Sega/Stern White Star Repair

Click to go back to the Data East/Sega/Stern pinball repair guides index.

2 Games

2.1 Sega

• Apollo 13
• Goldeneye
• Twister
• Independence Day
• Space Jam
• Star Wars Trilogy
• The Lost World Jurassic Park
• The X Files
• Starship Troopers
• Viper Night Driving
• Lost in Space
• Godzilla
• Harley-Davidson
• South Park

2.2 Stern

• Striker Xtreme
• Sharkey's Shootout
• High Roller Casino
• Austin Powers
• Monopoly
• NFL
• Playboy
• Roller Coaster Tycoon
• The Simpsons Pinball Party
• Terminator 3: Rise Of The Machines
• The Lord Of The Rings
• Ripley's Believe It Or Not
• Elvis
• Grand Prix
• The Sopranos
• NASCAR

3 Technical Info

3.1 The White Star Board Set

Typical Early Stern White Star Boardset (Monopoly)

The White Star System board set consists of three boards mounted to the the back box, but four boards total. The CPU / Sound board, I/O Power Driver board and the Display Power supply. Additionally, there is a dot matrix display controller mounted to the backside of the dot matrix display. On some games, the dot matrix display controller is mounted to the backbox instead.

The first two White Star games, Apollo 13 and Goldeneye, both used an additional solid state flipper control board. Starting with Twister, the solid state flipper board was abandoned, and flipper control became an additional duty for the I/O power driver board.

In addition to pinball machines, the White Star board set was used in several redemption games made by Sega and Stern. Some of the games are Wack-A-Doodle-Doo, Sega Sports, Titanic, and Monopoly redemption.

3.1.1 CPU / Sound Board

• Generations of the Sega/Stern White Star MPU
• 

  Sega White Star CPU/Sound Board 520-5136-00 Rev. D with 3 channel audio output (Apollo 13)
• 

  Sega White Star CPU/Sound Board 520-5136-16 Rev. D (Wack-A-Doodle-Doo)
• 

  Stern White Star CPU/Sound Board 520-5136-16 Rev. E (Monopoly)
• 

  Stern White Star CPU/Sound Board II 520-5300-00 (LOTR and forward). U201, U213, and U413 are socketed on this board as a result of alkaline damage abatement.

There are two versions of the White Star CPU/Sound board.

The board used in Sega machines starting with Apollo 13 (A13) and Stern machines up to Terminator 3 (T3) are the standard version as shown above with the BSMT2000. There were different revisions of this board, which included slight component changes and modifications. However, there were some distinct changes to the board from revision D to revision E. The Q9 transistor located just below U212 (RAM memory) was removed, and a two pin jumper (JP1) was added. Additional circuitry allowed for increased RAM memory from 8K to 32K. Jumper JP1 is used to select between use of a 6264 or a 62256 RAM. Note: While a 62256 RAM can be used in any game MPU, Sharkey's Shootout specifically requires a 62256 RAM.

Stern machines starting from Lord of the Rings (LOTR) used a different version of the White Star CPU/Sound board called White Star CPU/Sound Board II or more commonly WhiteStar (modified), part #520-5300-00. Stern had this in-between version made because the BSMT2000 audio chip wasn't available anymore and having it reproduced was too expensive. The White Star II therefore uses an BSMT2000 emulation circuit consisting of an Atmel AT91R40008 microcontroller and an Atmel AT49BV1614-11TC 16Mbit flash memory. Later boards use a AT49BV162AT-70TI flash memory chip. Both flash chips are obsolete now. The boards are known for sudden sound failures sometimes because of a bad flash chip and it is advisable to buy a spare. If the amplifier section is good the most common cause for sound failures is the U414 Xilinx CPLD. This chip is a programmed by Stern via the JTAG connector. While some repairmen have the skill to replace the chip, only Stern can "flash" the code into the chip.

Note: Stern states that the Atmel version board is backwards compatible with the standard (BSMT2000) version, so it can be used in *some* White Star based games (in most cases, Stern branded White Star games only - including South Park and Harley Davidson) prior to LOTR as well. However, the only games which the board can be installed in must use 8mb EPROMs for the sound section versus the 4mb EPROMs more commonly found on Sega White Star based games. If the CPU/Sound board II is used in early games, Starship Troopers for instance, some sounds may not render correctly.

Additionally...The original BMST2000 based board is not forward compatible with later games like LOTR, etc.

The BSMT2000 based White Star sound system is nearly identical to the sound circuitry found on Data East/Sega soundboards 5020-5050-0x, 5020-5077-00 and 5020-5126-02.

The board uses TIBPAL16L8 PAL ICs. These are programmable logic chips Stern used to condense a great deal of TTL logic into a single chip. Each PAL is marked with a colored dot for identification and configuration management. U213 is one of these PALs that is within the battery corrosion area. It is readily available, already programmed under part number 965-6504-00 (blue dot), the only exceptions to the blue dot being Sharkey's Shootout which uses part number 965-5023-00 (gold dot) and Lord of the Rings LE with a shaker motor installed which uses a different part number (the standard LOTR without a shaker motor uses the standard 965-6504-00 but the software version supporting the shaker motor will only run on the alternative U213).

There are two more PAL chips on the board at U19 (yellow dot, 965-0136-00) and U20 (white dot, 965-0137-00). These aren't needed very often for repair of the sound section but they do fail now and then.
Sega first used standard TIBPAL chips and later switched to reprogrammable PALCE devices. The latest boards from Stern use GAL chips. In the old PAL chips, fuses are burned so they do not have a data retention period. For the programmable logic chips the data retention period is generally specified at 20 years or more at operating temperature. The first White Star boards are from 1995 Beginning in 2015, all with reprogrammable chips begin to exceed the data retention period and although big problems are not to be expected, it wouldn't be surprising if these parts begin to fail.

The programmable logic in the board design makes it harder to repair the boards because of the grey area it causes in the schematics as there aren't any logic diagrams of these available. In short; you don't know what they do so you don't know what the output should be.

"Blink Codes" do not exist for the 68B09E or the Atmel based MPU. With one exception (Atmel based board) anything blinking is a side-effect of improper board operation. The blinking may help you diagnose the actual board problem, but blinking LEDs are not driven with intent by the processors. i.e. There are no "Bally -17-like Blink Codes".

The single exception to this is LED1 on the Atmel based MPU. LED1 on this board will blink anywhere from 5 to 8 times to indicate the sound operating system version contained in the flash memory at U8. Stern has provided a good explanation within Service Bulletin # 157. This service bulletin also describes the process to update flash memory to a newer version of the sound operating system (which is not recommended unless you encounter sound issues after moving an MPU from one game to another).

There are two JTAG (Joint Test Action Group) connectors available on the Atmel based board. One connects to the Xilinx CPLD (Complex Programmable Logic Device). The other connects to the Atmel R40008 processor. The JTAG interfaces to this board and their use have not been documented publicly.

As of April 2014, schematics for the 520-5136-16 board are no longer available from Stern's website.
As of April 2014, schematics for the 520-5300-00 board are no longer available from Stern's website.

3.1.1.1 BSMT2000 DSP Info

BSMT2000 audio chip

Like the earlier Data East boards, the first White Star generation uses a TI DSP for sound processing. The chip is just a relabeled mask programmed stock Texas Instruments TMS320C15NL-25 DSP. The 8K mask programmed ROM is custom programmed during manufacturing and can not be read out, reprogrammed, or altered in any way. Only the ones labeled BSMT2000 will work in a pinball. For prototyping, the TMS320P15NL-25 and TMS320E15NL-25 were available which included a more-or-less standard 27C64 EPROM. The pictures on the left shows a BSMT2000 and a TMS320P15NL-25 out of a Sega board. The hand labeled chips are rare but if you have one in your pin, you can read it out. The security fuse was not set on the device in the picture. If you build an adapter using a 27C64 and some logic which puts the DSP in microprocessor mode and does the address decoding it might even be possible to use the C10 and C15 devices because the internal ROM is not used in this mode.

3.1.1.2 WhiteStar Switch Circuitry

Both versions of rhe WhiteStar CPU/Sound board employ an 8 by 8 (8 switch columns, 8 switch rows) switch matrix. In addition to the typical switch matrix, an extra column of switches, which are dedicated to specific functions like coin door switches, appear on all WhiteStar games.

Flipper cabinet and end of stroke (EOS) switches are part of the dedicated switch set on games after Goldeneye.

A combination of microswitches, opto switches, leaf switches (on most drop target assemblies), and reed switches are used.

3.1.2 Power Driver Board

The Power Driver board supplies fused power, GI power and fusing, lamp matrix circuitry as well as coil and flash lamp drivers. A watchdog circuit also must communicate with the MPU periodically to allow the MPU to boot and operate properly.

As of April 2014, schematics for the 520-5137-01 board are no longer available from Stern's website.

As of April 2014, schematics for the 520-5137-64 board (Sharkey's Shootout Tournament Edition only) are no longer available from Stern's website.

As of April 2014, schematics for the 520-5137-01 board (100 ohm resistors added on J2 aux. port lines) are no longer available from Stern's website.

High current coils (50vdc) 1-8 (group 1) and 9-16 (group 2). Q15 and Q16 typically drive the left and right flippers respectively.
Low current coils (20vdc) 17-24 (group 1) and 25-32 (group 2).

Please be advised that four revisions of these board exist. The letter after the number indicates the revision. See below. All existing Stern schematics are wrong.

3.1.2.2 Revision E: Second revision

Stern White Star I/O Power Driver Board Rev E

• Not fully stuffed. Parts for flipper boards connector J4 and J5 are not stuffed.

3.1.2.3 Revision F: Third revision

Rev F of the Sega/Stern IO Power Driver Board.

• C25 changed from 470µF 100V to a 330µF 160V Snap-In capacitor
• U210 pin 3 is cut to allow for a 10% input voltage tolerance before resetting, aka "low line voltage". This was documented in Stern service bulletin #112 which can be found here. However, a better modification for U210 is to tie pin 3 to 5VDC, which sets the voltage watchdog to trip when voltage drops 5% (to about 4.75VDC). This is discussed in the PinWiki, here.

• J2 modified and 100 Ohm resistors R258 to R260, R273 -R277 added (see schematics)

3.1.2.4 Revision G: Fourth revision (not documented by Stern)

Rev G of the Sega/Stern IO Power Driver Board.

The most obvious change to the Rev G driver board is the addition of eight 100 ohm resistors under the J2 aux out connector (R258-260, R273-277). This board is backwards compatible with all machines that use the I/O Power Driver board. It is possible to add these resistors to the back of J2 for I/O board Rev F, jumping from the connector pins to the capacitors below the J2 connector to make the board compatible with those games that use the J2 connector.

