Williams WPC

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

1 Introduction

This guide covers Williams WPC (sometimes called WPC-089), WPC-S, and WPC-95 games.

2 Game List

2.1 WPC (Alphanumeric)

• Funhouse
• Harley-Davidson
• The Machine: Bride of Pin*Bot

2.2 WPC (Dot Matrix)

• Gilligan's Island
• Terminator 2: Judgement Day
• Hurricane
• Party Zone

2.3 WPC FlipTronics I & II

• The Addams Family / Addams Family Gold
• The Getaway: High Speed II
• Black Rose
• Fish Tales
• Doctor Who
• Creature from the Black Lagoon
• White Water
• Bram Stoker's Dracula
• Twilight Zone

2.4 WPC DCS Sound

• Indiana Jones: The Pinball Adventure
• Judge Dredd
• Star Trek: The Next Generation
• Popeye Saves The Earth
• Demolition Man

2.5 WPC-S CPU

• World Cup Soccer
• The Flintstones
• Corvette
• Red & Ted Road Show
• The Shadow
• Dirty Harry
• Theatre of Magic
• No Fear: Dangerous Sports
• Indianapolis 500
• Johnny Mnemonic
• Jack*Bot
• WHO Dunnit

2.6 WPC-95 CPU

• Congo
• Attack from Mars
• Safecracker
• Tales of the Arabian Nights
• Scared Stiff
• Junk Yard
• NBA Fastbreak
• Medieval Madness
• Cirqus Voltaire
• No Good Gofers
• The Championship Pub
• Monster Bash
• Cactus Canyon

3 Technical Info

At the heart of the WPC MPU is the Motorola 68B09EP microprocessor, running at 2Mhz. Note that the "E" suffix indicates externally clocked and is necessary to work in a WPC MPU. The "P" suffix merely indicates a "plastic" chip case. 6809 processors lacking the "E" suffix will not work in a WPC MPU.

The 68B09EP is an 8-bit/16-bit CPU with a 64KB address space. Bank switching is required to address more than 64KB. Bank switching is accomplished by the ASIC in WPC systems. The game ROM size varies from 128KB to 8MB, depending on the game. 8KB of battery backed RAM is available to the processor.

For more information, see The FreeWPC Manual

3.1 WPC CPU Generations

3.1.1 WPC CPU

The WPC CPU was the cornerstone of the new WPC pinball system, introduced by Williams in 1990.

The board has the following features:

• Motorola 68B09E processor
• Switch matrix inputs
• 8k of CMOS battery-backed memory
• A single EPROM, sized from 128K (1 megabit, 27C128) to 1M (8 megabits, 27C080 / 27C801)
• Watchdog circuit and master reset
• Three digital I/O ports - general-purpose; display; auxiliary (labeled "display")
• WPC ASIC - a 108-pin PLCC that has decoding, watchdog, and real-time clock circuitry.

What Goes Wrong

Loose Chips - resets and strange behavior

The natural heating and cooling cycle, and vibration, in a pinball machines cuases the socketed chips to slowly walk out of their sockets. This loosening can cause mysterious behavior and resets. Firmly pressing each socketed device - including the ASIC - and listening for a small click can solve many odd problems.

Battery Corrosion - rows of switches failing

When batteries reach old age, they always leak electrolyte. In WPC boards, this electrolyte drips onto the switch matrix circuits. There, it causes extensive corrosion. The corrosion will eat up traces under the green soldermask, creating a lumpy appearnce. This corrosion is extremely destructive. It must be caught early, and gently washed with vinegar. Damaged components must be replaced, and damaged traces cleaned with a fiberglass brush and carefully tinned with a soldering iron. Battery corrosion can render a board unrepairable, or undesirable.

The Switch Matrix - assorted switch problems

The WPC system, like many pinball machines, uses a switch matrix to read the switches. The matrix is 8 rows by 8 columns, and can read 64 switches with just 16 wires. Under the playfield, the wires snake between the lights, creating strings of rows and strings of columns. The lights are isolated with diodes.

The root cause of switch matrix problems on the WPC CPU is often a voltage short to the switch matrix. This can happen with residual charge from the solenoid or flasher supply, so the machine does not have to be on for damage to occur. One easy way to do this is to touch a coil lug while adjusting slingshot switches. The row detectors - the LM339s - are quite safe from everything but the 70V solenoid voltage, and will often survive that. The column drivers, however, are directly exposed to the wiring. So many shorts - from the GI; lamp matrix; flashers; or solenoid wiring can destroy the driver. A solenoid wiring short will often destroy the 74HCT244 ahead of the driver as well.

The symptom of a bad driver is a column short that does not go away when the playfield is disconnected. A bad driver may als be confused by neighboring columns, so switches may register twice.

3.1.2 WPC-S CPU

File:WPC-S CPU.jpg

WPC-S CPU

The WPC-S CPU is the second generation of WPC CPU board. The designer switched to a remote battery holder, located on a daughter board that clips to posts on the lower half of the board. This approach eliminated the PCB damage caused by battery leakage. The next version of the board, the WPC-95 CPU board, switched to a plastic battery holder that also protects the board. The board also has different switch matrix connectors to WPC and WPC-95.

The digital I/O connectors remain unchanged from the WPC CPU.

The most significant difference was the introduction of the security PIC. This devices contains a serial number and a handshake mechanism. The serial number is displayed at startup, and allowed the manufacturer to identify the dealer who sold a machine into a particular region. This approach was necessary when substantial price differences were common across Europe, to prevent grey imports.

To prevent simple removal of the PIC, the device also controlled the switch matrix, and would not operate without a correct machine code from the WPC CPU. Therefore, there is one PIC for each machine. An incorrect or defective chip will result in a "G13 error" at startup.

3.1.3 WPC-95 CPU

File:WPC95CPU.jpg

WPC-95 CPU Board

The WPC-95 CPU is the third generation of the WPC CPU design. It includes the security PIC added in the WPC-S system, and returns to an on-board battery holder. The new battery holder has a large plastic shell, designed to shield the PCB from any battery leakage. The pinout of the switch matrix connectors changed again from WPC-S.

3.2 WPC Power/Driver Board Generations

3.2.1 WPC-089 Power/Driver Board

The WPC power-driver board converts AC power from the transformer; drives the solenoids; drives the lamp matrix; drives the GI lamps; and detects flipper switch operation in non-Fliptronics machines.

WPC Power-Driver board with Flipper Relay

The power-driver board has the following functions:

AC rectification and smoothing for 70VDC solenoid voltage
AC rectification and smoothing for 20VDC flasher voltage
AC rectification and smoothing for 18VDC lamp matrix voltage
AC rectification and smoothing for unregulated 12V supply
AC rectification and smoothing for 5VDC flasher voltage
Regulated 12V supply
Regulated 5V supply
6.3VAC supply for GI circuits

Power control for high power solenoids
Power control for low power solenoids
Power control for flashers

Power control for lamp matrix
Power control for pre-Fliptronics filppers (not on all boards)
Sense optos for pre-Fliptoronics flippers (not on all boards)

Power control for 5 6.3VAC GI channels using four-sector triacs
Zero crossing sense for AC line in

Logic interface to WPC CPU

3.3 WPC Dot Matrix Controller board

WPC DMD Controller board

The WPC DMD display and DMD controllers are common features in Williams 1990's era pinball machines. The display is a high-voltage plasma system that has scanning electronics built in. The WPC dot matrix controller board, or WPC-95 AV (Audio/Video) board in later machines, provides the power for operation, and maintains the display dot images in its memory. The CPU loads these pages ahead of time, then instructs the display controller to "flip" the images onto the screen.

Clive at the Coin-Op Cauldron has an excellent article detailing some of the nitty gritty of the DMD controller here.

3.4 WPC Sound Boards

3.4.1 WPC pre-DCS Sound Board

WPC pre-DCS Sound Board, sound ROMs removed

The pre-DCS sound board uses a single LM1875 amp and can accommodate up to 3 sound ROMs. With a single "sound lane" and amp, stereo sound is not possible. The three speakers are connected in series, each of them attempting to accentuate different bands of the audio spectrum. A broken wire to any of the three speakers will result in no sound at all.

3.4.2 WPC DCS Sound Board

The WPC DCS Sound Board

With Indiana Jones: The Pinball Adventure, Williams introduced the DCS sound system. Its DSP architecture was a significant improvement on earlier systems, and it allowed Williams to use recorded music from EPROM rather than synthesizer music. The board also introduced dual amplifiers and an electronic crossover. DCS allowed Williams to use compressed sound samples, significantly improving the sound quality of the games.

The board dropped the microprocessor controlled synthesizer chip and sound bite system of earlier WPC machines, and replaced them with the now-obsolete Analog Devices AD2105 DSP. For its time, this was an extremely fast device (40MHz), and is capable of providing all of the sound processing needed in a WPC pinball machine. Williams also upgraded the chip amp. The new chip amp operates at a lower voltage, and forced a corresponding change in the transformer voltage (12VAC + 12VAC). Therefore, DCS machines have different transformer winding to previous WPC machines. The system has an electronic crossover, separating high frequencies to the backbox and low frequencies to the cabinet.

The card connects to the WPC CPU through the I/O port. It shares the port with the Fliptronics board and the display board.

What's on the Board?

