Midway Rotation VIII

Note: This page is a work in progress. Please help get it to a completed state by adding any useful information to it.

This section is dedicated to the repair, preservation, restoration and maintenance of Midway's Rotation VIII.

1 Midway's Rotation VIII

1.1 Introduction

This section is dedicated to the repair, preservation, restoration and maintenance of Midway's ROTATION VIII. Intended to be a repository of information as well as a resource to owners, it is hoped this will be of value to those interested in making one of history’s most unique pinball games operational and fun to play.

The information contained here was originally collected on a website run by Tim Polzin. He says "The content was collected in many ways. Much of it has been collected by word of mouth or gleaned from forums or other sources. Some information was gained from his first hand experience in repairing a number of these machines."

He has shut down that website and the information from it has been migrated here.

1.2 General Information

1.2.1 Rotation VIII Value:

Value and sale price of a Rotation VIII has come up on a number of forums. The author knows of a few working Rotation VIIIs that have been sold. The first was late fall of 2012 on E-bay at just under $2000 US near Chicago. A second was during the summer of 2013 in the Seattle area at $2500. Several others in the same area sold during 2014 in the range of $1500 $1700. All but one of these machines used my replacement daughter board. Most of the non-working but complete machines I know of seem to sell between $500 to $800. European prices for non-working machines seem to be considerably more. Condition and completeness are always deciding factors.

1.2.2 Original US Patent for Rotation VIII

Click here.

1.2.3 History

Designed in the infancy of solid state pinball machines, Rotation VIII is unique in that the Game is a cocktail table game with an automatic rotating playfield. Up to four players can sit around the table with the playfield turning sequentially to their position to play.

Known serial numbers range from 0216 to 0432 in North America and serial numbers from 0635 to 0737 appearing in Europe. 0696 appears in South America. It now seems likely that the lower number serials from 0200 to 0499 may be North American and the range from 0600 to 0799 represent machines exported out of North America. The difference noted so far between the North American machines and the ROW (Rest of World) machines seems to be in the power transformers with ROW machines having windings supporting 220 volts.

To date, there is no definitive number of how many ROTATION VIII's were produced. Guesses have placed the production run in the range of several hundred to 1000. The serial numbers found so far suggests that the number may be in approximately 400. Rumours also stated that some of the games may have been taken back by the manufacturer due to reliability problems. Production commenced in 1978, but we don’t know how long production continued. Many of the surviving games are now sitting in garages, basements and warehouses because they don't work. Only a handful are now operational.

1.2.3.1 Where is my Serial Number? Serial

The serial number is usually printed on a small light cardboard tag stapled to the inside of one of the coin doors. On some machines, the serial number may also be stamped into the horizontal shelf near the coin door. I have never found any serial number information on the playfield or anywhere else on the cabinet or frame.

1.3 Technical Information

1.3.1 Hardware Addressing

Port reads:

$21: matrix switches $22: keypad inputs

Port writes:

$00 .. $02: display data $03: lamp data $04 .. $07: solenoids $10 .. $17: sounds $23: switch / lamp / display column

Memory:

$0000 .. $17ff: ROM $1800 .. code extension $c000 .. $c1ff: RAM $e000 .. $e01f: NVRAM

Display data:

$00: Digit data A and B (bcd-coded, high and low nibbles) $01: Digit data C and D -"- $02: Digit data E and F -"-

$23: remember the strobe column value (0 .. 6) used, then: data A is written to position (strobe) of the 1up panel, data B is written to position (strobe) of the 2up panel, data C is written to position (strobe) of the 3up panel, data D is written to position (strobe) of the 4up panel, data F is written to position (strobe) of the lower panel.

You might wonder about data E? In fact it's not output, and contains some data the game needs for calculations (like the match digits).

Thanks to Gerrit Volkenborn for the addressing information.

1.4 Problems and Solutions

1.4.1 Summary of Issues

One of the primary causes of non-operating Rotation VIIIs is the original Characterization Board (daughter board). This board was subject to considerable corrosion from the memory battery. Further problems resulted from dirty or worn connector contacts also on the characterization board. As well, the characterization board used tri-power eproms which were touchy and had sockets that were prone to failure. In severe cases, the logic board has also been corroded from the Characterization board. Replacement of the Characterization board, updating the reset circuit and re-pinning connectors, especially the power supply lines on the logic board may bring a machine back to life. Unfortunately, many logic boards also have failed components and corroded traces.