This is the only version that will work with the TOPS system.

Parts deleted:

• Original R256, D228, R255 and C236 above U210 were removed

Modification 1:

J3 Pin 8 is no longer a key pin. The installed pin is connected to R255/R256 and to C236. The other side of these parts is connected to:

• R255 to D217 banded side like R245 to R251
• C236 to GND like C229 to C235
• R256 to U203 Pin 11 (LST lamp status)

The LST trace to D221 to D227 is missing. The diodes are still connected to the status output line of the VN02 solid state relays. It seems that this status is not queried by the game software or can just be ignored.

Modification 2:

• U210 Pin 3 is connected to R257.
• R257 is connected to GND but not stuffed!

Modification 2 is not necessary. A better modification for U210 is to tie pin 3 to 5VDC, which sets the voltage watchdog to trip when voltage drops 5% (to about 4.75VDC). This is discussed in the PinWiki, here.

New parts:

• R255 1K
• R256 39K
• R257 not stuffed
• C235 470pfd

3.1.2.5 Sharkey's Shootout Specific Driver Board

Cuts and jumpers applied to the Sharkey's Shootout specific I/O Driver Board.

The now defunct Stern web site containing schematics listed a few differences in the I/O Driver Board for Sharkey's Shootout. Those differences are identified in the image at left.

One other change to the board is shown on page 5 of the Sharkey's Shootout unique I/O Driver Board. Fuse F20 is changed from a 3ASB to a 4ASB.

3.1.2.6 RottenDog WhiteStar 5V Voltage Regulator

RottenDog LM317 Replacement.

The RottenDog aftermarket power/driver board uses the voltage regulator shown at left, primarily because the older LM317 voltage regulator has gone obsolete.

3.1.2.7 Lamp Matrix

The WhiteStar system is somewhat similar to previous lamp matrices designed by Data East, Sega and Williams. The main difference is instead of 8 lamp columns and 8 lamp rows (the typical 8 by 8 matrix), two additional lamp rows are used enlarging the matrix to 8 by 10.

3.1.2.8 TOPS Tournament System Driver Board Compatibility

Only the latest revision G is compatible with the Stern tournament system (TOPS). With earlier revisions while everything including the tournament sign seem to work ok a tournament can not be started. Pressing the tournament button will result in a 'Tournament Game Paused' message displayed on the DMD. The tournament button itself will also not light up during normal operation. The J3 Aux In connector has an additional pin instead of the key pin on these boards. Because of this the LST line to the lamp drivers goes nowhere on this revision.

3.1.3 Display Controller Board

Sega White Star Display Controller Board

As of April 2014, schematics for the 520-5055-00 DMD controller board (through hole components) are no longer available from Stern's website.
The display controller board doesn't contain circuitry to provide the on board 68B09E with a "reset" pulse. The reset signal is provided from the games MPU via a pin at the ribbon cable connection (J1 pin 20). The surface mount version of the board has provisions to provide a reset signal at U10, but the necessary components are not stuffed.

Stern White Star Display Controller Board, surface mount version.

As of April 2014, schematics for the 237-0255-00 DMD controller board (primarily surface mount components) are no longer available from Stern's website.

A fractured solder joint (circled) on this board prevented the "power on" LED from ever lighting.

The surface mount version of this board sometimes has dubious solder connections at some of the surface mount resistors and capacitors. The picture at left shows a fractured solder joint on the resistor in series with the "power on" LED. The LED would never light. Another fractured solder joint on the same board prevented the reset pulse from the MPU board from ever reaching the on board 68B09E.

3.1.4 Display Power Supply

Stern 520-5138-00 Display Power Supply

This WhiteStar game system high voltage power supply (for displays) is an evolution from the larger Data East power supply. The high voltage section of this board is nearly identical to the high voltage section of Data East power supplies 520-5047-01 and 520-5047-02. Of course, those earlier Data East power supply boards regulate and supply more than display high voltages (i.e. 5VDC). As such, the WhiteStar board could be made more compact while using almost identical parts.

3.1.5 128 x 32 Dot Matrix Display

All Sega and Stern White Star games make use of a 128 x 32 "standard" dot matrix display.

3.1.6 White Star Manufacturing Quality

Factory updates to the Sega/Stern White Star MPU.

Ceramic cap, 10 ohm 1/4 watt resistor, and a trace cut.

Generally, White Star boards are manufactured quite well. The traces are very fine, making alkaline damage repair difficult. But, those batteries shouldn't be on the board, should they? The Atmel based MPU board uses a multi-layer PCB, which increases the need for clean component removal and replacement.

Pictured at left, are two factory updates to the original Sega/Stern White Star MPU (BMST2000 version). These are normal, and not "hacks".

As can be seen in the picture at right, a ceramic 100pF cap and a 10 ohm 1/4 watt resistor have been added and a trace has been severed. This modification results in the 10 ohm resistor and the cap being placed between BMST2000 "CLKOUT" signal at pin 6 and ground at pin 10.

An additional 100pF cap is placed between pin 5 "E" and 6 "!E" of U1.

Another factory update to the Sega/Stern White Star MPU.

Another factory update to the left of U206 involves two cuts, and the addition of a diode between D3 and R430. Again, this is not a "hack".

Small traces lead to some components on White Star boards, enabling better manufacturing quality.

By the time White Star boards were being manufactured, improved methods of controlling solder temperature and junction temperature at the through-holes was much better understood. In the picture at left, the "spider" connections at some through-holes, allow less heat to be used to create quality solder joints.

3.2 Recommended Documentation

As always, it is highly recommended to possess a game manual. Every game manual is full of detailed information regarding game specific switch, lamp, and coil assignments. Equally, details for maneuvering through test, audit, and bookkeeping screen menus, schematics for all boards used, and game specific mechanical assemblies are included. Hard copy game manuals can be purchased through several of the recommended pinball parts suppliers. Some Stern White Star Game manuals are available in PDF form on the Stern Pinball website. They are on the game specific pages of the Stern website. An index of games can be found in the right margin here.

As of April 2014, the Stern Pinball website no longer archives theory of operation and board schematics in PDF format.

Likewise, Stern Pinball no longer keeps an online archive of service bulletins for Sega and Stern.

3.3 The Wire Coloring Code

White Star games do not use color coding system. Instead, the wire color was marked accordingly in the associated documentation, (ie. a green wire with a brown trace is referred to as GRN-BRN, orange with violet is ORG-VIO, white is just WHT, etc.).

3.4 Trough Opto Boards

++++ Need pics of the single opto transmitter and receiver boards ++++++

Stern White Star Trough Opto Transmitter (Later Games)

Stern White Star Trough Opto Receiver (Later Games)

All White Star Games use trough opto boards at the location of the trough VUK which feeds the ball to the shooter lane. Earlier White Star games (Apollo 13 to Viper Night Drivin') use only a single opto above the VUK plunger. Later White Star games (Lost in Space to NASCAR / Grand Prix) use a two opto system. The lower opto serves a dual purpose. It is the opto just above the VUK plunger like earlier games, but is also designated as trough switch #4. The upper opto is designated as the "stacking opto". Its purpose is to identify when a ball inadvertently gets "stacked" above a ball located at the VUK plunger.

To briefly summarize the operation of White Star VUK opto boards, if an object is blocking the light beam between the transmitter opto LED and the receiver opto LED, the CPU detects this as a switch closure. When an object is not present to break the light beam, the CPU detects this as an open switch. The components used on the opto receiver board are designed to do as such.

The transmitter side is simply an ultra-red LED with a current limiting resistor. The receiving opto is the same style ultra-red LED, and is able to detect the light wavelength emitted from the transmitting LED. By using discrete components on the receiving side, the appropriate signal is sent to the CPU via the switch matrix return (row). When LEDs are paired together in this type of fashion, the receiving LED emits a very small amount of voltage, when the proper light wavelength is present. By employing such a design, failure on the receiver opto side is more common.

This particular system is contrary to how Bally / Williams WPC CPUs handle opto pairs. Equally, WPC games employ infrared (IR) opto switch pairs. If interested in learning the theory of operations for the White Star opto trough upkicker boards, please consult the manual. Sega / Stern have included some excellent, detailed, technical documentation within their manuals.

As of April 2014, schematics and troubleshooting info for the trough upkicker dual opto boards are no longer available from Stern's website.

Once significant source of trouble with these trough boards is fractured header solder joints.

3.5 Magnet Processor Board

Sega 520-5143-00 Magnet Processor Board

Only two Sega games used this style of magnet processor board - Twister and Goldeneye. When U1 for Twister is installed in the board, the board's part number is 520-5143-01. The part number for the board when U1 is written for Goldeneye is 520-5143-02. Difference between the board used on each game is the different custom PIC / GAL chip used at position U1 and the Goldeneye board has R10, R11, R12, R13, D11 and D12 stuffed. This is not always the case with boards for Twister.

3.6 Flippers

Typical White Star Left Flipper Assembly (Monopoly)

All Sega / Stern White Star flipper assemblies are solid state controlled. Apollo 13 and Goldeneye are the only two White Star games which use a separate solid state flipper control board located in the game's lower cabinet. Starting with Twister, the flipper board was abandoned, and circuitry for the flippers was solely controlled by the FETs on the power driver board.

Please note that the adjacent pic of the White Star flipper assembly has a Williams flipper link / plunger installed.

The drive circuit for White Star games is different than traditional flippers. The flipper coil is single-wound. There is no "hold" winding, just a single "kick" winding. Because the flipper will get hot if the kick winding runs continuously, after the initial upstroke, the computer continuously cycles the coil on so that it stays up. According to the manual, the flipper is pulsed about 1/12 of the time.

The flipper coil must have a snubbing diode.

A normally closed EOS switch is present. The purpose of the EOS switch is to detect when a fast-moving ball has forced the flipper down ("foldback"). Unlike a traditional or Fliptronic mechanism, the EOS switch is not used to detect end of stroke. That is, the EOS switch closing does not cause a switch from "kick" to "hold". If the flipper is supposed to be up, and the EOS switch opens, the game will re-energize the flipper coil. This helps with some features, particularly "laser" kickers that fire the ball into the flippers. The EOS switch should be present for improved play, but if it's broken, the flipper will work for the most part. However, in some instances, a flipper with a broken or missing EOS switch may be noticeably weaker. If a flipper seems weak or lethargic, the flipper mechanisms have been rebuilt, and the EOS switch is broken, try replacement of the EOS switch first before considering other options.