The DCS sound board has the following subsystems:

• Rectifier and fused power supply for +/-16VDC amplifier power [check - WPC was 35V?]
• Dual-channel power amplifier with electronic crossover
• DSP-based sound synthesizer
• One EPROM for DSP code storage
• Other EPROMS for sound storage
• High-speed program memory for the DSP
• Mailbox interface for WPC-CPU

On power-up, the board runs a diagnostic and sounds a bong. It does this without any help from the WPC CPU. If there is no bong, the board has a problem.

What Goes Wrong

The smoothing capacitors C20 and C21 are often at the end of their useful lives. The chip amp, like the Honey Badger, really doesn't care, and will accomodate capacitor variance. However, replacing them will make even the pre-DCS sound much richer.

The coupling capacitors at C32 and C41 can fail.

3.4.3 WPC-95 AV Board

WPC-95 AV Board

The WPC-95 AV board integrated the sound and video system into a single board, siginifcantly reducing power and space.

The board has the following major subsystems

ASIC The ASIC is a 208-pin flatpack device. It incorporates the following functions:

• Interface registers to CPU
• Memory control and display output logic for the diaplay subsystem
• Memory interface for the AD2105 DSP

The ASIC has a tendency to fail in such a way that the sound disappears. New ASICs are available, but replacing them requires skill with surface mount devices.

Audio DSP The audio susbsystem uses a long-obsolete Analog Devices DSP part ADSP-2105. For its time, this was an extremely fast device (40MHz), and is capable of providing all of the sound processing needed in a WPC pinball machine. Previous boards used a second 68B09E for sound processing with an ADSP-xxxx

Location: lower center of board

Serial Port The serial port is rarely populated. Its uses include: machine debugging; machine linking (NBA Fastbreak); ticket printers

Location: lower left of board

Audio Amplifier The audio amplifier is a dual-channel TDA 20 operating at X volts. The system has an electronic crossover, separating high frequencies to the backboox and low frequencies to the cabinet.

Location: lower right of board

DMD Power The DMD requires 62, 95? and 107V to operate. The board uses a temperature-compensated design to supply these voltages. The 107V line is built as a 12V offset from the 95V line.

Location: upper right of board.

3.5 WPC FlipTronics I & II Boards

Beginning with The Addams Family, Williams implemented the FlipTronics flipper power and switch sensing system. The Addams Family (and only The Addams Family) used the FlipTronics I board. All subsequent games, until the WPC-95 board set debuted, used the FlipTronics II board. FlipTronics II boards are backward compatible with the FlipTronics I board.

WPC Fliptronics II

Prior to the FlipTronics system, flipper circuit design had essentially been the same simple circuit. This circuit consisted of power to two flipper coil windings (the power and hold windings). When the flipper cabinet switch was closed, the power winding and hold winding were both energized. As the flipper plunger was pulled in, the flipper EOS switch opened, interrupting power to the "power stroke" winding, leaving only the "hold winding" energized to hold the flipper in the up position. Flipper coil systems were designed this way so a flipper could be energized indefinitely (in theory). A failed EOS switch could result in weak flipper performance (poor EOS switch conduction) or result in burned flipper coils (EOS switch that never opens).

The FlipTronics system ended reliance on properly operating EOS switches. While FlipTronics systems have EOS switches, they are not used to control power to the coil windings. They merely act as feedback to the MPU that the flipper has completed a stroke. This information isn't used for anything, other than to alert the operator of a failed flipper via the "credit dot". A FlipTronics game will work perfectly even with flipper EOS switches completely missing.

FlipTronics game flipper operation:

1. Flipper power is always present at the flipper coil lugs. For the flipper to actuate, that coil power needs to find a path to ground.
2. The FlipTronics board switch sensing circuitry recognizes closure of a flipper cabinet switch.
3. The FlipTronics board communicates that flipper cabinet switch closure to the MPU.
4. The MPU commands the FlipTronics board to provide ground for the "power stroke" winding of the flipper coil (via a circuit passing through a 2N5401, TIP-102, and TIP-36C). This process is similar to how high powered coils are controlled by the power driver board.
5. A few milli-seconds later (thousandths of a second), the MPU commands the FlipTronics board to interrupt the ground path for the power stroke winding and to provide ground for the hold winding portion of the coil (via a circuit passing through a 2N5401 and a TIP-102).
6. The FlipTronics board switch sensing circuitry recognizes close of the flipper EOS switch, and communicates that switch closure to the MPU. Again, this switch closure isn't used to control power to the flipper coil windings. It's merely information.
7. When the player releases the cabinet flipper switch, the FlipTronics board again senses that open switch, communicates the change in state to the MPU which then commands the FlipTronics board to interrupt ground to the hold winding also, and the flipper returns to it's at rest position.

WPC-95 games use the exact same methods. However, the WPC-95 board set integrates the capabilities of the FlipTronics board into the power driver board.

The FlipTronics board also rectifies AC voltage sourced from the game transformer, and relays it from the Power/Driver board to create the 70VDC flipper power. A FlipTronics board can sense 8 switches, and control up to four flipper power and hold windings. In some games, spare flipper cabinet switches, EOS switches, and coil drive circuits are used for other game features.

Note that sometimes C2 is installed on this board and sometimes it is not. The board pictured does not have C2 installed. Since C2 is a 100uf/100V capacitor, it can't be doing too much AC smoothing of the full wave rectified flipper power. Some RGP discussions indicate that C2 may have been designed in originally to reduce EMI (electro-magnetic interference) from the board. Installing C2 will have a negligible affect on flipper power. Some report that installing C2 fixes a "jittery gate" symptom on BSD's "mist" assembly.

More information about bench testing FlipTronics board is available at Leon's website.

3.6 Miscellaneous WPC Boards

3.6.1 WPC 7 Opto Board

File:WPC 7 opto.jpg

WPC 7 opto board

This board is a typical example of a WPC opto board.

The connector on the lower right carry the power for seven of the IR emitters. The emitters are powered from the 12V supply through a 260 ohm resistor (the large blue resistors adjacent to the connector). Williams runs the opto emitters at about 40mA, which results in the resistors becoming hot. The opto boards are often darkened by the heat.

The connector on the upper right carries the signals for seven of the phototransistors. The phototransistor collector connects to 12V, and its emitter connects to ground through two 2kOhm resistors connected in series. When the transistor is illuminated by the emitter, it passes more current through the resistors, raising the input voltage on the LM339. The other pin on the LM339 is set to about 2.8V buy a 100k/22k voltage divider. Therefore, when the voltage at the junction of the two 2k resistors and the LM339 input is above 2.8V, the switch will read as closed. This happens at 1.4mA.

These two connectors can be used for optical switch troughs, where the wire bundling makes sense.

The three connectors on the left hand side each carry signals for one transmitter/receiver pair. These connections are often used for ramp entrances, and sometimes the pair is mounted on a single bracket.

The top connector connects to the switch matrix, and 12V power and ground.

What Goes Wrong?

The resistor discoloration is cosmetic, and rarely causes problems. The biggest problem for these boards is switch matrix shorts. If any of the power lines hits the switch matrix, it can easily destroy the LM339 and possibly its output diode. This issue will manifest as a ground short on one or more rows. It can be tested by removing the switch matrix connector. If the problem goes away, the associated LM339 and possibly its diode need to be replaced.

If the short was from the 70V supply, multiple LM339s will be damaged.

Replacing LM339s on these boards needs to be done carefully, it is easy to lift traces and pull the through-hole plating.

3.6.2 WPC 10 Opto Board

+++Add pic of 10 - opto board+++

The connector on the lower right carry the power for seven of the IR emitters. The emitters are powered from the 12V supply through a 260 ohm resistor (the large blue resistors adjacent to the connector). Williams runs the opto emitters at about 40mA, which results in the resistors becoming hot. The opto boards are often darkened by the heat.

The connector on the upper right carries the signals for seven of the phototransistors. The phototransistor collector connects to 12V, and its emitter connects to ground through two 2kOhm resistors connected in series. When the transistor is illuminated by the emitter, it passes more current through the resistors, raising the input voltage on the LM339. The other pin on the LM339 is set to about 2.8V buy a 100k/22k voltage divider. Therefore, when the voltage at the junction of the two 2k resistors and the LM339 input is above 2.8V, the switch will read as closed. This happens at 1.4mA.

These two connectors can be used for optical switch troughs, where the wire bundling makes sense.

The three connectors on the left hand side each carry signals for one transmitter/receiver pair. These connections are often used for ramp entrances, and sometimes the pair is mounted on a single bracket.

The top connector connects to the switch matrix, and 12V power and ground.

What Goes Wrong?

The resistor discoloration is cosmetic, and rarely causes problems. The biggest problem for these boards is switch matrix shorts. If any of the power lines hits the switch matrix, it can easily destroy the LM339 and possibly its output diode. This issue will manifest as a ground short on one or more rows. It can be tested by removing the switch matrix connector. If the problem goes away, the associated LM339 and possibly its diode need to be replaced.

If the short was from the 70V supply, multiple LM339s will be damaged.

Replacing LM339s on these boards needs to be done carefully, it is easy to lift traces and pull the through-hole plating.

3.6.3 WPC 16 Opto Board

WPC 16 opto boards, both versions

This board is used in at least Star Trek: The Next Generation and Champions Pub. Note the two different "versions". The bracket on one side is believed to be part number 5768-13739-00 while the bracket on the other side is believed to be part number 5768-13739-01. The boards are identical in layout and components. The only difference is bracket mounting, some silk screening, and one version has holes in the PCB which are not used.