As is the case with 40+ year old technology, many if not most of the active components on these boards are obsolete and may at times be difficult to find exact replacements. To date, some "drop in" substitutions have been found for several difficult to find components. This information is contained in the Technical page.

The rotating circuitry and equipment is robust but needs to be calibrated to work properly. The playfield will not rotate if a ball is not in the outhole. There are also several mechanical issues related to moving parts that will cause instability. Playfield parts are mostly stock and can be purchased from suppliers.

1.4.2 Getting your board to boot

The following summary is intended to be a troubleshooting guide for the electronics in Midway’s Rotation VIII.

As you have probably already noticed, the circuit boards used in this machine are not used in any other machine. As well, there are virtually no self-diagnostics available to help resolve machine electronic faults.

This guide assumes you have a reasonable understanding of electronic repair, have good soldering skills, can remove and replace components including DIP components, have basic electronic testing equipment and can read a schematic. An oscilloscope would be helpful, but a reasonable logic probe may still work.

1.4.2.1 Main Board

A little background on the main control board first. The main board is based on a Z-80 processor. The Z-80 processor uses separate pins for the address buss (16 lines) and the data buss (8 lines). This address and data buss all pass through to the daughter board which holds the 2716 style eproms and the NV memory. The daughter board also generates the RD (data read) signal used by a number of components on the main board.

Now, to add another level of complexity, the designer, I’m sure in his or her infinite wisdom, chose to use several additional components originally designed for the 8080 and Z-85 processors, and here is where the rub comes in. These two possessors used multiplexed data and address lines, in other works, the lower 8 address bits share the same lines as the 8 data bits. That means in addition to the 16 address lines and 8 data lines on the Z-80, you have another 8 other lines that can have data or address.

So, unless everything on the two separate address and data busses including the daughter board work, the main board will not boot. The main board is smart enough to send itself a reset pulse if it does not boot within a certain amount of time. Monitoring that reset line will give the trouble shooter an indication that the board is trying to boot, but is not successful.

OK, have I scared you yet?

Having repaired a number of main boards, there are a couple of common faults that may be encountered. Before starting any diagnosis, I am assuming that all voltages have been checked, connectors are solid and all boards are in place. I’m also assuming the main board does not show corrosion damage.

Situation 1) Turn on machine and nothing happens, only GI lights come on, no table movement and no controlled lamps are lit.

• First check pin 2 of J8. A solid high indicates the reset is not active and is normal. A machine booting normally will have a steady voltage of around 4.8 volts on this pin.481x446px This condition is highly unlikely in a non-booting board and, if present, is probably caused by a faulty reset circuit or loss of the clock signal. A solid low signal is also a reset fault and will not allow the processor to operate. Pin 10 of A1 should have clock. If it doesn’t, start trouble shooting the clock circuit including B1. If the clock signal is present at A1, check the reset circuit in chips A1, A2 and A3 for faults.
• If you have a continuous series of pulses on J8, pin2, you could have a fault in the daughter board, and/or either of the buss drivers F3 and H3.
• Examine the daughter board carefully for any signs of corrosion. I have had nothing but grief with this board. Even the slightest bit of corrosion seems to cause the daughter board to fail. If you have access to a known good daughter board, or have access to my replacement daughter board, try it to see if the main board will boot.
• Replace the buss drivers at F3 and H3 (8304 and 81LS95). Failure of these chips seems quite common and probably results from a failure on the daughter board. Even with an oscilloscope, I have never had much luck in determining which of the two chips is bad, so as a matter of practice, I replace both.  
• With a good chip in F3 and H3, hopefully you will now be in situation 2 below or better yet the main board boots.

Situation 2) Power on and random switched lights, solenoids and/or table movements and no start-up tune.