3.7 Accessing Bookkeeping, Settings, and Diagnostic Modes

Solid state pinball machines typically have a built in system for audits and adjustments. All White Star based machines use a system called "Portals". The Portals interface is a carry over from the last two Sega games (Baywatch and Batman Forever) which used the original Data East board set.

3.7.1 Portals (Coin Door Test Buttons)

The Portals control panel used on most White Star games (Monopoly)

The Portals system has a very user friendly, simple navigation system. Unlike previous Data East / Sega games, having to scroll through all of the audits and adjustments is a thing of the past. Diagnostic testing (including game specific assembly tests), audits, and adjustments can all be quickly accessed via the initial display menu graphical interface. The 3-button control panel (in most cases, except some earlier White Star games) used to access Portals is located on the coin door.

• DMD Screens during Various Tests in Portals
• 

  Switch test (similar to switch edge test in Williams speak)
• 

  Active switch test (similar to switch level test in Williams speak)
• 

  Dedicated switch test
• 

  Very flexible lamp test options
• 

  Coil test
• 

  Trough clear, but equally a good test for the trough and shooter lane switches
• 

  Advancing through trough clear

3.8 Coin Door Switches

Coin door Switches Located Inside a White Star Game (Monopoly)

Most White Star games have two switches located on a bracket inside the coin door (hinge side). The upper switch is the memory protect switch. This switch is used so that nothing can be written to the memory, unless the coin door is open. The memory protect switch is not part of the switch matrix, nor is it one of the dedicated switches. Although it does connect to the CPU / Sound board via the same connector as the dedicated switches, its state is processed by a different chip (U213) versus the dedicated switches (U206).

The lower switch is an interlock style "kill" switch. This kill switch is used to kill all power to the solenoids, when the coin door is opened. Since it is an interlock style switch, it can either be depressed (when the coin door is closed), or it can be extended outward to turn the switch on. To extend the switch to the on position when the coin door is open, grasp the switch, and gently pull it out to the lock position.

The kill switch is a double pole switch. Wiring to one pole of the kill switch is BLK-YEL and BLK-YEL, while the other is RED-WHT to WHT-RED (yes, the colors do not match!)

4 Problems and Solutions

4.2 MPU boot issues

4.2.1 Relocating the Battery from the MPU board

The aftermath of batteries which leaked. This board is beyond repair, and good only for salvage parts.

Relocating the 3xAA batteries off the MPU board or installing other alternatives for memory retention is always a good idea. Leaky alkaline batteries are the #1 killer of pinball boards.

One option is to install a remote battery holder, and place 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.

Another option is to install a memory capacitor. This method is considered to be better than a remote battery holder, because memory capacitors are not prone to leakage like alkaline batteries. The only drawback is that the memory capacitor has to be properly recharged to be effective. If a game has the potential of sitting dormate without being turned on for extended periods (typically months), the memory capacitor may not hold the memory.

A third option is to install a lithium battery with a battery holder. Lithium batteries are much less likely to leak than alkaline batteries, and they do not need to be recharged.

Whatever you do, GET THE BATTERIES OFF THE BOARD! It's not a matter of if they will leak, it's a matter of when. White Star boards are pricey to replace, and repair is difficult/expensive/even impossible due to the very fine traces used on the board.

A White Star MPU Board with Remote Battery Pack Connected. The Positive (+) terminal is circled. D200 and D201 are highlighted inside the rectangle. Note that the battery installation date is recorded with a Sharpie. Also note the long wires, allowing the holder to hang well below the backbox boards, another precaution for leaky batteries.

Adding a connector between the battery pack and the MPU board is a good idea. In doing this, the battery pack can easily be removed from the MPU board. Plus, if the batteries are forgotten, and do leak, the MPU board may 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. Using a diode instead of a jumper will prevent the batteries from being charged and 'cooked' by the game if blocking diode D201 on the MPU board fails. Keep in mind that a second (redundant) diode added to the circuit will reduce the backup voltage to the RAM memory by .5 to .7 volts. 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.

On the MPU, solder the battery cables: positive (red wire) to the "+" pad and ground (black wire) to the "-" pad.

Since the MPU board is already out, another good practice is to check the D201 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 rather quickly. In worse cases, the new batteries can even leak or explode if charged. Testing the D201 diode is quick and easy, and worth the trouble checking it out. When in doubt, replace the D201 diode with a 1N5817 (a 1N4004 will work in a pinch), or add a redundant 1N4004 to the battery pack. Once again, if a second (redundant) diode is added to the circuit, it will reduce the backup voltage to the RAM memory by .5 to .7 volts

Testing the D200 diode is a good idea too. The D200 (1N5817) diode is used to keep the backup batteries from powering the complete MPU board when the power is off. A symptom of a failed D200 diode are batteries which deplete rapidly.

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 VBATT test point on 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 some headaches.

4.2.2 Installing a Memory Capacitor Instead of Batteries

Installation of a Memory Capacitor

A memory capacitor is a great choice for memory retention, provided that the game will be turned on more frequently than not. If the game will not be turned on for several months, the memory cap may not stay charged. In turn, the memory may not be retained.

It is recommended to use a 1F (Farad) 5.5v memory capacitor. The lead centers on a common, button memory cap are shorter than the "+" and "-" through holes where the factory battery holder was installed. To overcome this, solder a short wire onto the negative lead of the cap. Then, bend the lead at a 90 degree angle, so it is now parallel to the bottom of the memory cap. Place the positive lead of the memory cap into the "+" of the board, and then place the wire from the negative lead into the "-" of the board. Solder both leads in place.

Please note that when using a memory cap, the D201 diode must be removed (or jumped), and a 0 ohm jumper or a 100 ohm resistor (or similar value) should be installed. This purpose of removing the D201 diode is so the memory cap can be properly charged when the game is on.

4.2.3 Repairing Alkaline Corrosion

Corrosion can hide in many hard to find places.

Sega/Stern White Star boards are well known for issues with leaky batteries. This is because the batteries are mounted on the top of the board with plenty of board beneath it for the corrosion to leach onto. Components well below the battery holder are often damaged.

Remember, battery corrosion is alkaline. Harkening back to the dim and distant past, remember your Chemistry 101. Alkaline is neutralized with acid. Most commonly used is vinegar, since it is an acid. Diluting the vinegar to half strength by adding an equal amount of water (50/50) is a good idea to prevent adverse impact to adjacent areas of the board.

This MPU sustained alkaline damage. Sockets for the 68B09E processor and the 6264 RAM have been removed.

More components removed, the alkaline damaged traces sanded, and the damaged area treated with 50/50 vinegar and water.

Argh! The MPU pictured at left has sustained alkaline damage. In this case, the damage was repairable. Most of the time, alkaline damage to White Star MPUs is more substantial and beyond economical repair.

Bare copper traces tinned.

All components (or sockets) reinstalled, and nearly ready for test.

The completed board shown on the right took a few hours to cleanup, repair, verify, and game test.

4.2.3.1 Continuity Chart of Pin Outs in Battery Damaged Area

Just below the battery holder, the CPU EPROM and RAM both reside. Both of these chips and their sockets are susceptible to battery damage. Provided that the alkali damage breached the chip sockets, it is always best to check continuity on the new work performed. Below is a chart of the pin outs between the RAM (U212), EPROM (U210), and CPU (U209) chip. This chart will hopefully come in handy should repair of this area is necessary.

Signal	68B09E Pin (U209)	ROM Pin (U210)	RAM Pin (U212)	PAL Pin (U213)	Other
Vcc	7	32	28	20
GND	1 & 39	16	14	10
A0	8	12	10
A1	9	11	9
A2	10	10	8
A3	11	9	7
A4	12	8	6
A5	13	7	5
A6	14	6	4
A7	15	5	3
A8	16	27	25
A9	17	26	24	16
A10	18	23	21	15
A11	19	25	23	8
A12	20	4	2	9
A13	21	28	26	3
A14	22	29	---	2	U211-2 (XA0)
A15	23	3	---	1	U211-5 (XA1)
A16	31	13	11		U211-6(XA2)
A17	31	13	11		U211-9 (XA3)
A18	31	13	11		U211-12 (XA4) - Prior to rev. E boards
D0	31	13	11
D1	30	14	12
D2	29	15	13
D3	28	17	15
D4	27	18	16
D5	26	19	17
D6	25	20	18
D7	24	21	19
/ROMCS	---	22 (BSMT2000) 24 (Atmel)	---	19
/RAMCS	---	---	20	18
/E	---	24 (BSMT2000) 22 (Atmel)	22	4	U1-6
/CS	---	---	26		Q9-1 - rev. D or JP1 (8K) - rev. E
R/W	32	---	27	7

4.2.4 Connecting a logic probe to the MPU

Connecting a logic probe to a Sega/Stern White Star MPU

Connecting a logic probe to a Stern White Star (Modified) MPU

Connecting a logic probe to a Sega/Stern White Star board set (left) is simple.
Both 5V and Ground test points are made available on the MPU board.

Connecting a logic probe to a Stern White Star (modified) board set (right) is simple too, but a 5V connection point isn't provided.
Connect the positive lead of your logic probe to either side of the inductor labeled FB1.

4.3 Low +5VDC and Game Resets

Some symptoms of a game which repeatedly resets upon boot up are:

1. General illumination lamps will repeatedly flash on and off (the GI relay can be heard clicking on / off).
2. The DMD will display the initial boot screen, show the display ROM revision number, go blank, and then repeat this process.
3. The sound board (depending on the age of the game) may only emit a "blip" sound or a shorter than usual initial sound clip.

In some of the above instances, the DMD display may or may not light up at all.

The +5VDC for logic power is sourced from the 8VAC secondary windings on the transformer. The 8VAC is fed to the I/O Power Driver Board, and rectified via bridge rectifier, BRDG21. The rectified DC voltage is regulated via an LM338K adjustable voltage regulator. Logic voltage can be adjusted via R116 on the driver board, which is a 50 ohm adjustment potentiometer.

U413, which is located on the CPU / sound board next to the reset button, is a Dallas Maxim DS1232 monitoring chip. In theory, should the logic voltage dip to less than 5% or +4.75VDC, the DS1232 will force a reset of the CPU. However, it has been determined that most White Star board sets will not function properly below +4.85VDC.

