AUGUST 1997 TECH TIPS

This months tech tips will aim at the Bally -17 and the -35 MPU. This system is now reaching 20 years of age for the -17 MPU and at least 10 years for the -35 series. That's some serious usage for any electronic board. Now, add in some elements like rough repair work (plus the occasional battery acid damage), and by any standards, none of these boards should still be running.

When finding problems with a locked up MPU, a history on the MPU is always helpful. If it "was running and just died", that is a world away from trying to fix an unknown, locked up MPU, that has one or a dozen problems.

Give yourself the advantage of knowing that at least the jumpers match the type of Game ROMS being used (Roms, EProms, Masked Roms, etc.) and that no extra jumpers were left behind from previous changes. An accurate jumper chart is mandatory here and a basic knowledge of how to recognize the different types of IC's is also a plus.

Next thing to look for is a working U9 (6800 processor) and the 6821 (PIA) at U11. This, along with a known good U6, is ALL that is needed to get a rise (flicker) out of a working MPU. After repairing thousands of these boards I have learned that whenever I come across an unknown, locked-up MPU, I will put in my own U9, U6 and U11 and in some cases, re-jumper the board, to get a head start on finding the problems.

If I still am faced with a locked up board, it is time to get out the logic probe and VOM for some in close measurements. Next month I will continue this lengthy check list for troubleshooting a "LED always on" Bally MPU.


SEPTEMBER 1997 TECH TIP

This continuing series on the Bally -17 and -35 MPU will focus a on deeper look into our "Led always on" Mpu. It is a really good time for purchasing that Radio Shack Logic Probe (22-303). There isn't a friendlier or more useful piece of test equipment around (for less than $20.00), that will give you this much help. Yes, you will have to get familiar with the signals on a WORKING MPU before you can identify a bad one but this probe is a safe unit to use and will not damage your board.

The only caution at this point is that, when probing with this or any tester, it is possible to hit two pins/ traces at the same time, which will probably derail the Processor. I make it a frequent practice to restart any board under test to give the Processor a chance to restart. Even on a 'dead' MPU, the U9 (6800) can be running. Now it is time to look for some key signals that should lead you to the problem(s).

We left our MPU last time with 3 major (known good) IC's in place. The U9 (6800), the Ull (6821) and the U6 Game (utility) Rom. The jumpers have been checked against the types of Game Roms being used and all extra jumpers removed. (You could spend the next year trying to get an MPU to run without the correct jumpers)! This is now a bare bones MPU that should have given that familiar flicker when turned on. But , all this LED does is to stay stuck on.

With all things considered, the Bally Electronic Pinball Games Repair Procedure (FO 560-X) is a very helpful book, when you consider that it was written before any games were sold or were being operated. They probably had to put some problems into the MPU's to be able to write all that good info and then wait to see how close they came to the real world. Actually, the only page that we have ever used is the one that looks at Symptom 1C: LED "on" continuously. This turned out to be the most useful information while the rest of the book tended to lead to several distractions, always coming back to the infamous "BUS-ASSOCIATED FAILURE PROCEDURE", using the AID2A module. Over time I have tried this unit and have come to the conclusion that by the time you have hooked this thing up and started the long test sequence, I will have found the problem with a VOM or Logic Probe! Add in the fact that this AID2A won't work if the 6800 is dead, and the following info will make sense.

Start with the U9 at pin 3 and 36-37,(o1 and o2 clock circuit). Pin 3 should read 2.4VDC and pins 36-37 should read 2.6VDC. But ALSO the logic probe will show a Hi-Lo indicating a wave form. The voltage could be within range but without the pulse it's not a correct signal. Either the IC15 or the IC16 is probably bad if the voltage is wrong. Sometimes, it's one of the 470pf caps (Cl4-Cl5), but the most common failure is IC15. While were on the subject and you are about to rob one of these hard to find MC3459L's out of an old board, relax and find a 74LS37. It is a perfect cross match. The IC16 however, is always a 9602.

This is the time to discuss the installation of ANY IC on this MPU. If you don't use a high profile Machined pin socket under any chip you put into this board, you stand a chance of shortening the life of this MPU. This is a 'one time repair' board that will not stand up under repeated attempts at desoldering/soldering. It will fall apart easily if you are not careful. That is why a good socket, with exposed pins, that can be soldered on both sides of the board - is a MUST. The majority of boards (that we have to send back as unrepairable), have been ruined by sloppy work - not battery acid damage.