• In most cases, this will happen with a bad daughter board and/or faulty solder joins of the connectors to the daughter board. Have I mentioned how much I hate the original daughter board?
• Reflow solder on daughter board and main board connectors.
• Check presence of all three voltages on the daughter board.
• Slightly wiggle and twist the daughter board as the machine is turned on, anything different?
• Reseat the three TMS2716 chips.
• Still nothing? You may be in a situation needing further diagnosis. Time for us to talk.

Situation 3) Power on, board boots and plays a tune, but as it boots, you get random solenoid firing.

• Upgrade the reset circuit as I outlined here.

1.4.2.2 Motor Driver Board

This board will commonly fail, usually when anything gets in the way of the rotating table. Wires, cables, almost anything can cause it and the fuse rarely protects it.
The good news is that this board is easily repaired. Usually what happens when one of the driver transistors fail is the table will turn in one direction only, cannot correct for an overshoot and will not start a game. Almost always, the failure will be one of the Darlington pair transistors, TIP 110 or TIP 115. All four are easy to replace and are commonly available.

Occasionally, if the Darlington transistor drivers are good and the table does not move, the problem may be in the large 3.3 ohm resistor. Check it for continuity. Rarely the LM3900 goes bad, but it can happen.

1.4.3 Reset Circuit

The Midway Service Note #2 identifies a problem with the reset circuitry. Resets on the pin2logic board are done two ways. One is a simple power-on-reset used to ensure the game starts properly. The second is a internally triggered reset to do an automatic reset on certain conditions (like pushing the "Game" button on the keyboard). The revised reset circuit Midway describes is on the power-on-resetreset.

I feel a better way of doing the power-on-reset is to clip and remove pin 2 of chip A2-3 (LM339) as shown above and solder in a DS1811 or MCP130 to the underside of the logic board to control reset as shown to the left. Lead 1 to pads 1-2 of A2, lead 2 to the +5 volt rail, and lead 3 to pad 12 (ground) of A2. This results in a clean 150 ms reset pulse on power-up with no coil chatter as often found on the original reset circuit.

1.4.4 Power Supply

It is very good practice to replace the 11,000 ufd filter capacitor on the power supply board. I was getting intermittent failures of supply voltage to a game that was traced back to this capacitor. Value is not critical as long as it is adequate (8000 ufd or more with at least a 16 volt rating).

1.4.5 Re-Pinning

It is a very good idea to re-pin at least the contacts on J0 and J1 which feed power to the logic board. For those connectors, I used the trifulcon series contacts to ensure a good connection.

1.4.6 Errata

I found a few errors in both the E627 and G627 schematics.

E3 (74ls174) - pin 13 is connected to AD5, not AD4, pin 11 is connected to AD4, not AD3

Reset line from pin 17 of K3-4 (0066-177XX) is connected to pin 15 of all 74ls259s, not just D5.

E3 pin 2 is not connected to pin 1 of F1 (74ls10)

F1 pin 1 is connected to pin 2 of B3 (74ls10) and pin 1 of C1 (74ls00)

If you find any more, let me know and I'll document them.

1.4.7 Substitute Parts

The NPN Darlington pair transistor (RCA D42D1) used in the Rows matrix driver appears to be not available anywhere. I found a TIP102 is a drop in replacement. The PNP Darlington pair transistor (RCA D45E1) used in the Column matrix is also not available and can be replaced with a TIP107.

The audio amp chip used on the G series logic board (LM377) is difficult to source. The later version LM1877 chip is pin-for-pin compatible and works well. Don't forget to replace the original heat sink or epoxy a replacement to it's back.

The audio generator (0066-117XX) is a rare custom chip. In addition to Rotation VIII, it was also used in Gorf, Wizard of Wor, Space Zap, Extra Bases, Sea Wolf II and perhaps others. It was also used in the Astrocade game console.

The pop bumper tops are the same used by the Bally Eight Ball (not Deluxe)

1.4.8 First Time Power-Up

To get started, here is what I suggest: (this is not necessary, but it makes access a lot easier, disconnect the 15 pin Molex connector by the one coin door and lift the frame out of the cabinet. A couple of short 2/4s over the top of the cabinet allows you to sit the frame on top and still reconnect the 15 pin Molex connector.) See below. This is a good time to check the power cable too.