If the voltage on the power I/O driver board is below the +4.85VDC threshold, adjustment can be made via the R116 adjustment pot, until a satisfactory voltage is achieved. The best location to measure the +5VDC is at the bottom leg of resistor R114. R114 is located in the vicinity of the R116 adjustment pot, and just below the +5VDC LED, L2. Another location to measure the +5VDC is right on the metal casing of the LM338K regulator (U19), If a satisfactory voltage cannot be achieved, turn the game off. Remove connector J16, located above the LM338K regulator. Turn the game back on, and measure the +5V again. If a satisfactory voltage can be acquired with J16 disconnected, a board or component which uses the +5VDC is "dragging" it down. Turn the game off, and remove all 5V input connectors on all other boards at this time. Reconnect J16 again, and review the logic voltage on the I/O Power Driver board. Repeat the process of turning off the game, and reinstalling logic power connectors one at a time to determine the suspect board. Keep in mind that all opto switch receivers used throughout the game use the same +5VDC logic lines. If no other boards in the backbox appear to be suspect, an opto transmitter board may be at fault.

Should the game start randomly resetting, the first course of action is to measure the +5VDC on the I/O Power Driver Board. If the logic voltage is within spec., measure the +5VDC on the CPU / sound board. The +5V test point on the CPU / sound board is located just to the left of the 6809EP CPU chip (U209) on the board. If the voltage drastically differs between the measurement of the I/O Power Driver Board and the CPU / sound board, turn the game off. Remove and reseat connections CN2 on the CPU / sound board and J16 on the I/O power driver board.

4.3.1 DS1232 Voltage Watchdog

While this looks to be an error, it's actually a factory modification to the DS1232 reset generator. While this part is on the CPU/Sound board, there is an identical part on the driver board, which will have pin 3 clipped also. On the CPU pin 2 (Time Delay Set) should also be clipped. Pin 2 and 3 are left unconnected from the factory on the WhiteStar (modified) CPU/Sound board.

Both the WhiteStar MPU and Driver boards employ a DS1232 voltage watchdog. From the Sega and Stern factory, pin 3 is almost always clipped. Pin 3 is used to select the voltage tolerance the chip is to enforce.

A discussion regarding the proper connection of pin 3 was offered by John Robertson on rec.games.pinball, and is copied below.

Begin discussion...

You read it here (rec.games.pinball) first!

Looking into a problem with our Roller Coaster Tycoon where the game would cut out the lights and coils for up to a second at random intervals. This would happen perhaps once or twice every day or two.

Digging into the game I noticed that the Dallas supervisor chip used in both the original Stern and the Rottendog replacement Driver Board both left pin 3 (LOP) floating. Now this pin is there to select the Vcc tolerance - if grounded it is 5% and if tied to Vcc it is 10%. However the schematics now show pin 3 as floating.

Sega Service Bulletin SB112 (June 7, 1999) had recommended cutting pin 3 (to floating) on both Dallas DS1232s (MPU & Driver PCBs) where the factory (at that time) had connected pin 3 to Gnd (5%). Sega, then Stern continued this practice of either cutting or leaving pin 3 floating. Rottendog also did this on their replacement board.

http://arcarc.xmission.com/Pinball/Stern%20Service%20Bulletins/sb/sb112.pdf

I was curious what would happen and if this idea of cutting the leg or leaving it floating was recommended by Dallas. So I pulled up both the original Dallas documents and the Maxim Application Note and in neither case did they suggest that pin 3 should be floating - no connection.They said it was to be connected to either Gnd or Vcc.

https://datasheets.maximintegrated.com/en/ds/MAX1232.pdf

https://pdfserv.maximintegrated.com/en/an/AN581.pdf

I then asked Maxim (who now make Dallas Semi chips) to comment on what happens when pin 3 is floating - they replied today (Nov 25, 2019):

Thank you for contacting us.

It is not recommended to leave the TOL pin open since it is a digital input and noise can make it oscillate, there is no internal pulling resistor, so please do indeed connect it to GND or Vcc.

Regards, Javier Monsalve Maxim Applications

We had already tied pin 3 to Vcc (+5) as of Nov 23, 2019 to check my assumption that floating was wrong, and the intermittent cutting out problems had indeed disappeared!

The game has been running for days now with no hiccups and we shall be leaving it running at least one more day before sending it to its new home. We are monitoring the Vcc with a Mooshimeter logging 5V, to see if Vcc drops below any reasonable threshold and a Heathkit logic probe monitoring the R (pin 5) of the DS1232 as the Heathkit logic probe stores momentary state changes (turns on a 'memory' LED).

Therefore, based on our tests, and the advice from Maxim, the manufacturer of the DS1232 IC Supervisor chip, we can now say that you need to tie pin 3 of the DS1232 to Vcc! Do NOT leave it floating as per Sega Service Bulletin SB112!

Hope this helps!

John :-#)#

An (probably) intentionally created solder bridge on a RottenDog power/driver board, bridging pins 3 and 4 on the DS1232 voltage watchdog, selecting 5% (4.75VDC) tolerance.

RottenDog WhiteStar power/driver boards may have a solder bridge across pins 3 and 4 of the DS1232 voltage watchdog. This is done to prevent pin 3 from "floating" and instead actively select the watchdog voltage.

Quoting Ed from Great Plains Electronics on PinSide...

It appears pin 3 (voltage tolerance pin) was left floating on their (RottenDog) design. That input is supposed to be tied high (reset at 4.5V or 10% VCC tolerance) or tied low (reset at 4.75V or 5% VCC tolerance) and should never be left floating. Adding the jumper from pins 3 to 4 selects the 5% tolerance. If the 5V line dips below 4.75V then your board will reset (which is what it is supposed to do with that setting).

Jumpering pin 3 to pin 8 (5VDC) selects the 10% tolerance.

4.3.2 Poor or Damaged Sockets Cause Reset

A damaged EPROM socket from a Sega/Stern White Star MPU

A damaged socket for either the game ROM or the 6264 RAM can cause a game to continually reset. This happens because the CPU can not fetch and execute instructions from the game ROM, or it is unable to read/write data to/from the RAM correctly. A simple replacement of damaged OEM sockets sometimes corrects a reset issue.

4.3.3 Game Resets with DMD Controller Board Connected

This may be a bit of an anomaly, but it is worth mentioning. If when the DMD controller board is connected without the data line ribbon cables, and the game continues to reset, replacement of BRDG21 may be required.

In one particular instance, a game was resetting continually with only the +5v / ground connection connected to the DMD controller board. If the board was not connected, the driver board would output ~+5VDC. If the DMD controller board was connected, the driver board's logic line would dip down to around +4.85VDC. Logically speaking, this type of symptom somewhat points to the DMD controller board having issue. However, the end result was replacement of BRDG21 per Stern tech support.

It is worth mentioning that the bridge rectifier did test correctly (using a DMM in diode test) both in circuit and out of circuit, but these tests were conducted while the bridge was not under load.

4.4 Solenoid problems

4.4.1 Game Notes

• NASCAR solenoid test swaps the order of output for coils 25 and 26. While the display calls out coils 25 and 26 correctly, in fact coils 26 and then 25 fire.

4.4.2 Testing MOSFETs and TIPs

Driver Test Points TPL1 above Q16 (left) and TPL3 above R302 (right) are used to fire the transistors on the driver board.

The Sega/Stern I/O Power Driver Board contains 32 driver transistors on the bottom right of the board. Transistors Q1-Q16 are MOSFETs – either 20N10Ls or P22NEs. These are used to drive high current coils like the flippers and other heavy loads. These are connected to devices through J8 and J9. Transistors Q17-Q32 are TIP 122s and are used for driving low power solenoids such as flash lamps. These are connected to devices through J6 and J7.

All of these drivers are computer controlled directly from the CPU board through a flip-flop on the driver board. Q1-Q8 are controlled through U1 on the driver board, Q9-Q16 are controlled through U2, Q17-Q24 are controlled through U3 and Q25-Q32 are controlled through U4. Each one of these chips is selected by the address decoder at U204. There is another address decoder at U205 and both of these decoders must be working properly for the I/O board to function.

It is possible to test the driver transistor and the solenoid in the circuit from the driver board. Above Q16 on the driver board is a test point marked “TPL1 FET”. Not a lot of people know what this is for. Per the manual: “This will apply 3.4v to the gate of the MOSFET transistor thus switching it on.” You’ll notice that the bank of resistors from R1 to R16 have a white stripe on one side closest to the MOSFETs. By using a jumper wire from TPL1 to the lug of each resistor in the white band, you can fire the coil associated with that resistor/MOSFET. Each resistor number corresponds with the same MOSFET number. So R1 fires Q1, R2 fires Q2, etc, etc. TPL1 is for testing all the FETs from Q1 to Q16.

If you look in that same photo above R302, you will notice “TIP TPL3”. This serves the same purpose as TPL1 except for the TIP transistors on the driver board Q17-Q32. By jumping from TPL3 to the leg of the resistor in the white band, you can test each of the TIP122s and their associated circuitry. This is a useful diagnostic step if the machine will not fire a coil in self-test or during a game. If you can fire the transistor with this test, you know the problem is with the computer control circuitry for the solenoids. However, if it doesn’t fire, it could be a problem with the transistor or the associated components in that solenoid circuit.

If one of Q1-Q16 is bad, it is best to replace it with an IRL540N (see below). If one of Q17-Q32 is bad, it is best to replace it with a TIP102.

4.4.2.1 Screening Error On the Board

You might notice in the photo that the white stripe is on the lower leg closest to the transistors of R1-16 and R25-32, but there is no white stripe on R17-24. Instead the stripe is on the lower leg of the resistors R200-R207. This is a screening error on the board but electrically match to the correct legs of resistors R17-24. As an example, if you want to test Q17, touch the leg of resistor R200 that’s in the white strip and Q17 should fire. The same holds true for R201/Q18, R202/Q19, etc etc.

4.4.3 Can an IRF540 replace a P20N10L, P22NE10L or an IRL540?

Danger Will Robinson...the IRF540 installed at Q6 shown above, is the wrong part. The original 22NE10L should be replaced with an IRL540, like the one to it's immediate right.

No.

The White Star driver board first used STP20N10L FETs to drive high current circuits, like solenoids. When they were discontinued later boards used the STP22NE10L. Since the 22NE10L is obsolete and even the current replacement, the STP40NF10L is hard to find, the IRL540 or IRL540N are the recommended replacements. If available get the IRL540N because it has a higher drain current rating of 36A.
While the drive specs of the IRF540 seem comparable, they are not. Using an IRF540 as a substitute can cause the associated coil to lock on, toast the coil, then blow the FET. The L in all of the original parts indicate that it is a logic level FET.

Per a good friend of Pinball, Ed Krzycki, via RGP post...

Nope -- an IRF540 is not a good sub for the IRL540 or 22NE10L.

The "L" parts have TTL compatible gate voltages. A TTL gate has a minimum high level output voltage of about 2.4 volts. To turn on any N-channel MOSFET transistor - the Gate voltage must be sufficiently high enough as listed by the spec sheets VGS(th) voltage.

