipper enable relay (K1) switches close, a path to ground is provided for the flipper coil. The EOS switch is closed, creating a short across the low power windngs, that removes resistance of the low power windings from the circuit. Only the smaller number of high power windings are "in circuit" with the EOS switch closed.

When the flipper cabinet switch is closed by pressing the flipper button, this is the final switch in the path to ground for the waiting power at the low resistance high power windings of the coil. The coil plunger is pulled into the coil and the flipper "flips". Near the end of the flipper's mechanical stroke, a pawl (a tab that sticks up) on the flipper crank mechanism will open the EOS switch, removing the short across the low power windings, and placing them in series along with the high power windings. Together, the total resistance of the coil becomes the sum of both the low power and the high power windings. With more windings, the coil will have higher resistance and hence, lower power. The combined high resistance windings act like a "hold" or "coin door lockout" relay coil, able to withstand being energized for extended periods of time. Without the mechanization of the EOS switch, the high power windings would overheat quickly and damage the coil. For this reason, EOS switch adjustment, and a clean EOS switch, are critical.

5.12.4 Parallel Wound Flipper Coil Operation

Parallel Wound Flipper Coil Operation. Diagram courtesy of Chris Hibler.

When the flipper enable relay (K1) switches close, a path to ground is provided for the low power windings of the flipper coil. And, since the EOS switch is closed, the high power windings also have a path to ground.

When the flipper cabinet switch is closed by pressing the flipper button, this is the final switch in the path to ground for the waiting power at both of the coil parallel windings. Since electrical current "takes the path of least resistance" the lower resistance high power winding takes most of the current and the flipper "flips". Near the end of the flipper's mechanical stroke, a pawl (a tab that sticks up) on the flipper crank mechanism will open the EOS switch, breaking the path to ground for the high power windings. The high resistance low power windings remain in circuit, and provide "hold" power to the flipper coil. The high resistance low power windings act like a "hold" or "coin door lockout" relay coil, able to withstand being energized for extended periods of time. Without the mechanization of the EOS switch, the high power windings would overheat quickly and damage the coil. For this reason, EOS switch adjustment, and a clean EOS switch, are critical.

5.12.5 Flipper Mechanical Problems

• Return springs
• Fractured nylon bushing
• Insufficient gap between nylon bushing and crank
• Heat damaged coil
• Damaged coil sleeve
• Damaged coil stop
• Flipper shaft slipping within crank
• Others?

6 Repair Logs and Game Specific Problems and Fixes

Did you do a repair? Log it here as a possible solution for others.

6.1 System 11 Display Single AlphaNumeric Character Won't Light

A solder bridge on the "digit strobe" connector.

In the category of "things that will make you crazy", the solder bridge pictured at left shorted pins 8 and 9 on the connector that carries the digit strobes to a System 11 display. This prevented one of the digits on the display from ever lighting. It's always a good idea to examine prior work closely. Assuming the prior tech knew what they were doing is not always valid.

6.2 Using a System 6/7 Speech Board in a Comet

A Modified System 6/7 Speech Board Functioning with a System 9 CPU

Should the need arise, a System 6/7 speech board can be modified to use 2732 EPROMs for a Comet.

6.3 Using a Data East MPU Board in a System 11 Game

Whaaaat?!?! Yes, a Data East MPU can be used in a Williams or Bally System 11 game. However, this is limited to System 11C games only, and only a DE 520-5003-03 (Version 3) MPU can be used. The reason for this limitation is because games previous to 11C have some or all of the sound section circuitry on the MPU board populated, the 11C does not. Likewise, DE boards do not have any sound section circuitry located on them by design. The board swap is not a simple drop in replacement though. Some minor modifications have to be performed for the DE board to function properly in an 11C game.