3.6.4 Auxiliary 8-Driver Board

This board sits in the top right of the backbox in many WPC machines. The connector labeling is tricky as the 8-Driver board is controlled by a ribbon cable that is connected to the "Display" connector on the WPC CPU board (J204). The "Display" label is a historical artifact and refers to the alphanumeric display used in Funhouse.

The board has 8 outputs, driven by TIP-102 transistors with a simple 2N4403 pre-driver, similar to the WPC power/driver board (which uses a 2N5401 as a pre-driver). It is used in the following games:

• Demolition Man
• Indiana Jones
• Road Show
• Star Trek: The Next Generation
• Twilight Zone

The board circuitry can also be used to add columns to the switch matrix (e.g. Indiana Jones, Star Trek: The Next Generation).

An example 8-Driver board

The board is designed to run flashers, low-power solenoids, and perhaps motors. The two lowest transistors can be repurposed to add columns to the switch matrix or lamp matrix circuit. This is accomplished via jumpers on the board. This makes the boards more flexible but also may require different jumpers to use the board in different games (i.e. the boards are not completely interchangeable between machines "as is").

What Goes Wrong?
This is a pretty simple board. Generally, the problem is going to be a blown transistor, or possibly a failed 74ALS576 latch.

One thing to watch for is that the board ground connection may be hidden under the ribbon cable between the CPU and the power-driver board. The board may partially work if this connection is open, which can be a big timewaster. So, check the ground wire if this board seems to be giving trouble.

Another thing to watch for is a "diode tieback" connection to the board for games like Star Trek: The Next Generation. Without this "diode tieback", transistors on the 8-Driver board WILL be damaged.

3.6.7 Bi-Directional Motor Board (A-15680)

A-15680 Bi-directional Motor Board

This board is used in Cirqus Voltaire, Dr Who, Party Zone and White Water.

One problematic aspect of the board's design is that the 100uf electrolytic cap is mounted over the trace that carries 12VDC power to the drive transistors. If the electrolytic cap fails and leaks it's corrosive contents, it will eat through the trace and the board won't work correctly. If this happens in WH2O, Big Foot will not turn his head during game test at boot up. Note: the board pictured at left exhibited this problem and was repaired.

3.6.9 24 Inch Opto Board

WPC 24 Inch Opto Board as used on BSD and Shadow

This board is used on Bram Stoker's Dracula to sense the "Mist Ball" and on Shadow to sense when the ball is at the bottom of the mini playfield. The usual problem with this board is that the large inductor (marked "106C" in the picture) vibrates off the board. The circuitry on this board is similar to a television remote control circuit. It senses a particular signal frequency.

3.6.10 Coin Door Interface Board

WPC/WPC-95 Coin Door Interface Board

The coin door interface board provides electrical connection between the coin door switches and the WPC switch matrix and dedicated switches like the "diagnostic" and volume control buttons. The coin door interface board also creates the "always closed" switch 24 by bridging column 2 and row 4 with a diode. WPC games use this "always closed" switch to test game functions.

If your coin door interface board gets dusty from flipper parts wearing, clean it. That dust conducts electricity and may cause some interesting coinage issues. There really isn't much that can go wrong with this board.

4 Problems and Solutions

4.1 Game blows power line circuit breaker immediately

Had a thunderstorm roll through recently? The varistor in the power box may have done it's job protecting the rest of the games circuitry from the possibly tremendous power surge caused by electrical storms.

4.2 Credit Dot

The credit dot is a decimal point (a single, lit pixel on the display), that appears after the number of credits or after the words free play on the DMD display on WPC games. It is a quick indicator used to tell the operator or owner that there may be an issue with the game. Most commonly, there is a switch or multiple switch problems with the game. However, a credit dot will appear for other reasons, such as when the date and time are not set.

If, after 30 balls or 10 games, a particular switch hasn't registered, the credit dot is posted to alert you to check the diagnostics. The credit dot may or may not indicate a real problem. For the example shown in the pic, it's simply reporting a switch that is hard to hit in the game. So, the credit dot is actually saying "Play Better". :-)

4.3 Relocating the battery from the MPU board

Relocating the 3xAA battery pack from the MPU board is always a good idea. Leaky alkaline batteries are the #1 killer of MPU boards. Simply removing the batteries is not an option with WPC games as you will always receive a "Factory Settings Restored" message when the game boots.

Options:

1. Remotely locate the battery holder somewhere below all other boards. This ensures that even if the remotely located batteries leak, they won't leak onto (or even drip onto) any circuit board. Replace the batteries annually, dating them with a Sharpie! as you do.
2. Replace the 6264 static RAM with a SIMTEK non-volatile RAM (STK12C68). These SIMTEK RAM chips are increasingly hard to find but offer a nice alternative to changing batteries annually. This method requires desoldering/soldering on the MPU and also has the down-side of not maintaining the Real Time Clock (meaningless in some games...nice in games like Twilight Zone that moves the playfield "toy" clock to the current time during attract mode, and Who?Dunnit which has a "Midnight Madness" feature).

4.5 Game resets

One of the more annoying faults exhibited by WPC machines is the chronic reset problem. The game will restart, often when both flippers are used at the same time.

There are several possible reasons for game resets. Before doing any work on the circuit boards, it is recommended that a number of much easier to fix, and just as probable root causes, be examined.

Possible causes are listed below, in the recommended order of examination. This order is a derived from a combination of "ease of examination" crossed with "probability of root cause". The standard advice dispensed on RGP to change BR2 and C5 is often wrong and unnecessary. If BR2/C5 are changed out and the game no longer resets, some will conclude that, sure enough, BR2/C5 was the problem. What is being missed is that along with replacing those components, at the very least, the connectors on the board are being reseated, the solder on the cap and bridge is being reflowed, and the board ground to the backplane is perhaps better due to torqueing the screws better. A poor board ground can cause enough voltage drop to induce resets.

4.5.1 WPC 5VDC Power Derivation Path

Following is the nominal path from the wall plug to eventual consumers of the clean 5VDC power supplied by the Power/Driver board.

1. 120VAC/60Hz power is nominally provided at the wall plug. Different voltages are accommodated also (see below)
2. An RFI (radio frequency interference) filter, which is seldom a contributor to game problems.
3. Line power is connected to a service outlet sometimes incorporated into the metal "power box" and made available to the tech via a short, uniquely plugged, extension cord. Line power is wired in parallel and does not require the power switch to be in the ON position.
4. Both sides of the AC power pass through the double pole/double throw (DPDT) game power switch
5. The "black" side of the AC power passes through an 8A FB fuse housed in a bayonet style (screw in) fuse enclosure.
6. The "black" side of the AC power then passes through the thermistor, which limits current inrush into the system.
7. "Butterfly" voltage select connector used to accommodate typical voltage systems used around the world (120VAC in US, 230VAC abroad, low line voltage, etc). When configured for 120VAC in the US, the connector and wiring resemble a butterfly.
8. Transformer Primary windings
9. Transformer Secondary windings
10. Yet another Molex connector, mating with the harness that will eventually lead to the Power/Driver board.
11. Some games incorporate a "coin door interlock switch" which interrupts coil and flasher power when the coin door is open. If your game is equipped with this switch, there will be an extra molex connector necessary to implement this switch.
12. J101 on the Power/Driver board
13. Full Wave AC rectification and smoothing via BR2 and C5
14. Regulation to 5VDC via the LM323K at Q1
15. Lighting LED4...
16. And finally, delivery to 5VDC power consumers via J114, J116, J117, and J118

4.5.2 Low Line Voltage

Step 1: measure line voltage

Low line voltage can occur for several different reasons...

• Peak summer usage may cause a drop in line voltage (at the wall). Nominal voltage should be 115VDC minimum.
• Too many games or "power hungry" devices on the same circuit. This is common when numerous devices are plugged into a power strip.

If your machine is resetting, open the coin door and measure the AC voltage at the auxiliary socket with the machine turned off, and then with the machine turned on. If the voltage at the auxiliary socket is over 105VDC, the problem is most likely inside your game. If the voltage is below 105VDC, then you have a line voltage problem and you'll need your local electricity company's help to correct the problem.

4.5.3 Poor Ground Connection for Power/Driver Board

Step 2: Make sure the power/driver board is screwed down tightly

A poor ground connection from the power/driver board to the backbox ground plane can reduce 5VDC by as much as .3 volts. Ensure that the screws that hold the power/driver board down secure it tightly.

4.5.4 Cracked Power or Ground Header Pins

Step 3: Examine the header pins at J101, J102, and J103

Similar to step 2 above, cracked header pins where the power enters the board or where ground is returned to the board can cause game resets. In the picture at left, cracked header pins at J103 (ground) in a Twilight Zone caused game resets. Either reflow the solder, or better yet, replace the headers with new pins.

4.5.5 Missing diodes, open diodes, or cold solder joints at the Flipper coils

Step 4: Examine flipper coil diodes

WPC flippers are, in general, the only coils in the game that should have diodes across the coil lugs. While diodes seldom fail, sometimes the pounding environment that flipper diodes must live within causes solder joints to fail. Ensure that each flipper coil has the required two diodes, that the diodes are good, and that they are soldered to the coil lugs with a quality solder joint.