The IRL540 requires a Gate threshold voltage of 1 to 2 volts. A TTL part exceeds this voltage so a TTL gate's high level output would guarantee the gate would turn completely on. The IRL540 is guaranteed to have "turn on" and "turn off" voltages that are compatible with a standard TTL device.

The IRF540 requires a Gate threshold voltage of 2 to 4 volts. A TTL gate *might* turn this part on if the specific part being used has a threshold in the 2 to 2.4 range. If the specific part being used has a threshold closer to the 4 volt range then the MOSFET probably won't turn on properly or fully. The IRF540 *might* work or it might not work. And no guarantees that a working part would remain working if ambient temperature changes.

Another aspect to an IRF540: An IRL540 operates more as an on-off switch but an IRF540 does not. Depending on voltages, an IRF540 can operate as an on-off switch and can also operate anywhere in between on and off (half on) depending on input voltage. If the TTL device cannot quite drive the gate voltage high enough then the MOSFET may end up partially conducting. This can have different results depending on how much it is turned on.

The 22NE10L has similar characteristics to that of the IRL540 as far as gate threshold voltages.

4.4.4 Adding a Real Knocker to White Star Games

Some White Star games (like Starship Troopers) are not equipped with a real, old fashioned, knocker. We all love knockers, don't we?

To add a real knocker to your White Star game, follow this procedure. Note: in this procedure, we locate the knocker in the head, attempting to increase the knocker's impact. For cabinet installation, take advantage of the wiring already provided for the European Token Dispenser.

In the picture at left, we show all the components necessary to add a knocker. The new knocker is circled, top left. We've used an in-line connector along the wiring, also circled, lower left. The power and ground connections are noted by the rectangles.

A snippet from the Sega Starship Troopers manual showing coils, and in particular, coil 8, the European Token Dispenser.

First we need note that the European Token Dispenser solenoid drive is unused in domestic (USA) games. Coil power and drive connections are wired into the game. They can be found in the cabinet bottom. We could simply connect the knocker to that connection, but in this example, we mount the knocker in the game head.

Adding a wire to provide the knocker with 50V power.

Here I've stuffed another wire (yellow with purple tracer) on top of the solenoid power supply at J10, pin 5. This wire will eventually be connected to the banded side of the knocker coil.

Adding a wire to provide a path to ground for the knocker, typically called the solenoid drive.

Here I've stuffed another wire (purple with black tracer) on top of the solenoid 8 drive connection at J8, pin 9. This wire will eventually be connected to the non-banded side of the knocker coil.

The new knocker, screwed to the cabinet upper left corner...not much room there. Also add the "strike plate" to the cabinet side.

Of course, you need to procure the actual knocker hardware, and connect it as shown in the prior pics. In the picture at left, we've added a Sega knocker assembly and strike plate.

The last thing to do is to enter the game adjustments and turn the electronic knocker sound and the European Token Dispenser settings to off. On Starship Troopers, Enter Portals, Adjustments, Sega, and then adjustment 33 (Knocker Volume - set to OFF) and adjustment 40 (European Token Dispenser - set to OFF).

That's it. Enjoy those knockers!

4.5 Flasher problems

One of the 4 sets of 4 colored flashers on Starship Troopers.

More regarding flashers to be added later...

Starship Troopers, Apollo 13 and perhaps other White Star games, drives 4 flash lamps via a single TIP-122 transistor. The TIP-122 is rated for a maximum drain of 8A (and a working current of 5A) while 4 flash lamps typically have a cold resistance of under 1 ohm (0.8 ohm is typical including wire and socket resistance.) While incandescent lamps only present their minimum resistance when extinguished (they are negative-temperature coefficient devices by design) the surge drain when these lamps are lighted radically exceeds the transistor's rating and is likely to damage or destroy it over time.

These channels should be replaced with TIP-102s (15A peak drain) or replace two (or more) of the flash lamps with LEDs to bring the drain in-line with the maximum ratings of the TIP-122.

4.6 Controlled lamp problems

4.6.1 Overview of The Lamp Matrix

The WhiteStar lamp matrix is an 8 column, 10 row lamp matrix, powered by 18V.

Lamp matrix power is derived from 13VAC presented to the I/O Driver board at J17, pins 8 and 9, via the transformer secondary winding. The circuit is fused by F22, an 8ASB fuse. BRDG20, a 35A bridge rectifier creates the 18VDC which is then smoothed by two 15,000uf/25V filter capacitors at C201 and C202. Of interest is the combined filtering of 30,000uf matches the original lamp matrix filtering capacity of old Williams System 3 games. When lit, LED202 indicates the presence of 18VDC power.

A blown fuse at F22 probably indicates a shorted bridge at BRDG20.

Components of the WhiteStar I/O Driver Board Lamp Matrix

Lamp rows are implemented with STP19N06L FETs, controlled by the 74HCT273 at U6 on the I/O Driver Board. 39K current limiting resistors protect the 74HCT273 from sinking too much power. The 74HCT273 is strobed (enabled) via the LMPSTB which originates at U205 (a 74LS138 at U205, pin 15).

If a STP19N06L transistor fails, you can replace it with an IRL530N or an IRL540N. An IRFnnn (i.e. IRF530) or BUZ22 can not be used as a replacement.

Lamp columns are implemented by the obsolete VN02N PENTAWATT solid state relay (5 legs). They are controlled by the 74HCT273 at U18 on the I/O Driver Board. 18VDC is "strobed" to the lamp column headers. The 74HCT273 is strobed (enabled) via the LMPDRV which originates at U205 (a 74LS138 at U205, pin 13).

A replacement for the obsolete VN02N has not been identified. In addition to the WhiteStar I/O Driver board, VN02Ns were used on Capcom driver boards.

Be sure to check the fuse clips for the controlled lights. They go bad from either metal fatigue or they crack. This causes poor, erratic, or no connection with the fuse and thus, no lights.

4.7 General illumination problems

The four general illumination (GI) strings are turned on and off via the relay on the power driver board. On the White Star board the relay is controlled via transistor Q200 and latch U206. This is more or less the same setup found in earlier Data East and Williams System 11 games where the relay is located on the power supply or PPB board. These relays seldom fail. The part number for the 24VDC relay is any of the following:

• LB2-24DP (Great Plains Electronics number)
• 190-5002-00
• LY2-0-DC24
• FRL264 P024/02CK
• R14-11D10-24P

If one or more GI strings are out check fuses F24 to F27 and then connector J15. Although WhiteStar GI connectors don't burn and tarnish like early Bally/Williams WPC GI connectors, checking both the male and female connectors at J15 is worthwhile.

A leaky capacitor at C32

If the GI goes out during game play and turns back on when opening the door, check capacitor C32 on the power driver board. It normally has nothing to do with either the GI circuit or the relay voltage but a leaking cap can short the trace coming from Q200 to ground. In this case the relay pulls in and disconnects the GI. When you open the door the game cuts the coil power for the relay coil and the GI turns back on. This has been observed in a Twister game.

The fuse clips for the G.I. circuits also go bad by either cracking or just plain fatigue. This causes poor, erratic, or no contact with the fuse thus killing power to the lights.

4.8 Switch problems

4.8.1 Overview of the Switch Matrix

The WhiteStar switch matrix is the typical 8 by 8, 64 switch matrix. As with other switch matrix implementations, isolation diodes are used on every switch in the matrix.

Switch columns (strobes) are connected to CN5 on the top edge, left side, of the CPU/Sound board. The columns are controlled by the 74HCT273 at U203 which sequentially controls an array of 8 2N3904 transistors. These transistors switch the 5VDC, normally held high via 560 ohm pull up resistors R423 through R430, to ground.

If a switch in the matrix is closed, the ground driven through the switch column will be presented at the row associated with the column/row intersection of that switch.

Switch rows (returns) are connected to CN7 on the top edge, right side, of the CPU/Sound board. 5VDC pull up resistors (560 ohms at R401 through R408) normally hold the switch row high. However, when the ground is presented at the row header of CN7, that 5VDC is pulled to ground. The signal is compared to 5VDC by the LM339 comparators at U400 and U401 with the result presented the the 74HC245 at U205 and then read via the 68B09E's data bus.

Note that the WhiteStar LM339 comparison implementation differs from the Bally/Williams WPC implementation which also uses 5VDC as a comparison voltage (reference) but compares that 5VDC to either 12VDC (open switch) or ground (closed switch).

4.8.2 General Switch Matrix Testing

Testing Whitestar CPU/Sound board switch matrix circuitry with a jumper. A diode is not necessary.

Like most modern game manufacturers, Sega/Stern games employ a switch matrix. The switch matrix circuitry on the CPU/Sound board may be tested (as pictured at left) by jumpering between switch row and switch column header pins while in switch test. This works with all versions of the Whitestar CPU/Sound board.

Note that some games, that do not use the full switch matrix, will fail to report a switch closure when jumpering row and column header pins for a switch that is shown as "not used" by the game. i.e. Sopranos does not use switch column 6. While a strobe is sent out on column 6, jumpering between column 6 and any row will not return a switch closure indication in switch test.

4.8.3 Playfield Switch Problems

4.8.3.1 Modular Stand Up Targets

The following Sega games employ unique playfield "Stand Up" targets:

• Goldeneye
• Twister
• Space Jam
• Star Wars Trilogy
• The Lost World Jurassic Park
• Starship Troopers
• Viper Night Drivin'
• Lost in Space
• Godzilla
• Harley-Davidson
• South Park
• Playboy (used behind the "Peek-A-Boo" stand up)

They are self-contained modular targets, made by Happ Controls, that are quite easy to remove or replace. The target "faces" may be purchased separately and may be interchanged easily. The unit has a self-contained diode, like all switches that are sensed via a switch matrix. It also contains a small micro-switch, that the "tail" of the target face actuates. The "Achilles Heel" of this target system is the IDC connector. As with all IDC connectors, the wires can pull out fairly easily.

Detailed disassembly / assembly instructions can be reviewed in Stern Service Bulletin 110.

• Sega Modular Targets (from Starship Troopers)
• 

  An example target. The target "faces" are interchangeable.
• 

  "Exploded" target parts. The gray switch part must be oriented correctly. Note the integrated diode just under the micro-switch, center/bottom.
• 

  The Achilles Heel of a Sega target...the IDC connector.
• 

  The first step to disassemble a target. Slide the rear face of the target assembly up.
• 

  Testing a closed target switch, step 1. In diode test, with the switch closed, your meter should read 0 or null.
• 

  Testing a closed target switch, step 2. In diode test, with the switch closed, your meter should read .5 to .7, the nominal voltage drop across a diode.