With this part of the lecture out of the way, we will get back to our locked up MPU. If both clock signals appear to be OK, then next is the RESET signal at Pin 40 of U9. This is about 5VDC and comes from the collector of the Q5 (2N4403). Because this transistor is located in the ACID ZONE, it might be a good idea to change it and its opposite, Q1 (2N4401). This is also true of any RUNNING MPU that doesn't always want to start on the first try. Without this signal to the 6800, at the right time, not much will happen. This is the end of part 2 of this extended series and we will continue with our "LED always ON" MPU next month.


OCTOBER 1997 TECH TIPS

Continuing with our "LED ALWAYS ON" Bally MPU series on a board with known good U9, U11 and U6. The jumpers have been verified and any acid damage has been corrected. When trouble shooting any Bally MPU, these are the absolute minimum requirements before starting.

The Reset line and both Clock Signals are within range and the next signal to check is the HALT line at U9 pin 2. This signal is held high thru R135 and is only connected to J5. If the 4-5 volts is missing then either the resistor is open or there is a short in the J5 connector. Look for some bent pins or possible corrosion under J5. This is a rare problem and the next signal is a better suspect.

That signal is the VMA which originates within the processor and goes out from U9 at pin 5 and goes into Pin 8 of U19. Experimenting with this signal has shown that the processor will start even if this signal is removed from the circuit at U19 but the 6800 will become unstable and keep resetting. If the signal is grounded as it goes into U19 then the LED locks on. Try lifting pin #8 (on U19) when the reading on test point 7 (VMA) is missing or out of regulation. The correct reading is around 2.8- 3 VDC.

Assuming that U19 is good, the next part of the VMA circuit combines the VMA signal with Memory Enable (U19 Pin 9) and exits out pin 10. This Memory Enable is tied high thru R15 and reads a steady logic 5 volts. This combo signal is inverted by U19 and then goes thru another Inverter at U14. It goes in U14 at Pin 11 (1.2 VDC) and comes out at Pin 12 (5VDC). Now it gets combined with a O’2 signal from IC 16 (9602) Pin 10 and goes into U15. This combined signal (3 VDC) exits at pin 6 and goes right back into U15 at both Pins 1 and 2. This buffers and inverts the signal and exits out U15 pin 3 -(1.8 VDC) to the system as the VUA-02. This VUA-02 is combined with other signals to become a chip select line going to both U2 and U6. This means that without a correct signal from U15 Pin 3, nothing much can happen. Time spent with this area is important and the voltages shown on the schematic are right on the mark.

The VMA line can be correct but it has to be combined and then exit U15 pin 3 with correct voltage for it to work. When looking at U15, consider that it is serving double duty as part of the Clock Buffers circuit as well as the VMA, so suspect this chip first, if voltages are off.

In case you missed this tip last month, the U15 can be crossed with a 74LS37. This makes changing this chip less of a big deal. Before, when the origional MOTO chip was hard to find, no one wanted to change this expensive chip until every thing else had been tried!

At this point you may have changed 2 or 3 chips and when you arrived at the next one, the MPU started. Don’t think for a minute that the other IC’s were OK and you were wasting your time changing them. Any time that you remove an IC on this MPU (and put down a Machined Pin socket), you have made future repairs/ troubleshooting better for everyone. If Bally had only used sockets under these common failure IC’s then thousands of these MPU’s would still be running.

Next month we will wrap up this series with the final circuits that could be involved in our "LED ALWAYS ON" Bally MPU.


NOVEMBER TECH TIPS

This will wrap up the series of articles on the "Led always on" problems with a Bally MPU. As previously stated, we have a clean MPU with jumpers that match the game ROMS, a known good U9 (6800) and a known good U11 (6821). The next signal that should be checked on a locked up MPU is the VUA-02. This signal originates at pin 3 of U15 and is gated with other signals to enable IC'S and as pull-ups to Data lines. It should measure 2.1 VDC at pin 3 of U15. If out of range or missing, it will have to be traced back thru U14 and U19 to find where it went south. Again, the most common failures here are U15 and U14, then U19.

We are now going to look at the U9 6800 Processor. The thing to look for here is what signal is missing or totally wrong. There are address signals that must be seen at pins 9 thru 24 as well as data signals that must be seen at pins 26 thru 33. Starting with the address signals- note the following:

1) The logic probe should show the presence of a distinctive HI/LO on each pin with the possible exceptions of the following. Pin 23 (Al3) and Pin 25 (Al5) are not used and will be floating. Pin 24 (A14) will be floating if board has not been jumpered for 2732 EPROMS. (These 2732 Game ROMS will use this A14 signal). But if A14 is missing or low, suspect the INVERTER U14 (it goes in at pin 3-out at pin 2).

2) Note here that Address lines AO thru A8 are held high thru Resistors R122 thru R130.