1) Do a visual inspection of the underside of the playfield to look for shorted wires and open connections. Check for any discoloured coils which may have heated.
2) Install the PCBs and connect all the connectors.
3) Power up the game. If any coils immediately fire or if random lights (not just the rack) come on, turn off the power and trouble shoot the logic board. If the logic board works, you should get a short tune, the display goes to all zeros, then alternates between 0 and 400000 and the balls on the rack sequentially illuminate. If the rack balls light sequentially, but no tune, you may have a bad speaker which is pretty common. A thumbwheel pot on the left side of the logic board controls volume.
4) On the keypad, push "end" then push test 1. All bulbs should continue to flash quickly. You can see if any bulbs are out.
5) On the keypad, push "end" then push test 2. The switch closures should show on the display. The electrical manual provides details on numbers. With the ball removed from the outhole, you should see something around 07 or so, depending on the position of the shaft encoder. You can sequentially go through and close the switch contacts while watching the display.
6) On the keypad, push "end" then test 3. The coils should fire sequentially, except for the flippers. You can make the flippers fire also by holding down the flipper buttons, but it also will depend on playfield positioning as to which buttons to push. Push "end" to end the test.
7) On the keypad, push "end" then test 5. The displays should all count up starting at 0.
8) Once you have successfully done tests 1-3 and 5, go to my calibration page for calibrating the playfield.

1.4.9 Troubleshooting (Electronic)

First, check power voltages. I've found a slightly low logic supply of 4.85-4.90 volts typical. A supply of less than 4.5 v will cause problems. The original Characterization board also uses -5 volts, +12 Volts and +9 volts.

For a non-booting machine, put a scope or logic probe on pin 2 of J-8 (an unused logic board connector). That is the reset line. If there are continuous resets, the board is not properly loading the software. Time to look at the Characterization board (if it is the original) the controller (8156), or bus drivers (81LS95 and 8304). A non-booting board without continious resets may indicate a missing clock signal or a bad counter (A-1).

1.4.10 Troubleshooting (Mechanical)

On a booting machine (in attract mode) access the keypad and press Test #4. The encoder should display it's position as shown on the Calibration page. Press "End", then Test #6 and on manual rotation of the playfield you should see the display pass from 0 to 1 to 3 to 2 when at each compass direction (or backwards depending on rotation). If you are not getting these displays, there is a problem with related circuitry. Missing or inorrect signals from the shaft encoder (Test #4) or the optical sensor (Test #6) can result from cold solder joints on the encoder board, or on the optical sensor board. If all the displays are ok, review my Calibration page. Remember that the ball must be in the outhole for the table to rotate.

If the playfield oscillates and can't settle to position, there may be loose hardware in the drive mechanism or on the base where the main spindle bolts to the base.

1.4.11 Access access

The General Instructions manual provides details on game assembly. Easy access to all the boards can be gained by disconnecting the Molex connector near the tilt switch and lifting the entire unit out of the case. A couple of 2x4s across the top of the case can then support the frame while allowing re-attachment of the Molex connector.

1.4.12 Using the Keypad

With the ball in the outhole, the machine on and in attract mode, open either coin door and push the small service button over the coin mechnism. The playfield should turn to expose the keypad. Some keypads do not have Test #4 to Test #10 written on them. These keys nevertheless exist in sequence after Test #3. As an example. pressing the first key on the fifth row will be Test #4, and so forth.keypad

The keypad has several functions.

1.4.12.1 Diagnostics and Calibration:

The "Test" keys provide diagnostics to assist in machine repair and calibration.

To access these tests, push the "End" key, then the desired "Test" key. Press "End" to go on to the next test.

Test #1 - Lamp Test - all bulbs are strobed and will continue until the "End" key is pushed.

Test #2 - Switch Test - remove ball from outhole. Closed switches will be displayed. Note that some of the encoder switches will always display closed.

Test #3 - Solenoid Test - all solenoids are powered sequentually until "End" key is pressed. The flipper switches must be closed to activate the flippers.

Test #4 - Encoder Switch (see calibration page)

Test #5 - Display Test (Credit/Ball display only shows the right and left 2 digits)

Test #6 - Table Position Test (see calibration page)

Test #7 - Not Used

Test #8 - Reset all game parameters to factory presets

Test #9 - Not Used

Test #10 - Not Used.