4.5.6 Poor Connections between the Transformer Secondary and the Power/Driver Board

Step 5: Examine the "cube shaped" molex connector in the cabinet

The connections from the transformer secondary to the Power/Driver board can degrade over time, a natural result of hours and hours of "power on" time. Try re-seating each connector. If the reset frequency is reduced or disappears, then you've probably discovered the root cause. You should re-pin these connectors with at least new pins (the connector housings may be salvageable) for a long term solution.

Note: the molex pin extractor for these "round" pins is relatively expensive but there is no other good way to remove them.

4.5.7 Poor Connections between the Power/Driver board, the CPU, and other PCBs

Step 6: Examine other wiring harnesses and the connectors

Measure DC voltage at TP2 on the Power/Driver board.

1. DMM set to DC volts (if not an auto-ranging meter, expect to measure 5VDC)
2. Black lead secured to game ground (the ground braid in the head is a good place to pick up ground)
3. Red lead on TP2. Remember the reading

Measure DC voltage at the CPU board.

1. Again, DMM set to DC volts
2. Black lead in the same place
3. Red lead on pin 32 of the game ROM, or on pin 2 (center pin) of the MC34064 under voltage watchdog at U10
4. Record this reading

Compare the two readings. ANY difference indicates voltage loss across the connectors. Some voltage loss is expected/natural since no circuit path can provide zero resistance. A difference of .1VDC is enough to warrant examination of the connectors between J114 on the Power/Driver board and J210 on the CPU board.

Reseating the wire harness that connects the power-driver board to the CPU board may identify the root cause (but reseating is not a long term solution). Some games use an interceptor harness to power extra boards (IJ, TZ, WH20, etc). There is a inline splice connector (commonly called a Z-connector) between the driver board and the CPU board. You can replace the Z-connector or, more robustly, eliminate it by splicing the wires together, soldering and heat-shrinking. This removes the Z-connector as a contributor to game resets entirely.

Tarnished or heat damaged header pins create resistance and are sometimes a contributor to game resets. You can bet that if the male header pins are tarnished, their female mates are tarnished also. Should you find tarnished pins, now might be a good time to remove that aspect as a contributor to game resets.

If you are following this step-by-step process, now is the first time that you'll need to use a soldering/desoldering station.

<place a link to "best practice" soldering/desoldering here...note to self by Chris Hibler>

Replace the male header pins, remembering to clip the correct "key" pin before reinstalling the board into the game.

Replace the female connector and pins with good quality Trifurcon crimp-on pins. IDC (Insulation Displacement Connectors) connectors as found originally on your game were used by the OEM to speed the manufacturing process and do not provide the level of robustness that can be achieved using Trifurcon crimp-on pins and good technique.

Another possible contributor to poor power connections is cracked header pin solder joints. Although, with WPC double-sided PCBs and the relatively large header pin through-holes, this is rarely seen. Should you identify cracked header pin solder joints, it's best to remove as much of the old solder as possible before reflowing new solder at the joint.

4.5.8 Using a Multimeter to Test the Bridge Rectifier and Capacitors

Step 7: Bridges and Caps in circuit

There's one last test we can do before disassembling parts of the machine. This test can clearly identify a bad bridge rectifier or capacitor. You will need a clip lead. With the machine off, clip the lead onto the top left lead (positive) of BR2. You will have to do this by feel, as you cannot see the lead. This clip point is the output of the bridge rectifier, and is connected to the capacitor and LM323K regulator input. It should be at 9V DC.

Then, place the other clip on the red (+) lead of your meter, and connect the black lead to ground by tucking it under the braid. Check your installation to make sure that nothing is shorting.

Now, set your meter to DC volts and turn on the machine.

The meter should read 9V. This will be true with good and bad capacitors. If the bridge rectifier is bad, it will read about 7V.

For the capacitor, turn your meter to AC volts. Note that inexpensive meters may give false readings here. Ideally, your meter should be a true RMS meter.

Good capacitors will read about 300mV AC. Bad capacitors will read up to 2VAC. Anything over 1VAC indicates a failed/failing filter capacitor at C5 which should be replaced.

4.5.9 Failed Thermistor

Step 8: Examine the Thermistor

The thermistor is (generally) a black or grey disk, about the size of a dime. It is located inside the power box which is found just inside the coin door and to the right. If your game has one (and not all WPC games do), it will be connected in series between the power box fuse and the black side of AC power. Note that the power box may also contain a "varistor", or MOV, which is essentially a surge protector. The varistor will be wired in parallel with (across) the AC power. The varistor is not a factor in game resets.

The thermistor's job is to limit current inrush into the capacitors when the game is first powered on. This reduces stress on the bridge rectifiers or diodes in the game's power circuits (which is the primary cause of bridge rectifier failures). After a few seconds, the thermistor heats up and drops to a very low resistance. Failing thermistors pass less current and have to get hotter to work. This heating takes time, so the game will often reset in the first 30 minutes of operation, and then be fine afterwards. Obviously, a cold environment will make the symptoms worse, and a warm room may appear to cure the problem.

Resetting while the game warms up is therefore a key indicator of a failing thermistor. Note also that DCS and WPC-95 A/V boards may reset independent of the MPU. If this is the case, you'll hear the characteristic same "bong" as when the game boots.

***Safety Warning*** Unplug the game AND turn the game off before conducting the following test.

Sometimes, the thermistor may be visibly damaged. However, it may look good and still be bad. An easy test of the thermistor is to jumper across the legs of the thermistor with a heavy gauge wire. If the game resets no longer occur, replace the thermistor with the correctly rated part. The original Williams part number is 5016-12978-00. A replacement is available from Great Plains Electronics.

4.5.10 Questionable Prior Rework

Step 9: Examine prior rework

The printed circuit boards used in pinball machines sometimes have traces on both sides of the board. Most WPC boards are manufactured in this way. The traces are joined through the board by a thin layer of copper, plated to the inside of the hole. This plating is delicate. Without great care, the plating can pull out or crack when components are removed, especially alkaline damaged components as found on MPU boards (after batteries have leaked) and "snap caps" as found on OEM WPC Power/Driver boards (at C5 for instance).

Extreme care as well as good technique must be exercised when removing these components as it is very easy to lift traces or damage through-holes on the board. If the component is known to be bad, it is sometimes easier to snip the component from the board with a flush cutter and remove each leg of the component individually.

An excellent way to repair a lifted trace or cracked through-hole is to create a 'solder stitch' between the traces on each side of a two sided board. Some PCB repairmen will install jumpers as a "repair" for cracked through-holes or to "guarantee" connectivity between components. Some PCB repairmen will avoid the use of jumpers in favor of better looking rework.

4.5.11 Failed Capacitors

Step 10: Power smoothing capacitors at end of life

Electrolytic capacitors do not last forever. They are designed to operate for about 1,000 hours at their full rating. Capacitor life is a function of temperature: the cooler the capacitor operates, the longer the life. The system designer must select higher voltage and capacitance ratings to achieve the design life, which can be as short as 5 years in a pinball machine operated continuously in a warm environment.

The WPC Power/Driver board uses a 15,000uf, 25V capacitor at C5 to smooth the full wave rectified AC from BR2. When this capacitor is no longer smoothing the rectified AC well enough to produce a clean 5VDC, the voltage may drop below the threshold enforced by the watchdog circuit on the MPU (an MC34064 at U10). When this happens, resets sometimes result.

Note that commonly available replacement capacitors are rated to 35V, and will work fine for this application. In general, you may use a capacitor rated for higher operating voltages but the spec capacitance (uf, or micro-farads) should not be changed.

C4 (spec'd at 100uf, 10V, axial) can sometimes fail and cause the game to reset also.

4.5.12 Failed Bridge Rectifier

Step 11: Test Bridge Rectifier #2 (BR2)

The thermistor protects the bridge rectifiers from inrush current. Still, bridge rectifiers will occasionally fail.

Testing a bridge rectifier is simple.

1. Place your DMM into "diode test".
2. Put the black lead of your DMM on the "oddball" lead of the bridge rectifier. This will be the lead that isn't oriented the same as the others (as with the "spade" type of bridge) or the lead that prevents the four legs from forming a square (as with the "wire lead" type of bridge). This will also be the DC positive lead of the bridge.
3. Place the red lead of the DMM on each of the adjacent legs, one at a time.
4. A reading nominally between .5 and .7 should be seen (this represents the voltage drop across the bridge's internal diodes).
5. Now place the red lead of the DMM on the lead opposite of the "oddball" lead, or the DC negative lead of the bridge.
6. Place the black lead of the DMM on each of the adjacent legs, one at a time.
7. Again, a nominal reading between .5 and .7 should be seen.

Readings outside of these ranges indicate a failed or failing bridge. Note that these readings are not "hard and fast". For instance, a reading of .452 is probably acceptable. We are looking for an "open" or a "short". Note also that this test is not conducted "under load" and it is possible for the bridge to test "good" when it will in fact fail under load (this is also true when testing diodes, transistors, etc).

Should you decide to replace the bridge, it is best to screw the heat sink to the bridge before soldering it in place as the heat sink is shared between BR1 and BR2. A small amount of heat sink compound between the bridge and the heat sink is necessary. This compound improves conduction of the heat away from the bridge and into the sink.

Some repair tips suggested cutting the heat sink in half, separating it into two parts, one part for each bridge. This technique is no longer considered to be a best practice.