The first hurdle is to properly power the DE board. Connection 1J17 is where power originates on the 11C board. The complimentary connection on a DE board is CN17. The problem is that these connections are keyed differently between the two boards. 1J17 has a key at pin 7, while the key on CN17 is at pin 8. The safest way to make a connection to CN17 with the 11C game harness is to remove pin 7 on the DE board. This will have no ill effect, if the DE board is to be placed back into a DE game. The worst case scenario if pin 7 is removed and the board placed back into a DE game is that the CN17 connector housing could be installed off by one pin, thus shorting the +5v to ground, and blowing the logic voltage fuse.

Correct jumper setting on DE MPU for 11C game ROMs

Addition harness and Z-connector installed to correct special solenoid drive outputs

Next, the game ROM jumpers have to be set correctly. Since all 11C games use dual ROMs, jumper setting J4 must be in on the DE MPU, while J5 is out. Install the 11C game ROMS in the following positions. U27 (ROM 1) is to be inserted into the 5C socket, while U26 (ROM 2) is installed into the 5B socket. Notch on the chip / socket is facing to the right. There is an alternative method. The code for the dual ROMs can be burned into a single 27512 EPROM.

Finally, there is a problem with how DE boards poll the special solenoid drivers differently than 11C boards. Two of the drivers are correct (solenoids 17 and 21), while the four others are mismatched. To overcome this problem, either a harness can be created (DE board to 11C harness) with a Z-connector installed between 1J19 and the game harness, or the 7-pin housing (the 9-pin housing at 1J19 is split with a 2-pin housing and a 7-pin housing) at 1J19 can be repinned. The benefit of building a harness and adding a Z-connector is that it is easily reversible, if an 11C board is to be placed back into the game at a future time.

The following wires will have to be connected to 1J19 / CN19 on the DE board for the game to function as intended.

Pin Connection	Stock Wiring	Swapped Wiring	Notes
1	ORG-VIO	ORG-VIO	No change. This is for the flipper enable.
2	ORG-GRY	ORG-GRY	No change. This is for the flipper enable.
3	BLU-ORG	BLU-YEL
4	BLU-RED	BLU-ORG
5	Key	None
6	BLU-YEL	BLU-BLK
7	BLU-BRN	BLU-BRN	No change.
8	BLU-GRN	BLU-GRN	No change.
9	BLU-BLK	BLU-RED

6.4 Ball Will Not Kick Over to Shooter Lane at Game Start (Pinbot)

Trough and shooter lane switches test alright in switch test. Ball feeder coil from trough to shooter lane tests fine in solenoid test. If the motor for the visor is disconnected, and the visor does not cycle at the start of the game, the ball will not kick to the shooter lane. This may be an issue if one or both of the visor switches are not recognized too. Not certain what code revision this particular game was running.

6.5 Ball Will Not Kick Over to Shooter Lane and Drop Target Will Not Reset (Cyclone)

Fuse under the playfield of a Cyclone, which powers solenoids 1A and 5A

Both the outhole solenoid (1a) and drop target reset solenoid (5a) are the only multiplexed 24v coils in Cyclone. The power wires for these coils are brown. For some reason, Williams decided it would be a good idea to put a fuse under the playfield for these coils (which is not included in the documentation). The fuse/fuseholder is located just to the right of the spook house subway.

6.6 Special Solenoids Are Not Pulsed During Solenoid Test on System 9 Games

The System 9 circuit board architecture provides no physical way to control the special solenoids other than via the enabling switch associated with the special solenoid device.

6.7 CPU energizes a coil 5 times

Police Force - After the pinball is switched on the ramp diverter coil engages 5 times. After the game start the diverter coil engages 5 times again. Diverter mechanism doesn't hold in position. So after the coil is energized the mechanism falls back and so the switch is closed again. Then the software tries another 4 times to resolve the problem.

Resolution: Clean the diverter mechanism and in particular the plastic. Lubricate the plastic lanes!

6.8 Elvira and the Party Monsters drop target reset coil blows the lamp matrix

The drop target reset coil should be mounted with the solder tabs down. If mounted with the solder tabs up, (toward the playfield) the solder tabs can contact the J-A-M lamp sockets. This blows the lamp matrix ICs on the MPU board by sending coil voltage through the lamp matrix.