When finished with the tests, press the "Game" key.

1.4.12.2 Resetting Game Parameters:

Game features and parameters are summerized on pages 27 to 29 of the Electrical Operating Manual. To display a feature, press "End", press "Set", press ".", press desired number, then press ".". The feature number will be displayed in the "Match" display and the value will be displayed in the score displays.

To enter another value, enter the value desired using the number keys, then press "Game".

1.4.13 Calibrating the Rotating Playfield

The position of the rotating playfield is determined by two different sensors. There is an encoder mounted on the centre shaft which provides an absolute rotational value of the playfield position. Secondly, there are optical sensors which provides exact positioning for the four play positions. In simplified terms, the encoder tells the logic board which player to turn the playfield to, and the optical sensor provides an exact position to the logic board for the playfield to stop.

To calibrate the playfield, remove the glass cover and remove the ball from the outhole. Make sure all the hardware is securely fastened, the V-belt is snug and the set screws on the pulley shafts are tight.

Test #4 , accessed from the keypad, provides a read out of the position encoder. After Keypadpressing the Test #4 key, manually rotate the playfield. A value between 00 and 77 will be displayed in the "Ball" digits of the centre display. Value 00 should be North, 20 should be East, 40 should be South and 60 should be West. When finished with Test #4, press "End" before entering another test key.

Test #6 provides a read out of the optical sensors. When approaching a play position, the number 1 or 2 will be displayed, also in the "Ball" digits. When at the correct position, the number 3 is displayed.

If either of these tests do not display these values, playfield rotation will not occur properly or even at all. Before proceeding any further, these repairs must be made.

If the readouts are displayed as described, manually position the playfield to the North position and press #6. Position the playfield so that the number 3 is displayed. Now, press "End" then press key #4. If 00 is displayed, the table is calibrated and should be ready to play. if a value other than 00 is displayed, the encoder will need to be adjusted.

To adjust the encoder, there are several small set screws on the rotational shaft just Encoder Adjustabove the pulley. Loosen the set screws and turn just the encoder shaft to read 00 while keeping the large shaft with the pulley stationary. Tighten the set screws, ensuring the reading remains at 00, then go to test #6 and ensure that 3 remains displayed.

The playfield should now be calibrated. Don't forget to place the ball back in the outhole.

If during play, the table continues to "search" for the exact position, there may be loose hardware. This can be caused by loose pulleys, a loose belt or the bottom bearing bracket may be loose on its mounting screws. This bracket can be accessed (with some difficulty) by lifting the game out of the cabinet and tipping it to access the bottom of the mount. The 4 mounting carriage bolts can then be seen. Replacing them with 1/4" bolts will secure the hardware.

1.5 Repair Support

1.5.1 New Service Provider

John Robertson of John's Jukes has taken on the existing stock of parts and board servicing. John can be reached at https://www.flippers.com/.

1.5.2 New Integrated Logic Board

Tim Polzin created a replacement logic board and it integrated the daughter Logicboard onto the mainboard and have updated the power on reset and clock circuit. It was basically a re-do of the original board with a few modifications. The prototype board has been stable and reliable. The photo to the left is the first production board. All boards will only have sockets for the memory, Z-80, 8156 and audio generator.

This board was a very limited production run and is no longer available.

1.5.3 New Display Boards

Like most items on a Rotation VIII, the display board is absolutely unobtainable. On one Display machine I was working on, all but one of the displays had burned segments and warped red lenses. I found a conventional 7 segment display that worked on the signals available and had a replacement PC board etched. The result is not an exact appearing replacement, but is functionally identical and actually has a slightly larger display. These require a slight bit of filing on the apron opening, but otherwise are a plug and play replacement.

1.6 References

The Gallery has a new series of detailed photos. They were taken of Chris Welsh's machines. One machine was like new. These were the machines used in the recent Northwest Pinball and Arcade Show.

A supply of replacement daughter (Characterization) boards are on hand.

1.7 Northwest Arcade and Pinball Show - Tacoma

Two machines belonging to Chris Welsh

1.8 Videos

Below are three YouTube videos showing a fully operational machine in three different modes.