Solder both the top and bottom solder pads of the bridge. Should the through-hole of the bridge leads be cracked or otherwise damaged, use the "solder stitch" technique noted above or install jumper wires between the DC outputs of the bridge and the terminals of C5. When clipping the leads of the wire bridge, do not cut into the solder "meniscus" (the "volcano" of solder around the wire lead). Cutting into the solder meniscus can cause cracked solder joints later.

With proper desoldering tools, like the Hakko 808 or Hakko 470D, removal of bridge rectifiers and other components is greatly simplified.

4.5.13 Failed Voltage Regulator

Step 11: When all else fails...

The LM323K 5V regulator is a robust device, but it can drift over time to below the design requirement of 4.8VDC. If your 5VDC reads less than 4.8VDC at the test point, or less than 4.75V on the CPU board, you have to replace the regulator. Note that these voltages are compliant to the LM323K's spec. However, at the lower end of this spec, this lower voltage may be the cause of resets.

4.5.14 The Absolute Last Resort

Step 42: We really didn't want to get to this point!

There is a method of absolute last resort. This is absolutely, positively a hack. And yes, it can push some extra margin into the design, covering up other problems, which is why it is a hack.

The LM323K draws a small current to operate, which it passes to ground. We can use that current to raise the voltage of the whole LM323K, pushing the output closer to 5V.

Procedure:

1. Isolate the LM323K's metal case from board ground by cutting the ground traces that run left and right from the nut on the back of the board that bolts the LM323K to the board. The neater you make these cuts, the easier it to reverse this hack.
2. Some versions of the Power/Driver board include a ground trace on the component side of the board that provides a ground to the tantalum capacitor at C9. This trace must be cut on the component side of the board.
3. Add a 1/2W resistor in series, from the LM323K ground to the board's ground trace.

A 12 ohm resistor will nominally raise the LM323K voltage output by .1 VDC.
A 24 ohm resistor will nominally raise the LM323K voltage output by .2 VDC.
Note: The LM323K is designed to operate this way.

4.5.15 The Conductive Grease Hack

What is a "Hack"? Some techs will recommend the use of conductive grease to improve conductivity at a connector. While some techs will "swear by the stuff", other techs will "swear at the stuff", and as such, this is not considered a best practice. The grease can extend the life of new header pins, and protect against hostile environments that would cause corrosion. However, when used on header pins that are already damaged, grease can only slow further deterioration. Worse, the heat in the connector will turn the grease to gunk. If you find a flaky connector with gunk in it, you've been "greased" and you will want to re-pin it.

4.6 Check fuses F114 and F115 message

This message is sometimes displayed when the game boots and a game cannot be started. Most of the time, F114/F115 will be found to be just fine.

The game issues this message because it cannot read the switch matrix normally. The crafty designers at Williams inserted an “always closed“ switch into the matrix as switch 24. Switch 24 isn’t a switch at all, but instead provides the game software with a “known closed” switch matrix return. Switch 24 is actually a diode across column 2, row 4 of the switch matrix and is located on the coin door interface board.

Since the WPC switch matrix circuitry on the MPU uses 12V to determine switch state, the game assumes that the 12VDC power has been interrupted. F114/F115 fuse the 12V generation circuit; F114 on the AC side of BR1; F115 on the DC side of BR1. Hence, the game assumes that one of these fuses is blown and the message is displayed.

Diagnosing the problem

Start by checking DC voltage at TP3. You should see about 12VDC. If not, follow these steps.

1. Actually check fuses F114 and F115 using the procedure here. If a blown fuse is found, replace it. If F114 immediately blows again, then BR1 is probably shorted. Skip to the paragraph below to test the bridge.
2. Check LED6. If not lit (indicating the absence of 18V at TP8), then suspect BR1 or the filter caps for BR1, C6 and C7. If BR1 had failed shorted or if C6 or C7 where shorted (rare) then you would expect to see F114 blown. BR1 can be tested using the procedure below.
3. Check LED1. If not lit (indicating the absence of “digital” 12V at TP3, then test D1 and D2 (both 1N4004 diodes). If D1/D2 test good, then check continuity from pin 2 of the 7812 voltage regulator at Q2 to J114, pin 1. This verifies the path through F115.
4. If this all checks out, yet 12VDC is still not at TP3, then suspect the 7812 voltage regulator at Q2. Q2 is in a TO-220 package (like a TIP-102) and has a small heat sink attached.

If TP3 does, in fact, have 12VDC present, then we need to dig deeper.

1. First check for 12VDC on the CPU board to ensure that the 12V is getting to the board via the 2 (and possibly 4 if your game has a “Z-Connector”) connectors that carry the 12VDC. An easy place to measure 12VDC on the CPU is at pin 10 of the ULN2803 (U20). Pin 10 is on the bottom row of pins, furthest left. Note that this connection (an also the 12VDC connection to pin 3 of the LM339s) is not shown on the schematic.
2. If the CPU is receiving 12VDC, and you still have the “Check fuses F114 and F115” message, then the problem lies within the CPUs switch matrix logic circuitry.
3. Prime candidates for CPU switch matrix failure are U20 (a ULN2803), the two LM339s at U18 and U19, the 74LS374 at U14, and the 74LS240 at U13. All of these ICs are in the “corrosion zone” caused by leaky alkaline batteries. If your CPU exhibits the “blue/green fuzzies”, address the alkaline damage first.
4. Testing the 74LSXXX ICs is simple using the procedure here.
5. If coil power, flasher power, or even lamp power was somehow shorted to the switch matrix, there is a near 100% probability that the ULN2803 at U20 has been damaged. There is no good way to test the ULN2803 other than using a logic probe with the game powered on. Both the input side and the output side of the ULN2803 should be toggling between logic 0 and logic 1. If not, socket and replace the ULN2803.

File:Test-br.jpg

Checking an uninstalled bridge rectifier

Testing a bridge rectifier is simple.

1. Place your DMM into "diode test".
2. Put the black lead of your DMM on the "oddball" lead of the bridge rectifier. This will be the lead that isn't oriented the same as the others (as with the "spade" type of bridge) or the lead that prevents the four legs from forming a square (as with the "wire lead" type of bridge). This will also be the DC positive lead of the bridge.
3. Place the red lead of the DMM on each of the adjacent legs, one at a time.
4. A reading nominally between .5 and .7 should be seen (this represents the voltage drop across the bridge's internal diodes).
5. Now place the red lead of the DMM on the lead opposite of the "oddball" lead, or the DC negative lead of the bridge.
6. Place the black lead of the DMM on each of the adjacent legs, one at a time.
7. Again, a nominal reading between .5 and .7 should be seen.

Readings outside of these ranges indicate a failed or failing bridge. Note that these readings are not "hard and fast". For instance, a reading of .462 is probably acceptable. We are looking for an "open" or a "short". Note also that this test is not conducted "under load" and it is possible for the bridge to test "good" when it will in fact fail under load (this is also true when testing diodes, transistors, etc).

The heat sink covering BR1 and BR2 on a WPC-board

Take special care when replacing a bridge rectifier. It is easy to lift the traces of the plated through holes when removing these. After removing the heat sink, cut the old bridge rectifier off of the board close to the top leaving as much of the lug as possible. Then add a small amount of solder to the connection on the solder side to improve heat transfer. Finally apply the iron to the solder side to heat the joint and remove the remaining lug with small pliers.

When replacing the new bridge rectifier you should remove the old thermal compound and apply a new thin layer. The old compound can be removed with a little rubbing alcohol. Apply the thermal compound to the top of the bridge rectifiers and reattach the heat sink with the screws. It is much easier to replace the heat sink before resoldering the new bridge rectifier. Make sure that the heatsink is sitting flush with the tops of the bridge rectifiers and that you leave about a 5/8" gap between the bridge rectifier and the board to improve air flow.

Splitting the heat sink between BR1 and BR2 was at one time thought to improve reliability. This is no longer considered a best practice and is not recommended.

Diodes replaced BR1 and BR2 on WPC-95 boards

In WPC-95 games, BR1 and BR2 were replaced by a series of diodes. D11-D14 replaced BR1. D7-D10 replaced BR2. These are much less prone to failure. Still, they do fail on occasion. If they fail, it's best to replace all four in the "gang".

4.7 Voltage Problems - LED2 & LED3

LED2 and LED3 on the Power Driver Board are indicators for the line voltage detection circuit. On on-board comparator circuit uses +5V and +18V to check the line voltage. The comparator consists of a resistor voltage divider and an LM339 comparator.

The following table shows how to interpret the LED states. Note that other problems with the +18V or +5V supply could cause a high or low line voltage indication.

Line Voltage Indicator Chart

LED2	LED3	Voltage
On	Off	OK
Off	Off	High
On	On	Low

4.8 Solenoid & Flasher problems

Before proceeding to diagnose solenoid or flasher problems, see this section: How coils and flashers are turned on

4.9 Lamp problems

Lamps (or globes for those of you in the UK) fall into two categories. "General Illumination" and "Controlled Lamps". Your game probably has other lamps which are actually "flashers" that require an 89 or 906 bulb. Flashers are covered elsewhere in this Wiki.

General illumination lamps (GI) provide the ambient lighting for the playfield, backbox, and coin door. These lamps are list most of the time. The only "controllable" aspect of these lamps is their brightness.

Controlled lamps are under complete CPU control via the lamp matrix. These lamps illuminate the various "features" of the game such as mode inserts, pop bumper lamps, inlane/outlane insert lamps, etc.