3) The A9, A11 and the A12 signals are sent thru various buffers and inverters so they might be pulsing low or missing altogether. If A9 is missing, look at U19 pin 13 or U18 pin 14. If A 11 is missing, look at U18 pin 3. If A12 is missing, look at U17 pin 13 or at U18 pin 11. Even if the signal ends there and does not go to another circuit (some signals are created for chip select lines and never used), a bad IC will swallow the address line and shut down the processor.

The quickest way to check if the IC is dragging down the address signal, is to clip and lift the pin at the input of the address line into the IC and then check at the Processor to see if the signal jumped back up.

Use some caution here, it is prudent to simply change an IC rather than do a pin lift and re-solder. Only someone that is comfortable with this procedure should be using it. I feel that the time and cost of removing and replacing the suspected IC is worth the effort. This board is really fragile and the cutting/ re-soldering might damage the traces near the chip. Even if this replaced IC doesn't cure the lock up, the benefit of having a machined pin socket under ANY IC is worth the effort.

If the Address Lines seem OK, then repeat the procedure with the Data Lines. The rules change a little because the Data Lines are not buffered or gated thru IC'S and go directly to all IC'S, U1 thru U11. They are tied to either logic 5volts (D0 thru D3) or to VUA02 (D4 thru D7). DO thru D3 go thru Resistors R114 -R117. D4 thru D7 go thru resistors R118 - R121 AND Diodes CR45 -CR48. I have seldom seen any of these components fail but it is usually a good idea to check out the diodes.

The last step in the search for missing signals (if the Processor seems to be OK), is to check EACH pin on the U6. If the 6800 is running, it has nothing to do, until the U6 gives it some instructions. If you find a dead or missing signal on any pin of U6, then the problem has been found. Trace this signal back to its origin and get it working. There is an (distant) outside chance that the LED itself is faulty, or the Q2 transistor that drives it may be stuck on.

For your information, out of the last 100 MPU'S sent to us for work, the MPU's repairs came out this way:

20% Were clean with only IC failure, 20% were clean but had previous (hack) work damage but were repairable, 50% were acid damaged and only half of those were repairable.

This is the end of this series and we will cover another system next month.


JANUARY 1998 BALLY TECH TIP

WITHOUT A DOUBT, THE MOST COMMON QUESTION THAT I GET IS ABOUT SWITCH PROBLEMS IN BALLY PINBALLS. RATHER THAN ANSWERING EACH ONE, I WILL TRY TO COVER THEM ALL IN THIS ARTICLE.

THIS INFORMATION WILL APPLY TO BALLY / STERNS AND THEIR SWITCH TEST PROGRAMS. OF THE BUILT-IN TEST PROGRAMS, THE SWITCH TEST IS THE MOST IMPORTANT BECAUSE IT CAN SHOW WHAT IS WORKING AND WHAT IS NOT. IN ADDITION, IT SOMETIMES WILL SHOW WRONG INFORMATION, WHICH CAN BE JUST AS HELPFUL.

I WILL BREAK IT DOWN INTO THREE SECTIONS: STUCK SWITCHES, SWITCH NOT WORKING, AND ALL OTHER SWITCH PROBLEMS.

1) STUCK SWITCHES - WHENEVER THE SWITCH TEST IS RUN, IT IS NECESSARY TO REMOVE ANY BALLS, RESET ANY TARGETS AND PUT SOME INSULATING PAPER IN ANY SPINNER SWITCH SO THAT THE PLAYFIELD CAN BE LEANED AGAINST THE BACKBOX. WITH THE PLAYFIELD IN THE UPRIGHT POSITION YOU WILL HAVE ACCESS TO ANY SWITCHES THAT ARE SHOWN IN TEST. THIS WILL REQUIRE A LITTLE HEAD STANDING TO READ THE DISPLAY IN THE BACK DOOR, BUT HAVING THE DOOR OPEN WILL MAKE THE JOB HARDER. NOW THAT YOU ARE ALL SET, PUT GAME IN SWITCH TEST AND LOOK FOR ANY NUMBER ON DISPLAY THAT SHOWS A STUCK SWITCH. HAVE THE GAME BOOK READY BECAUSE YOU MUST KNOW WHERE THAT NUMBERED SWITCH IS LOCATED. IF IT IS UNDER THE PLAYFIELD IT COULD SIMPLY BE A SWITCH THAT HAS CLOSED UP AND NEEDS ADJUSTING. IF IT'S A STAND UP TARGET OR ROLLOVER, THEN MAYBE THE CAPACITOR, USUALLY FOUND ON THESE UNITS, HAS FRIED SHUT. SIMPLY CLIP ONE END OF THE CAP AND SEE IF THE NUMBER GOES AWAY. REMEMBER THAT THE HIGHEST NUMBER IS ALWAYS DISPLAYED FIRST SO THERE COULD BE LOTS OF STUCK SWITCHES AND YOU ONLY ARE DEALING WITH THE TOP OF THE LIST.