4.8.3.2 Magnetic Reed Switches

Sega Magnetic Reed Switch (from Space Jam)

Starting with Space Jam, Sega introduced magnetic reed switches. This type of switch is a sealed, non-serviceable switch. Theory of operation of the reed switch can be reviewed in service bulletin 89. Replacement part number is 180-5145-XX.

4.8.3.3 Ball Trough Switch Problems

Typical Sega ball trough assy. using 1/8" screws heads

Typical Stern ball trough assy.using 3/16" screws heads

As with any sub microswitch, the switches can become sporadic or fail over time. Another thing that can happen is the switch arm can break at its pivot point. This happens with this style of switch more frequently than other styles, because the metal arm is placed between two plastic ears. With nearly every other style of sub microswitch, the arm has the metal ears which go over plastic nubs on the switch body. The correct replacement switches are 180-5119-00 roller arm switches. Over the years Sega / Stern used both 1/8" and 3/16" screw heads on the screws which secure the trough switches. At this time, it is uncertain when Sega / Stern changed the screw size.

Xcelite 1/8" Nutdriver used to remove screws from trough switches

Ball trough assembly with divots. Although it's not a White Star game, Jurassic Park is used for this example.

There may be other issues, which appear to act like ball trough switch failures. One of the problems is divots in the ball trough itself. After many iterations of the balls slamming into the trough over and over, divots start to develop. This divots can keep balls from rolling down the trough assembly. In turn, the balls will not be closing the proper switches in the trough. The end result is a game that "loses its mind". If the balls are not where they are expected to be in the trough, the game cannot function as intended.

A similar issue to divots in the trough is magnitized balls. This occurs primarily in games where a magnet captures a ball and holds it for a length of time. Some examples of games with magnets that hold the ball are Apollo 13, Goldeneye, Twister, Space Jam, and X-Files. This problem typically happends when two or more balls are grouped together in the trough. They will literally be held in place and stick to the trough. This happens to the balls located furthest from the ball trough VUK. To determine if a ball or balls are magnetized, remove the lower apron, and the top bracket over the trough. Gently place a non-metallic object between the balls to see if they separate themselves.

For ball trough switch issues pertaining to the opto boards, see the section below.

4.8.4 Opto Switch Problems

With White Star games, opto switch pair closures are analyzed by the CPU in the same manner as leaf switch pairs or normally open (NO) microswitches. When a leaf or NO microswitch closes, the end result is identified by the CPU as a valid switch closure.

Two sets of opto pairs installed in a Sega X-Files

If an object is blocking the light beam between the transmitter opto LED and the receiver opto LED, the CPU detects this as a switch closure. When an object is not present to break the light beam, the CPU detects this as an open switch.

The transmitter side is an ultra-red LED with a current limiting resistor. The receiver is an identical ultra-red LED. When light from the transmitter shines on the receiver LED, a small amount of voltage is "excited" in the LED. By using discrete components on the receiver side, that small voltage is detected and the appropriate signal is sent to the CPU via the switch matrix return (row). In this design, failure on the receiver opto side is more common. The transmitter LED will be lit (and visible with the naked eye), while the receiver LED will not be lit.

• Sega/White Star Transmitter, Receiver, and Opto 'Long Hop' Board Pair, as found on Starship Troopers. Note that the ultra-red LED on the transmitter side is not lit until power is applied. The receiver side never illuminates.
• 

  The opto transmitter side.
• 

  The opto receiver side, with accompanying electronics. The red colored LED is a diagnostic aid to indicate a detected switch closure.
• 

  An opto pair. The pair must be in perfect alignment or the opto switch will operate unreliably.

This particular system is different than how Bally/Williams WPC CPUs handle opto pairs. WPC games employ infrared opto pairs.

Theory of operation for these circuits is described in the manuals. Sega/Stern have included some excellent, detailed, technical documentation within their manuals. Stern Service Bulletin #67 includes detailed theory of operation of these opto switches.

4.8.4.1 Two Balls Served, Continuous Balls Served to the Shooter Lane, or the Trough VUK Fires Repeatedly

Common problems with some White Star games are:

1. Two balls are served to the shooter lane.
2. Balls are continuously served to the shooter lane.
3. The trough VUK repeatedly fires, when there are no balls above the VUK trough plunger.

The first test is to determine if the balls are sliding down the trough. If they are repeatedly magnetized they tend to stick to the side of the trough and not slide down. Replacing the balls with will fix this. There are balls available which are low magnetic, but you can also line the floor of the ball trough with a thin sheet of PETG breaking the magnetic contact.

Depending on the exact circumstances, and the era of the White Star game, one or two of these symptoms can occur. The crux of the problem is either one or both of the two trough LED opto transmitters or receivers is failing. Rarely do the components on the trough opto boards other than the red LEDs actually fail though. And in most cases, it is not the LED itself, although, there are some instances when the problem is a failing LED. It is noteworthy that the issue is more common to the opto receiver side more so than the opto transmitter side.

A Stern Monopoly in "Clear Ball Trough" Mode Identifying One Ball in the Trough

A Stern Monopoly in "Clear Ball Trough" Mode Identifying One Ball in the Trough and One Ball Stacked

If the any of the previously mentioned symptoms are evident, the best approach is to first put the game into "Clear Ball Trough" mode via the Portal buttons on the coin door. Consult the manual documentation for details how to enter this test. Once all of the balls are removed from the trough, no balls should display on the DMD display. If either of the two images shown to the left are displayed on the DMD, and the ball trough no longer has any balls present, there is a problem with either the transmitting or receiving opto boards.<br=clear all>

The first thing is to observe the two red opto LEDs on the transmitting board. This board is located on the backside of the ball trough. If the transmitting LEDs are lit, there probably is not a problem with them. However, this is not always necessarily the case. Secondly, inspect the opto LEDs on the receiving board. To do this, the game will have to be turned off, and the receiving opto board removed via the three screws which attach it to the trough. Look for hairline cracks around the .100" angled header pin solder joints. Likewise, inspect the solder joints around the opto LEDs. If cracks are evident in any of these locations, remove, and apply fresh solder to these connections. Reinstall the receiver board, and observe the results. Conversely, if hairline cracks are not found, or if the same symptoms are still evident after reflowing solder joints, replacement of the LED optos is suggested.

White Star games initially used MT5000UR ultra-bright red LEDs, while later games used TLRH180P ultra-bright red LEDs for both the transmitter and receiver boards. Both styles of LEDs have become expensive and fairly difficult to source. Great Plains Electronics offers the MV8114 as viable replacement for either original, factory LED. The only caveat is that both the transmitter and receiver opto LEDs must be replaced in pairs to ensure proper function.

When replacing the opto LEDs, make certain the base of the LED is placed squarely on the PCB. An LED which is poorly installed may result in sporadic switch closures.

Lock Ball Assembly (from Jurassic Park - used for illustrative purposes)

Please note that a "Lock Ball Assembly" was employed in the trough design on early White Star games. The lock ball assembly is a unit, which was carried over from later Data East and early Sega games. It consists of a coil mounted horizontally above the left side of the trough VUK. This assembly was later removed presumably either due to cost factors, or so this coil assignment could be used as a playfield coil feature elsewhere instead.

If a lock ball assembly is used in the game, and trough opto failure has occured, the symptom will more than likely be that the trough VUK will fire repeatedly without a ball present above the VUK plunger.

4.8.4.2 Phantom Opto Switch Closures

Similar to the discussion above, "Two Balls Served, Continuous Balls Served to the Shooter Lane, or the Trough VUK Fires Repeatedly", other opto pairs in the game may fail over time and become unreliable. Usually this manifests as phantom switch closures. Games like Starship Troopers that use an opto pair for both the left and right orbit shots may score the orbit shot randomly or even continuously. If the "long hop" opto assembly is removed, you may find that the opto pair registers correctly if the optos are aimed with great accuracy at one another. However, the slightest misalignment (like flexing the PCB) will cause the opto pair to register a "hit".

The correction for this is again to replace the original MT5000UR or TLRH180P ultra-bright red LEDs, with MV8114 LEDs from Great Plains Electronics.

4.9 Display problems

A few of the many possible display problem root causes are...

• Failure in high voltage power supply
• Failed Dot Matrix Display
• Failed ribbon cable
• Failed power connector on display driver board
• Failed 74LS374 buffer IC on the MPU board

Each possible root cause is discussed below.

4.9.1 Measuring High Voltages for the DMD Display

Testing DMD Controller voltages at the display. Note: This picture is a Williams game, but the concept is the same.

To test for the voltages necessary for the DMD to operate, first open the backbox, and rest the DMD panel face down on the glass. Or, if the game has a "Showcase" style head (i.e. StarShip Troopers), swing the panel open to gain access to the HV connection to the DMD. The simplest place to test the voltages is at the display end of the power wire bundle as shown in the picture. SAFETY WARNING: Your left hand should be in your pocket, to avoid a potentially serious shock. We are working with high voltages, so be VERY CAREFUL. Keeping one hand in your pocket is a good practice to follow as it eliminates the easy electric current pathway across your heart. A DMD controller provides enough electric current to kill you!

The dot matrix display needs three voltages to operate, nominally -112, -100, and 62 VDC. All three voltages must be present for the display to operate correctly.

Place the black lead of the DMM under the ground braid in the game's head or use clip leads to secure the black lead to ground. Remember, you are working one-handed, so this becomes a necessity. Place the red lead of your DMM on each of the pins supplying power as shown in the picture. As long as there is a 12VDC difference between the Anode and 12VDC offset voltages, your DMD controller is providing correct voltage levels.

Function	Pin	Spec	nominal voltage	Notes
Anode	1	-120V	-112V
12VDC Offset	2	-100V	-100V	Must be 12V less than anode voltage. Will be about 20V less than anode voltage if display not connected
Key	3
Ground	4
Ground	5
5VDC	6	5V	5V
12VDC	7	12V	12V
Cathode	8	60V	62V

If the high voltages are off by more than a few volts, turn the game off, disconnect the power connector at the HV Power Supply, and test the voltages at the HV Power Supply male header pins. If the voltages return to nominal values, the zener diodes on the power supply are probably OK but the resistors and transistors around them are suspect. As shown in the table above, if HV power is not connected to the display, the nominal 12VDC offset will be about 20VDC.

Note that a failing or failed DMD display glass can drag these voltages down. Before you begin any work on the HV Power Supply board, perform the above test with the DMD display power disconnected. If the voltages return to nominal values, it's possible that your display is the culprit. Test the high voltages with a known good display to isolate the fault to either the HV Power Supply or the display panel.