4.9.1 General Illumination Problems

Background

The WPC power/driver board provides 5 GI circuits under CPU "brightness control". The WPC-95 power/driver board provides 3 GI circuits under CPU "brightness" control and 2 circuits that are always powered. Those two circuits are always connected to the backbox lamps.

GI power is provided by the transformer in the form of 6.3VAC at power/driver board connector J115. Although there are five pairs of wires, they are all connected together at the transformer. One side of the AC GI circuit on the power/driver board is fused by F106 through F110 which are all 5ASB fuses. This side runs to the GI lamps, and carries 6.3VAC to the backbox, playfield, and coin door via power driver board connectors J120, J121, and J119 respectively. J119 is a 3-pin header that is always connected to the coin door lamps. J120 and J121 are electrically and physically identical (keyed at pin 4) and therefore can be connected to either the female playfield GI connector or the female backbox GI connector without care. The other side of the GI AC feed is connected to ground.

The GI power is controlled by a triac in the return line from the playfield. When these triacs are off, no current passes and the lamps are off. When the triac is turned on, the GI lamps glow. The triac is a slightly special part - it must be a "four sector" triac. This means that it can switch both positive and negative sides of the AC cycle with only a positive signal.

Each triac is switched on/off by a 2N5401 transistor which is controlled by U1, a 74LS374, an octal D-Type Flip-Flop with three state outputs (a fancy way of saying that the device is an 8-bit data buffer).

The CPU can dim the GI by switching on the GI for only parts of the AC waveform. To do this, it uses a zero crossing signal generated from the 6.5VAC that feeds the 5V circuit.

Wire colors to J115 during the early WPC games were always yellow (AC) and yellow-white (AC return). Pin 1 of J115 provides the "ground reference". Later WPC games used different wire colors at J115 and instead of "looping" power through the connector at J115, connected two wires to the same pin at the 9-pin connector between the transformer secondary and J115.

J119 is always keyed at pin2, with a white-Violet wire at pin 1 and a Violet wire at pin 3

J120 and J121 are 11 pin connectors. Wire color/positions (when used) are identical for all WPC games and are shown in the table.

Pin	Wire color at J120/J121
1	Brown
2	Orange
3	Yellow (yellow)
4	Key
5	Green
6	Violet
7	White-Brown
8	White-Orange
9	White-Yellow (yellow)
10	White-Green
11	White-Violet

Common GI Problem Causes (listed by probability/ease of testing and correction)

GI Bulbs Not Lighting

Let's start by determining if the problem is "off board", i.e. not on your power/driver board, or "on board".

1. Take note of the wire color of the GI string that is not working.
2. Set your DMM to AC voltage. If your DMM is not an "auto-ranging" model, expect to measure about 7 volts AC.
3. Turn the game on.
4. Insert one probe of your DMM (either one) into the rear of the female connector position with the same wire color as noted earlier. Make sure you make contact with the conductor.
5. Insert the other probe into the rear of the female connector with the white wire that has the same color "tracer" as the color noted earlier.
6. You should be measuring 6 to 7 VAC.
7. If there is voltage at the connector, then the problem is "off board".
8. If there is no voltage at the connector, then the problem is "on board".

Off board GI problems

1. All lamps blown. Don't laugh. Just for giggles, put a known working lamp into one of the lamp sockets of the suspected GI string. I've seen it more than once. Usually the result of operators never changing a bulb, or somehow connecting higher voltage to the circuit.
2. "Open" in GI wiring. The GI wires leading from J120/J121 may be cut/broken between the connector and the GI lamp sockets. Remove both of the female connectors at J120/J121. Use your DMM to check continuity between the lamp socket and the appropriate wire/pin at J120/J121.

On board GI problems

1. Blown fuses. This is also the easiest problem to fix. Check each GI fuse following this procedure. Don't trust your eyes.
2. Burned connectors. <need a picture> Over long periods of time, both the male and female GI connectors at J115, J120, and J121 often burn. This was especially prevalent in early WPC games like Terminator 2. Numerous "hacks" have been employed over the years by well meaning repairmen but the only real way to fix burned connectors is to replace both the male header pins and the female housings/pins. "Trifurcon" phosphor-bronze crimp-on pins are recommended to provide a better connection at the female connector. Note that these must be phosphor-bronze (7A rating, as original). Don't buy them unless the supplier calls that out as the brass version looks the same but is only rated at 5A. If you replace burned female connectors, the male connectors are probably burned too and should be replaced. The damaged surface of the male pins will generate heat, and cause early failure of the replacement connector. Plus, the GI will be dimmer than it should be. J115 is a special case. The black Panduit connector that Williams used on many machines is far superior to any crimp pin, with a 12A rating (compared to 7A for phosphor-bronze Trifurcon pins). These, however, can be hard to find. Fortunately, they rarely burn and therefore rarely require service.
3. Burned AC power input connector. There is also a 9-pin (3x3) connector between the transformer secondary and J115. This connector/pins sometimes burn and may require replacement. Check to ensure that AC power is being delivered to the power/driver board at J115.
   1. Set your DMM to AC voltage. If your DMM is not an "auto-ranging" model, expect to measure about 7 volts AC.
   2. Turn the game on.
   3. Insert one probe of your DMM (either one) into the rear of the female connector at J115, pin 3.
   4. Insert the other probe into the rear of the female connector at J115, pin 11.
   5. You should be measuring 6 to 7 VAC.
   6. If no voltage is present, examine the 9-pin connector between the transformer secondary and J115.
4. Burned traces. Usually, this is a secondary problem created after the connectors burn. Visual inspection of the traces might uncover the problem, but "buzzing" them for continuity is the best practice.
5. Failed 2N5401. These don't fail very often, but are easy to test following this procedure.
6. Failed resistors. Again, these don't fail very often but are easy to test with your DMM. Keep in mind that "in-circuit" measurements may not be reliable.
7. Failed 74LS374. This is rare too but easy to test following this procedure. You can also test the outputs of the LS374 by probing pins 2, 5, 6, 16 and 19 with your logic probe. At full GI brightness, these pins should test as "high".
8. Failed Triac. These rarely fail. But, if you've gotten to this step in the procedure, it's time to replace the Triac. Remember, you need a 4-sector triac.

As these boards age it is easy to pull out the little copper hole liners that connect traces from one side of the board to the other. If you install new connector pins, be sure to check continuity from the new pins onto the board somewhere. This only takes a couple minutes and can save you a lot of time.

GI Lamps Not Dimming

The primary cause of this problem is broken traces at BR2, usually caused when BR2 was replaced. These traces feed D3 and D38. Both of these traces must be intact for dimming to work.

4.9.2 Controlled Lamp Problems

Under construction...

4.10 Switch Problems

Switches in WPC games fall into two categories, those within the switch matrix, and "direct" switches.

4.10.1 Direct Switch Problems

Direct switch operation

Direct switches include:

• Left coin chute
• Center coin chute
• Right coin chute
• 4th coin chute
• Service Credits/Escape (referred to as "escape" here)
• Volume down/Down
• Volume up/Up
• Begin Test/Enter

Direct switches are not part of the WPC switch matrix. All of the direct switches are located on the coin door, and connect to the MPU at J205. The MPU senses these switches individually, and apart from the switch matrix. Therefore, isolation diodes are not used with direct switches.

Normally, with the switch open, the LM339s at U16 and U17 compare 12V (supplied on the MPU to both the switch and to the LM339) with 5V (as a comparison level) and signals the 74LS240 at U15 that the switch is open. When the switch closes, it shorts the 12V to ground and the comparison at the LM339 then indicates to U15 that the switch is closed. U15 is "clocked" by pin 48 of the ASIC (SW DIR), causing U15 to present its data to the data bus. The 6809/ASIC "debounces" the switch. Debouncing is not a factor to be considered here and doesn't factor in switch testing at all.

Direct Switch Pinout at J205 for both WPC and WPC-S MPUs
Notes:

J205, pin 5 is the "key" pin.

J205, pin 10 provides ground (black wire)

J205, pin 11 is unused

J205, pin 12 is an "enable" back to the coin door interface board

Signal	Wire color at J205	MPU pin	Diode	LM339 & input pin	U15 input pin
Left chute	orange/brown	J205-1	D15	U17-5	11
Center chute	orange/red	J205-2	D16	U17-7	13
Right chute	orange/black	J205-3	D17	U17-11	15
4th chute	orange/yellow	J205-4	D18	U17-9	17
Escape	orange/green	J205-6	D11	U16-9	2
Down	orange/blue	J205-7	D12	U16-11	4
Up	orange/violet	J205-8	D13	U16-7	6
Enter	orange/gray	J205-9	D14	U16-5	8

Debugging direct switch problems

To discover the problem as quickly as possible, we'll divide the problem into smaller pieces. We must first determine if the problem is on the MPU board or the problem lies with the wiring to the coin door and/or the specific switch. If your MPU has obvious alkaline damage at U15, U16, or U17, address the alkaline damage first.

Begin testing with the game OFF.

Isolate the problem to the MPU or to the game wiring/switch

1. Remove connector J205 from the MPU.
2. Build a jumper wire with alligator clips on both ends.
3. Clip one end of the jumper wire to the game's ground braid in the game head.
4. Turn the game on
5. Carefully touch the other end of your jumper wire to the appropriate pin on J205 as shown in the table.
6. If the MPU carries out the function of the switch, the problem is not on the MPU.
7. If the MPU does nothing, the problem is on the MPU.
8. Skip to the appropriate section below.

Identifying problems NOT on the MPU

Test the direct switch's path to ground

1. Your game should still be turned off.
2. DMM set to continuity.
3. Clip the black lead of your DMM to any game ground, like the lockdown bar or ground braid.
4. Red lead on the solder joint between the switch and the black wire that provides ground.
5. You should hear "tone". If not, further diagnose the break in the black wire between the solder joint and game ground.

Test the direct switch itself

1. Black lead of your DMM still clipped to game ground.
2. Red lead on the solder joint opposite the black wire (or bare wire jumper) of the switch under test.
3. Depress the switch. You should hear "tone". If not, the switch is defective and not "making". See the section below that describes cleaning these switches.

Test the signal path to the MPU

1. Clip the black lead of your DMM on the solder joint opposite the black wire (or bare wire jumper) for the switch under test.
2. Remove the connector plug at J205 from the MPU.
3. Red lead on the appropriate pin of J205 for the switch under test. See the table above. It's easiest to access the pin through the rectangular hole in the back of the connector where the pin's "tang" snaps in.
4. You should hear "tone". If not, there is a discontinuity in the wire between the direct switch and J205. Note that the coin door interface board is between these two points. The coin door interface board is a "pass-through" for these signals and rarely causes a problem. Still, reseating connectors J1, J3, and J4 on the coin door interface board might uncover the problem. More likely, the wire between J1 on the coin door interface board and J205 has a break in it.

Identifying problems on the MPU

Test the signal path through J205 and onto the MPU
J205 is right below the battery holder and as such, sometimes receives the unwanted gift of dripping alkaline from depleted batteries. Carefully examine both the male and female connections of J205 for alkaline "greenies". Assuming no alkaline damage...

1. Begin with your game turned off.
2. Connect J205 to the MPU.
3. Clip the black lead of your DMM to the solder joint opposite the black wire (or bare wire jumper) for the switch under test.
4. Red lead on the banded end of the appropriate diode shown in the table above.
5. You should hear "tone". If not, there is either a problem with the physical connection at J205 or the alkaline "greenies" are sneakier than you gave them credit for. Re-examine J205 and the surrounding area of the board for alkaline damage. The traces from the switch connectors are very small and it takes very little alkaline damage to compromise them. You may also re-pin the female side of J205.

At this point, a logic probe would be the best tool to use. You can pick up 5V power for your probe across the electrolytic cap at C31 which is immediately to the right of the battery holder. Black lead on the negative side (top of the cap). Red lead on positive side (bottom). The board is silkscreened with polarity markings. Set the logic probe to "CMOS" test mode, as you will be measuring 12VDC.

If you don't already have a logic probe, you should. Although for this test, you can still get by with your trusty DMM. Clip the black lead of your DMM to game ground. The ground braid in the head is a good place to pickup ground at this point. Set your DMM to DC volts.

Test the LM339 inputs

1. Start with the game turned off.
2. Again, set your logic probe to "CMOS"
3. Clip one end of your jumper wire to the appopriate pin of J205. Leave the other end of your jumper wire unconnected, but handy as you'll touch it to ground later.
4. You now need to turn the game ON.
5. Measure the signal at the appropriate LM339's appropriate pin shown in the table above. Either place your logic probe on the pin or place the Red lead of your DMM on the pin. The signal should measure high (or about 12VDC with your DMM).
6. Now, with your other hand, touch the free end of your jumper wire to the ground braid in the backbox as you observe the results.
7. You should see the signal transition to low. If you still measure high, then the pin isn't being grounded correctly. Candidates are a failed diode/resistor in the circuit, the board trace between the diode and the LM339 is compromised, or a badly failed LM339.

Test the LM339 outputs/74LS240 (U15) inputs

1. Set your logic probe back to TTL as you will be measuring 0 - 5V signals
2. Measure the signal at the appropriate pin (see table) of U15. The signal should measure high (or about 5VDC with your DMM).
3. Touch the free end of your jumper wire to the ground braid in the backbox as you observe the results.
4. You should see the signal transition to low. If you still measure high with the jumper wire grounded, then either the LM339 outputs have failed, the board trace between the LM339 and U15 is compromised, or the 74LS240 has failed and is corrupting the signal. You can test U15 using the procedure in the "How to..." section of PinWiki, here.

The output side of U15 can't be tested effectively since that is the processor/ASIC data bus and should be constantly and irregularly changing states.

If you've followed this process step-by-step, you should have identified the problem with the signal and will be able to effectively perform the appropriate repair.

Cleaning direct switches

Coming soon...will describe disassembly and cleaning... Sometimes, several rigorous open/close cycles will "clean" corrosion from the switch.

4.10.2 Switch Matrix Problems

Isolate the problem to the MPU or to the game wiring/diodes/switches

Follow these steps to determine if the switch matrix problem is on the CPU or somewhere in the game wiring.

1. Remove connectors J206/J207, J208/209 and J212 from the bottom of the CPU (J206 and J207 are electronically the same and J208 and J209 are electronically the same).
2. Clip one lead of a test jumper to pin one of J207 making sure you don't touch any of the nearby pins on J207.
3. Power the game on, and enter Switch Edge test (you've left J205 on the CPU so the diagnostics switches still work).
4. Touch the other end of the jumper to pin one on J209, then pin two and so on. You should see (and hear) the CPU indicate that switches #11, #12, #13, etc. are being "made" as you touch the pins of J209. You should *never* see multiple switches being made.
5. Move the jumper from J207 pin 1 to pin 2. Re-test to every pin on J209, listening for the machine to read every switch as closed. Continue in a similar manner testing each pin of J207 to every pin in J209.

If only one switch shows on the Switch Edge test for each pin combination, then the CPU's switch matrix is working properly and you can confidently assume that any switch problems lie elsewhere. (Shorted wire, bad/broken/shorted diode, bad under playfield opto board, bad coin door interface board, etc.)

If multiple switch closures are reported when you touch one pin of the connector, or if the CPU reports a row or column short, then there is a fault with the switch matrix circuitry on the CPU board itself. Be aware that an additional fault may still be present on the playfield; switch matrix chips do not simply blow for no reason!

4.11 Display problems

Under Construction

Display problems are usually the result of failing dot matrix displays themselves, flakey ribbon cables or connectors, ribbon cables installed "one row off" (or even one column off), failing high voltage sections of the dot matrix controller board, and rarely logic IC problems on the dot matrix controller board.

Start by checking the easy things first...

1. Reseat the narrow ribbon cable between the DMD controller board and the DMD.
2. Reseat the wide ribbon cables from the MPU to the DMD controller.
3. Ensure each ribbon cable is correctly mated, and not "one row off" or even "one column off".

4.11.1 Display Panel Problems

The display panel itself may have problems. The glass component has a limited life. The continued electrical discharge releases gases inside the display, which eventually stop the pixels from glowing, and the dots will slowly fade over time. If the voltages are too high, the panel will show sparkles, clouds or ghosting.

4.11.2 Testing DMD Controller Power

Testing DMD Controller voltages at the display

After you've checked the easy things, measure the voltages supplied by the DMD controller to the DMD display panel. To test the DMD controller voltages, first open the backbox, and rest the DMD panel face down on the glass. The simplest place to test the voltages provided by the DMD controller 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. Williams designed the DMD controller to drive the the DMD display panel just a bit below its specified voltages, presumably to extend the life of the panel. You can find the specification of a panel here.

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. At some point in the production of DMD controllers, Williams changed the Anode and offset voltages from about -124VDC and -112VDC to about -112VDC and -100VDC. As long as you measure voltages close to these, and there is a 12VDC difference between the Anode and 12VDC offset voltages, your DMD controller is providing correct voltage levels.

Function	Pin	Spec	WPC	Notes
Anode	1	-110V	-112V	-124V in early DMD machines
12VDC Offset	2	-98V	-100V	12V less than anode voltage
Key	3
Ground	4
Ground	5
5VDC	6	5V	5V
12VDC	7	12V	12V
Cathode	8	75V	62V	Williams runs this voltage low

If the high voltages are off by more than a few volts, turn the game off, disconnect the power connector at the DMD controller, and test the voltages at the DMD controller male header pins. If the voltages return to WPC nominal voltages, the zener diodes on the DMD controller are probably OK but the resistors and transistors around them are suspect.

For the 62V and -112V circuits Williams used a relatively sophisticated regulator circuit. The problem is, the resistors and diodes dissipate quite a lot of heat. In a warm operating environment, this leads to burned components and circuit boards. The board traces are often weakened so repair in the area is tricky.

Here's a description of the function of the parts in the -112V circuit.

Q1 is an output buffer. It just follows the voltage set at the top of zener diode D2. Q2 and Q3 form a constant current 5mA driver for the diode. Q2 is a current control for D2. By pulling on the base of Q2, the zener current can be reduced to compensate for the diode leaking as it warms. Q3 is a current sensor for D2. Note the very low 120 ohm resistor. If D2 passes more than 5mA, Q3 will start to turn on, and reduce the current through the diode by pulling on the base of Q3. So the zener diode always operates at 5mA.

The transistors can be tested in circuit, just watch out for the 120 ohm resistors across the base and emitter of Q3.