Should your HV Power Supply fail, the circuit is almost identical to the circuit used by Williams in their WPC DMD Controller. Similar parts are required to rehab the Sega/Stern HV Power Supply.

As of April 2014, schematics for the HV Power Supply are no longer available on Stern's site.

4.9.2 Typical Failures of High Voltage Power Supply

A Stern 520-5138-00 Dot Matrix High Voltage Power Supply

The nominal output voltages of this board are shown below.
The voltages measured between game ground and each pin of CN2 (right side as installed in the game).

• -110V on CN2-1
• -98V on CN2-2
• +5V on CN2-6 (piped directly through from CN1-7. This was a method of minimizing connector complexity)
• +12V on CN2-7 -- 20V is supplied at CN1-6 and is regulated down to +12V by VR1, a 7812.
  Note: the voltage at this pin will be missing if the board isn't stuffed with the 7812. It is not known when Stern/Sega stopped stuffing this location.
• +68V on CN2-8

The fuse on the board is a 3/4 ASB fuse.

Common failures:

• Fuse blowing
• -110V missing or too low
• -98V missing or too low
• +68V missing or too low

Solutions:

A blown fuse is indicative of a failed rectifier diode at D1 or D2 (1N4004. Anything from a 1N4005 to a 1N4007 can be used too). When D1 fails, Q1 and Q3 often fail along with it. When D2 fails, Q2, Q4, and Q5 often fail along with it.

• -120V problems are often caused by a failed transistor at Q2 (MPS-A42) and/or Q4 (MJE15031). When one of the transistors fails, often the other one fails too. Replace zener diode D4 (3.9V, 1N4730), D6 (100V, 1N4764A) and D7 (13V, 1N4743A) as well.

• -100V problems are often caused by a failed transistor at Q5 (MJE15030) or zener diode D6 (100V, 1N4764A). It's best to replace both. The 2k/5W resistor on R8 might also be the problem source, mostly due to failed solder contacts.

• +60V problems are often caused by a failed transistor at Q3 (MJE15030) and/or Q1 (MPS-A92) When one of the transistors fails, often the other one fails too. Replace zener diode D3 (3.9V, 1N4730) and D5 (68V, 1N4760A) as well.

As of 2014, schematics for the display power supply board are no longer available on Stern's website.

The Data East display power section of the 520-5047-01 and 520-5047-02 boards is an almost 1:1 copy compared to the one on this board.

4.9.3 Failed Dot Matrix Display

A failed Dot Matrix Display will obviously cause display problems. Once you've tested for proper voltages, swap in a known good display, and swap the suspect display into a known working game. If the suspect display doesn't work in the known working game, it's time for a new DMD display.

4.9.4 Corrosion Under Header Pins

Corrosion under the dual row headers of a DMD Controller board

Corrosion from leaky batteries, being the bane of pinball PCBs, finds it's way to many places that might not normally be suspected. Sometimes, alkaline corrosion invades the header pin carriers, as shown in the picture at left. In this case, enough of the signals between the MPU board and the DMD controller board were corrupted to prevent proper rendering on the DMD display. Cleanup via typical methods restored this board to proper operation.

4.9.5 Failed ribbon cable

The ribbon cables on Sega/Stern White Star games connecting the CPU to the display controller are particularly long and twisted. There is ample opportunity for the ribbon cable to be damaged.

You should reseat the ribbon at both ends several times to knock off accumulated corrosion. This might temporarily solve the issue.

A simple test of the ribbon cable integrity would be to reverse it, causing the conductors used within the ribbon to be "swapped". Remove the ribbon from both ends, and instead of seating it with the red stripe at pin 1, seat the ribbon with the red stripe at the opposite end from pin 1. Make sure you reverse both ends of the ribbon. If the results witnessed on the display are then different, there is a good chance you need a new ribbon cable.

Gently bending and twisting the ribbon cables may help uncover intermittent connections and a compromised cable. Don't bend/twist too vigorously as this in itself may compromise the cable.

4.9.6 Failed power connector on display driver board (CN2)

Sega/Stern (and even Data East) chose a very odd power connection to the display controller. CN2 is a .156, 3 pin connector, which is the +5v and ground connection for the display controller. It can be inadvertently and easily reversed. The solder joints for these header pins fracture often, or the pins in the female housing lose their grip (and hence connection). It's good practice to repin both sides of this connection, using crimp-on TriFurcon pins in the female housing.

4.9.7 Failed 74HCT273 buffer IC on the MPU board

One manifestation of a failed U201 or an open between U201 and the plasma display connector. Image courtesy of Chris Hibler.

The 74HCT273 at U201 on the CPU board buffers the output to the display controller via the ribbon cable. Should this IC fail, incorrect or no data will reach the display controller. Most of the time, this results in only the "splash screen" appearing which is displayed by the DMD controller on it's own, or in rare instances, the odd displays shown at left. First an over-scaled "Location ID:" image is displayed then "Bad Schematic Table Index!". This particular failure was caused by a corroded trace between U201 pin 2 and the via just northwest of pin 2. However, the same message may be displayed any time that the data path between U201 and the DMD Controller fails (i.e. poor header connections, failed ribbon cable).

U201 can be tested with a DMM following the process described here.

4.9.8 Display Issues Caused by Mismatched Game and Display ROMs

This message appears when the display controller software can't correctly decode display "calls" made by the MPU. Note that hexadecimal numbers other than "F7" may be displayed. Image courtesy of Chris Hibler.

Over the course of a game's manufacturing run, and even post production, game and display software was updated. It is important to ensure that the game ROM version is compatible with the display ROM version. Newer versions of the game code may request a display routine from the display controller that the display controller can not make sense of. When this happens, the display controller will display "Effect NN", where "NN" is the hexadecimal number of the display routine received.

Note that a less likely failure of IC U17 (a 74LS245) on the display controller board, which buffers inputs from the MPU, can cause the same or similar behavior. Corrosion on the ribbon header pins (both ends) or a faulty ribbon cable are other possible reasons for this message.

4.10 Sound problems

Before going any further, check all of the basics in the Basic Sound Troubleshooting section.

Basic operation (simplified) of the White Star CPU/Sound board audio part: The main CPU (U200) makes sound calls to the sound CPU (U6) which instructs the BSMT to play a certain sound. The BSMT fetches the required data (sound sample) from the ROM memory (U17-U21-U36-U37) and feeds the 16-bits wide parallel data to a conversion circuit which makes it serial data (shift register, U23-U24). The serial data is fed into a DAC (U26, Digital to Analogue Converter) that outputs to a pre-amplifier (U30, OPAMP). The output from the pre-amplifier is fed into the TDA2030A amplifiers at U101/U102 and sometimes U100 (some boards have 3 amplifiers, one for each speaker) and from there the audio signal goes to the speakers.

The White Star (modified) version of the MPU board replaces almost all of the prior version's sound circuitry with 3 surface mount ICs.

• Xilinx CPLD at U414 (XC9572XL)
• Atmel RISC processor at U402 (AT91R40008)
• Atmel flash memory at U8 (AT49BV162AT)

The Xilinx CPDL (Complex Programmable Logic Device) is programmed by Stern (or it's board manufacturer) via the JTAG connector on the board.

The Atmel RISC processor (Reduced Instruction Set Computing) replaces the 68B09EP of the prior board.

The Flash Memory contains the Sound Operating System. The correct version of the sound operating system my be "flashed" into the memory. LED1 on this board will blink anywhere from 5 to 8 times to indicate the sound operating system version contained in the flash memory at U8. Stern has provided a good explanation within Service Bulletin #157. This service bulletin also describes the process to update flash memory to a newer version of the sound operating system (which is not recommended unless you encounter sound issues after moving an MPU from one game to another). In April, 2016, "Chas" at Stern Pinball noted that sound operating system 8 (O/S 8) is backwards compatible with all WhiteStar (modified) games. It may even work with sound ROMs predating the WhiteStar (modified) board. However, there are known compatibility issues with early WhiteStar Sega games like StarShip Troopers.

The sound section of both board versions is quite reliable. The failure occuring most is a failed TDA2030A amplifier. If a sound problem can be isolated to one of the three fine pitch surface mount ICs on the WhiteStar (modified) board, the only option known as of April, 2016 is to return the board to Stern for repair. Stern does carry the parts necessary, but removal and replacement of such fine pitch parts is above the capability of most pinball hobbyists.

Sound missing from one of the speakers: Most likely a failed TDA2030A amplifier.

No sound at all: Check the voltages on the board. On the DC input connector CN2 there should be approx. +12V DC between pin 2 (ground) and pin 6 (+12V) and -12V between pin 2 (ground) and pin 3 (-12V), these voltages come from the Power Supply Board. On the board itself are two voltage regulators to convert the +12V and -12V to +5V and -5V for the pre-amplifiers, check these voltages by measuring between the ground (GND test point) and U30 pin 8 for +5V, U30 pin 4 for -5V.

Other causes of no sound at all are the U7 game ROM being corrupted or improperly seated. You can re-seat this ROM and it often solves the issue. If any of the voice ROMs (U17, U21, U36, U37) are corrupted or improperly seated, the sound will still work but will be corrupted or have intermittent issues such as scratchy sounds or improper sounds at specific times.

An incorrectly jumpered or installed RAM chip can also cause no sound. JMP5 should be installed and JMP4 removed when a 6116 is used. JMP4 should be installed and JMP5 removed when a 6264 is used. JMP4 and JMP5 are just above U7 (Sound ROM code) and U4 (RAM socket).

Finally, if you have no sound at all, you might also want to check that you have the correct sound board. Some operators may have swapped the 520-5300 with a 520-5136 board for the older series White Star games. The 5300 board will work in all White Star systems. The 5136 will work but will present no sound at all in all games newer than lord of the rings (2003).

Samples playing randomly This happens on Whitestar (modified) CPU boards that emulate the BSMT2000 (which Stern could not source cheaply enough at the time). Reflashing the sound operating system firmware per Stern Service Bulletin 157 is recommended.

Sound is distorted If the sound is distorted, the most likely problem is the TDA1543 (U26) chip. This sounds as if the volume is very over modulated and are “clipped”. Be sure to use the TDA1543 version and NOT the TDA1543A. The TDA1543A uses another protocol (Japanese). If the sound is distorted at master volume levels below 4, see the "Loud Hum and/or Distortion at Low Sound Levels" section below.

Sound cuts out at the exact same point in a clip every time If the sound consistently cuts out, or volume is lowered, or the clip is muffled, every time at the same point in the clip, there is a possibility that the 6116 RAM at U4 has failed.