The board is subject to transistor shorts, which can also blow out the resistors. Open resistors are a common problem if Q2 and Q3 are bad.

The transistors can be replaced with generic parts, but they must have 150V ratings.

4.11.3 Schematic Diagram for WPC and WPC-95 DMD Controller Board

This link explains my experience repairing a Twilight Zone ('93 WPC) DMD Controller Board. Also repaired the actual display PCB, which was damaged from a from a faulty +62v circuit:

TZ DMD Controller Board Repair

Here’s a schematic with a list of the parts needed for WPC and WPC-95. Both board are extremely similar: WPC DMD Controller

Another area which will help with troubleshooting the DMD Controller Board is in the DE area (DE copied the Williams board designs):

Data East/Sega#PS_520-5047-01_-_High_Voltage_Missing_at_the_DMD_display

The following images of a fairly crispy DMD controller will assist you in ensuring that all traces are repaired.

• DMD Controller, High Voltage Section Parts removed
• 

  Front side traces. The yellow line shows where a trace had burned off of this board.
• 

  Rear side traces.

4.11.4 DMD Repair Warnings

Be careful when repairing the DMD Controller Boards, you are dealing with High Voltages when the boards are running. A good safety measure is to put one hand in a back pocket (or behind your back) when testing voltages.

It's also fairly easy to lift traces on the PCB or ruin the 'through holes' when removing components. If this happens, you may have to check continuity and 'stitch' the component hole with a thin wire strand. It's a good idea to practice on some useless PCB boards first, before taking on the DMD Controller Boards. For details on a stitch see the "How to Stitch" section of the Wiki.

If you do not have decent soldering and de-soldering skills, DMD HV work should be left to a professional.

HV Repair Kits and advice are available from: Ed at Great Plains Electronics

• WPC-95 HV REPAIR KIT $5.00
• WPC HV REPAIR KIT $5.50
• GPE kits do not contain capacitors, which can be ordered separately as needed.

HV Repair Kits and advice are also available from: Rob Anthony at Pinball Classics

• WPC-95 or WPC HV REPAIR KIT $10
• Indicate in the notes section of the order if you need the WPC or the WPC-95 HV kit, Pinball Classics kits do contain the 150uF 160V capacitors.

Bally/Williams and Data East have similarities with the 128x32 display and the display itself is interchangeable between manufacturers. DMDs have a limited lifetime and will eventually outgas. Please see the information about outgassing displays in the Data East section.

4.11.5 Failed RAM

Most failures of the DMD controller that are not attributed to the high voltage section are caused by failed 6264 static RAM at U24. Since the RAM is very susceptible to electrostatic discharge, they are often damaged by improper handling. Sometimes, they just fail. Failure of this RAM manifests in many ways, including some shown below.

• Various display issues resulting from failed 6264 RAM
• 

  Example 1
• 

  Example 2
• 

  Example 3
• 

  Example 4

4.12 Sound problems

4.12.1 Lowering the Minimum Volume

By default the game volume can be set no lower than '8' to prevent an operator from setting the volume too low. This can be overridden by entering the adjustments menu and setting Adjustment 1.28 (Minimum Volume Override) to yes. You can now set the volume as low as you like.

4.12.2 Unbalanced Sound with pre-DCS sound boards

Pre-DCS sound boards are not 100% interchangeable across games. Games from Funhouse to Party Zone (including Funhouse, Harley Davidson, Bride of Pinbot, Slugfest, Gilligan's Island, Terminator 2, Hurricane, and Party Zone) use different resistor values at R23/R24 than the remaining games that used the pre-DCS sound board (games from Hot Shots basketball through The Addams Family Gold). If you find the sound "unbalanced" between background sounds and voices, for instance, check the following resistor values.

R23/R24 values for...

Funhouse through Party Zone - 150K ohms

Hot Shots through TAFG - 56K ohms

Note that the resistor value indicated in the schematics for pre-DCS sound boards at R22 and R25 are wrong. The schematics indicate 150K ohms. The correct value is 120K ohms.

4.12.3 Distorted or slightly "off" sounds

The pre-DCS sound uses the capacitors as filters in the sound circuit; C15, C36, and C38. If one or more of these caps fail, certain frequencies may not sound as loud or as clear. The game will just sound "off" with perhaps slightly imbalanced sounds.

4.12.4 Wrong sounds being played

Generally, this problem boils down to a communication problem between the MPU and the sound board. Possible failure points are:

• Ribbon cables
• Male header pins on both MPU and sound board
• 74LS374 data buffer ICs used to communicate between the MPU and the sound board.
• Rarely, a failed 6264 RAM can cause this. The RAM is "tested" by the MPU. Sometimes it will pass the test, yet fail in game play. Odd...I know.

4.13 Flipper Problems

Note: This section mostly in outline form now...more beef to be added later...Chris Hibler
In most instances, an inexperienced tech will immediately blame the flipper coil itself for weak or non-functioning flippers. It is logical to blame the coil for the source of the problem. The flipper coil is, after all, the end reason why a flipper becomes engaged. However, its failure is the least probable reason that a flipper is not working.

4.13.1 Flipper does not work

In this case, you've started a game, you press the flipper cabinet switch, and nothing happens.

The first thing to do is inspect the wire connections to the flipper coil lugs. The most common reason a WPC flipper does not function is that either one of the wires broke from the game to the lug, or one of the coil winding wires, has broken away from the coil lug and is no longer connected to the coil. This is a pretty simple fix. Make a clean cut to the wire, strip or sand 1/4" to 3/8" of insulation back, and make a nice solder joint on the coil.

A good reason to nip spade connectors off and solder directly to the coil lug. This spade connector simply fell off the coil lug. Luckily, it didn't come to rest on an adjacent lug.

Some WPC-95 titles used "spade connectors" to attach wires to the coil lugs. This was done for ease of manufacture and for that reason only. Similar to a wire breaking free, these spade connectors sometimes do not provide a solid connection and sometimes simply fall off. In that case, cut the spade connector off and solder the wire directly to the coil lug. This makes a much better connection that is also much more reliable. Should a spade connector fall onto an adjacent coil lug, it's very possible that a short circuit will be created, possibly damaging the FlipTronics board.

4.13.1.1 Mechanical Reasons

There are a couple of mechanical reasons a flipper may not be working, including:

• Broken flipper link
• Flipper crank not "biting" flipper shaft (this would most likely manifest as a flipper that continues to move backwards, out of adjustment)

Simple inspection of the flipper mechanism should guide your corrective actions.

4.13.1.2 Pre-FlipTronics Games Electrical Reasons

• Failed flipper power supply - make sure that DC flipper power is present at the coil using your DMM. Place the red lead on a coil lug, the black lead on game ground (siderail or ground braid anywhere in the game). You should read about 70VDC.
• Flipper cabinet switch not "making" or fouled so badly that it doesn't conduct flipper power
• Fouled or missing flipper end-of-stroke switch. Pre-FlipTronics games relied on EOS switches in the same way that older games do. That is, power is routed to the power stroke winding of the coil via the EOS switch. If the EOS switch is fouled, only the hold winding will receive power; perhaps inadequate to move the flipper at all.

4.13.1.3 FlipTronics Games Electrical Reasons

• Failed flipper power supply. Flipper power originates at the transformer. It is routed through the power/driver board (which does nothing to it) and then to the FlipTronics board. The FlipTronics board rectifies the AC power via an onboard bridge rectifier. Power may be interrupted anywhere along this path, including a failed bridge rectifier or blown fuse.
• Failed flipper coil "power stroke" winding
• Flipper cabinet switch not "making" or flipper opto board not working
• Damaged connection between flipper cabinet switch and FlipTronics board
• FlipTronics board not recognizing flipper cabinet button switch closure

A missing pin in the FlipTonics-to-MPU interface. The missing pin prevented transmission of the F8 signal (D7), which is the upper left flipper switch closed signal.

An unusual reason for the flipper switch closure signal not being sent to the MPU is shown at left.

• Failed logic path on the FlipTronics board including the transistors. See: WPC FlipTronics I and II Boards for a description of the signal path.
• Improperly seated or failed ribbon connectors between MPU and FlipTronics board

4.13.2 Flipper is weak

In this case, the flipper makes a stroke, but it's a weak stroke, not striking the ball with normal force.

4.13.2.1 Mechanical Reasons

• Not enough vertical movement in flipper bat (crank tightened too close to bushing)
• Cracked or broken flipper bushing
• Mushroomed coil stop and plunger (replace coil sleeve too)
• Missing link sleeve or oval link "eyelet"
• Tired flipper crank (will not tighten on bat shaft)

4.13.2.2 Pre-FlipTronics Games Electrical Reasons

• Flipper cabinet switch not "making" or fouled so badly that it does not conduct
• Failed flipper end-of-stroke switch

4.13.2.3 FlipTronics Games Electrical Reasons

• Dirty flipper cabinet switch optos
• Power stroke winding of coil failed

4.13.3 Flipper locks on or fails to fall to resting position

In this case, the flipper locks in the up position, either at game power on or when the flipper cabinet switch is pressed.

4.13.3.4 Fodder (may be useful to complete this section)

• Fouled or broken "normally closed" EOS switch
• Fouled or broken flipper cabinet switch
• Flipper relay not enabled
• Fouled / dirty or non-functioning flipper opto switches
• Upper flipper switch closures in switch test, when there aren't any upper flippers on game