Sound volume is extremely loud Possible TDA1543 (U26) chip failure or counterfeit chip.

Loud hum If there is a loud hum coming out of the speakers as soon as the game is turned on, make certain that the +12V is present.

Loud hum/burnt speakers/12V DC coming out of the speaker connector If you have a machine with all speakers dead, check for voltage on the speaker wires. If the -12V wire from the driver board to the CPU board is disconnected, it can cause 12V DC to be sent down the speaker line, causing a loud hum and fairly quickly destroying any speakers connected!

Music Sounds Like Alvin and the Chipmunks

Schematic diagram showing the physical location of jumper W6 between the sound section 68B09E and the BMST2000, enabling the use of 8Mb EPROMs

The original WhiteStar schematics mention jumper W6, which "must be installed for games that use 8Mb Sound EPROMs (U17 U21 U36 U37). For games which use 4Mb Sound EPROMs this jumper is not installed but 4Mb EPROMS will work on boards with W6 installed". This boils down to...install jumper W6. The manifestation of using 8Mb ROM images on a board with W6 not installed is sound that appears to be "Alvin and the Chipmunks"-like.

A YouTube video showing an incorrectly jumpered WhiteStar MPU with Harley Davidson ROMs can be found here.

Note that the WhiteStar (modified) MPU board does not provide jumper W6, nor is it needed. The WhiteStar modified board works with both 4Mb and 8Mb sound ROMs, as if W6 was installed.

4.10.1 Sound Section Jumper Settings

JMP5 should be installed and JMP4 removed when a 6116 is used. JMP4 should be installed and JMP5 removed when a 6264 is used. JMP4 and JMP5 are just above U7 (Sound ROM code) and U4 (RAM socket). An incorrectly jumpered or installed RAM chip can result in no sound at all.

Important: Under normal operation...

• JMP2 should be on the right pair of pins
• JMP3 should be on the top pair of pins, and
• JMP4 should be on the bottom pair of pins.

Placing the jumpers in any other position will cause sound to be missing. These jumpers are provided to update the sound operating system within the flash memory at U8.

4.10.2 Loud Hum and/or Distortion at Low Sound Levels

Trim POTs connected across R106 and R110. This mod was offered by an aftermarket seller. Image courtesy of Jack Jarrett

Trim POTs connected across R106 and R110 in a rather kludgy way...but it works. Image courtesy of PinSider Capt_Pinball

Stern released Service Bulletin 133, which discusses a fix for a loud hum and / or distortion when the master volume is set below level 4. The machines that are specifically effected are Monopoly and Austin Powers. However, it has been over 10 years since these games have been released, and boards get transplanted to other games over time. So, it is important to check the two resistors (R106 and R110) in the amplification section of the CPU / sound board to see if the correct values are installed, regardless of the game. The correct resistor values should be 33K ohms 1/4 watt for R110 (ORG-ORG-ORG-GOLD), and 10K ohms 1/4 watt for R106 (BRN-BLK-ORG-GOLD). Consult the service bulletin for more details, and the layout of the amp section. The R106 and R110 control the analog volume. When these resistors are broken (open) the analog volume goes to 100% and can't be controlled via the buttons (digital volume) on the coin door. The higher the resistor value the higher the volume.

Another method of accomplishing the same thing, without removing and replacing a resistor, is to add a resistor to the back of the board, across R106 and R110, creating the equivalent resistance as mentioned above. A 4.7K resistor seems to be a sweet spot.

Yet another method used is to clip a 10K trim pot across each resistor as shown in these images. The trim pot can be adjusted to a good compromise between hum and volume.

4.10.3 Low Background Sound Volume

Background sound level setting on Roller Coaster Tycoon

Some games, at least Viper Night Drivin' and Roller Coaster Tycoon, provide a means of adjusting the level of the background sound. The sound can be adjusted from level 1 (loudest) to level 15 (lowest).

4.10.4 Adding a Second Backbox Speaker to some Sega Games

Adding a second backbox speaker may improve the sound in games shipped with a single backbox speaker. Buy the speakers in pairs. This image shows the new speaker installed onto the left speaker bracket, and the right speaker position "blank" from a StarShip Troopers game. The "blank" must be modified so that the new speaker can be attached.

A chap named Martin Riley submitted a nice article to the PinballNews explaining how to add a second backbox speaker to games like StarShip Troopers.

From the article...
"Mid-life Sega games from Twister through to and including Viper Night Drivin' only had the left hand backbox speaker plus the cabinet speaker and only two of the 3 audio amps are populated on the CPU board.

Late Sega games from Lost in Space onwards have two backbox speakers but they are effectively mono as they are both connected to the left channel audio amp on the CPU board (the right channel amp being unpopulated)."

The complete article can be found at http://www.pinballnews.com/learn/speaker/index.html

The article lists all components required to add the 3rd speaker channel. Some (maybe all) two channel boards are not stuffed with C122, which is not listed in article parts list. You'll need one more 0.1µF axial ceramic capacitor for C122.

Adding a 3rd speaker to StarShip Troopers improved the sound quite a bit. The sound is richer and cleaner. The emphasis on some sounds is more pronounced and easier to hear. Edit: months later, I was testing a board in my StarShip Troopers that didn't have the 3rd amp. Forgetting that I had added the 3rd amp to my MPU, I initially thought something was wrong with the sound that the MPU under test was producing. It was merely the difference between a single and dual backbox speakers.

Mounting the additional speaker to the "blank plate" (as found in StarShip Troopers) required drilling the rivets that attach the speaker grill, and replacing them with low profile head screws. This is only necessary for Sega "Showcase" backbox games. For games like Independence Day, mounting the additional speaker should be trivial.

4.10.5 Is a Second Backbox Speaker Really Worth It?

While some have reported good results adding a third amp and second backbox speaker to these early Sega games, there are questions about the efficacy of such a solution. While better speakers themselves will improve the sound, the game-specific sound samples in the ROMs are mastered as mono sound clips. There are very few sounds in the game that are true stereo sound and these are usually simple sound effects which were used in multiple games.

You can listen to and judge for yourself by running the Bridge front end and the M1 MAME music player. Bridge and links to the M1 player can be found here.

The mono mastering continued through the White Star II and SAM systems. As David Thiel, who has created music and sound for numerous Data East and Stern pinball machines, states on this page of his website, he "composed, arranged, performed, mixed and mastered (all the music for Family Guy) for 24 kHz 16 bit Mono."

5 Game Specific Problems and Fixes

Lord Of The Rings

If your Balrog stops registering hits, a little checking will sort it out. The mini microswitch with a roller arm used to register hits may need adjusting. Known past problems are a flaky or bad switch, and/or wiring. The Balrog moves in such a way that wires can break inside where you can't readily see them. It's easy to check the switch and wires with a meter. The wires push into an IDC connector nearby; be sure they are pushed in tight.

There isn't any adjustment for the micro switch, except for bending the blade with the mini-roller; try bending the blade outwards slightly.  According to the manual, sw #28 stays normally closed and then a Balrog hit opens the switch. This is not the normal target arrangement. So using the Balrog test mode in diagnostics, the switch should stay solidly closed, and then only open when Balrog is tilted forward.

The Shire and Gimli VUK switches.
These get hammered on LOTR and will eventually need replacing, There isn't any easy solution other than replacing the flaky switch. The tiny black screws on the Stern switches are 3/16".  You will need a nut driver that size if you want to attempt replacement of the switches.

The Shire (and Gimli) VUK switch, are fork blade microswitches. Stern part number 180-5116-01.
Spec on side of switch says Cherry .1A 125v E63. Try parts4pinballs.com or your favorite Stern parts supplier.

The Shire VUK.
It's easiest to remove the entire Bottom Left VUK coil and bracket from under the PF (playfield) to change the fork switch.
Take out the two largest bolts first, using a cross-head (phillips) screwdriver on the top and 11/32" nut runner (or socket) on the bottom.
Remove the metal VUK arch from the top of the playfield.
Remove the two lower wood screws (1/4" drive) leaving the one on the corner near the T-nut.
Supporting the Coil Bracket, remove the last 1/4" screw.
Unclip the connector from the coil and VUK.  Remove coil and bracket assembly.

Microswitch small screws are 3/16".  As stated before, to remove the switch correctly you need a nut driver (or socket) that size. You should be able to slide the heatshrink tubing all the way up the wires, solder on the new switch and then slide them back down in place. If not, use new heatshrink tubing. Assembly is above steps in reverse. Test in switch tests before replacing assembly and then by throwing a ball in the Shire in attract mode one the assembly has been correctly replaced.

Ripley's Believe it or Not!
Not certain what code versions apply with this. When the coin door is open, and the first ball drains, the game will not kick the next ball over to the shooter lane. This is even with the solenoid voltage interlock switch pulled out. It is presumed that the memory lock switch has to be closed for the game to function properly.

6 Repair Logs

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

• Striker Xtreme:
  • Problem: Display is blacked out or only coming on for a few seconds before blanking out.
  • Cause: Resistors R6 and/or R7 on the display power supply board have gone bad (open or increased in value).
  • Solution: Replace resistors R6 AND R7 on the display power supply board. Both are 330k @ 1/2 watt
• Simpsons Pinball Party:
  • Problem: When last ball drains, game still thinks its in play. Ball search or power cycle will eventually find it.
  • Cause: Cracked solder joint on trough board.
  • Solution: Pull both trough boards. A 1/4" wrench can be useful in getting the back board off. In this case, the obviously cracked solder connection was on the receiver board in front. But while the board is out, the optos were cleaned with Windex.
  • This can be verified by waiting for the problem to show up, then go into switch test. Previously, switch #5 (leftmost) had been suspected and had been re-aligned, but that was probably not necessary.
• Starship Troopers, but applies to all:
  • Problem: No display, no sound, upper amplifier went up in smoke.
  • Cause: Connector CN4 on MPU / Sound board and CN1 of display power supply board were swapped. Both CN4 and CN1 are keyed the same. In haste, these connectors can be mixed up. Keep in mind that the speaker output is a very light gauge wire versus the heavier gauge for the display power supply input.
  • Solution: This was not as bad as expected. After studying the schematics, only +20V (CN1 pin 6 - pin 4 is GND on both connections) can potentially backfeed through the sound circuitry. Amplifier U101 failing spectacularly possibly prevented additional damage. Amps U101 and U102 were replaced (U100 was not stuffed on this board). After replacing the amps, sound came back, but was very faint. Further investigation revealed 3 100Kohm resistors had failed, R21, R22, and R24. Also, U26, the TDA1543 DAC, had also failed.