FOR NOW, WE WILL DEAL WITH SWITCH AND WIRING HARNESS PROBLEMS AND SAVE THE MPU STUFF FOR LAST. AFTER THE STUCK SWITCHE(S) HAVE BEEN CLEARED IT ALWAYS A GOOD IDEA TO TEST THE REST OF THE SWITCHES. THERE MAY BE A PROBLEM THAT ISN'T SHOWING UP NOW. LIKE ONE OF THE FOLLOWING:

2) SWITCH NOT WORKING - THERE CAN BE A COUPLE OF REASONS FOR A MISSING SWITCH NUMBER. IF IT IS ONLY ONE, THAN IT COULD BE A BROKEN WIRE OR DIRTY SWITCH. IF IT IS MORE THAN ONE, IT COULD BE A BROKEN WIRE AT THE TOP OF THAT STRING OR EVEN A MISSING SIGNAL AT THE CONNECTOR, DUE TO BROKEN TERMINAL (INSIDE THAT CONNECTOR). ALL PLAYFIELD SWITCHES ARE RUN INTO MPU AT UPPER RIGHT CONNECTOR (J2). ALSO IN THE MIX IS THE POSSIBILITY OF A BAD OR BROKEN DIODE ON ONE SIDE OF THE SWITCH. THIS CAN BE DETERMINED BY JUMPING THE TWO SWITCH WIRES TOGETHER (BEFORE THE DIODE) AND IF THE SWITCH NUMBER SHOWS UP, THE DIODE WAS BAD. AT THIS POINT, LET'S DISCUSS THE REASON FOR DIODES ON SWITCHES.

TO KEEP THE NUMBER OF WIRES DOWN TO A MINIMUM, THE USE OF DIODES ON ONE SIDE OF THE SWITCH WILL ISOLATE THAT SWITCH FROM OTHERS IN THAT COLUMN. NOW, UP TO 48 SWITCHES ARE POSSIBLE WITH ONLY 14 WIRES. THE PROBLEM HERE IS THAT NOT ALL GAMES HAVE DIODES ON EVERY JUNCTION.

THE REASON IS, BALLY FELT THAT DIODES WERE NOT NEEDED ON COIN SWITCHES, BECAUSE THEY ARE BEING CLOSED BEFORE ANY OTHER (PLAYFIELD) SWITCH COULD BE CLOSED, SO IT WAS SAFE TO OMIT THEM. IF YOU LOOK UNDER THE PLAYFIELD OF SOME EARLY GAMES (FREEDOM - 1976), YOU WOULD BE SURPRISED TO SEE HOW MANY DIODES ARE MISSING FROM SWITCHES. BACK IN THE 70's, IT WAS THOUGHT THAT DIODES WERE NOT NEEDED ON ALL SWITCHES BECAUSE THE BALL COULD ONLY CLOSE ONE AT A TIME.

MOVE FORWARD TO THE MID-1980's, AND NOW WERE DOWN TO JUST A COUPLE OF COIN SWITCHES WITHOUT DIODES. WOULDN'T YOU KNOW IT, THE COIN SWITCH CAN STICK CLOSED AND SEND ALL KINDS OF WRONG INFO TO THE MPU!! BALLY FINALLY RECOGNIZED THE PROBLEM AND PUT DIODES ON THESE LAST 3 HOLDOUTS.

SWITCHES AND THE MPU

WHENEVER A SWITCH IS CLOSED, THAT INFORMATION IS SENT TO THE MPU THRU THE WIRE HARNESS TO J2 AND ON TO U10, THE 6821 PIA. THE SWITCH CLOSER BRINGS TOGETHER 2 SEPARATE OUTPUT SIGNALS FROM THIS U10 0. THIS INFO IS THEN SENT TO THE PROCESSOR WHERE THE GAMES PROGRAM ACTS ON THAT INFO. IF THE INFORMATION IS CORRECT, THE PROPER ACTION WILL TAKE PLACE ON THE PLAYFIELD. IF THE CLOSURE SHOWS A DIFFERENT (WRONG) NUMBER THAN THE WRONG ACTION WILL OCCUR. IT IS THAT SIMPLE! WRONG INFO IN MEANS WRONG INFO OUT.

THE WIRING HARNESS CAN SEND THE CORRECT INFORMATION BUT THE MPU CAN STEER THIS INFORMATION TO THE WRONG PLACE DUE TO MINUSCULE TRACES OF ACID CORROSION LEFT ON THE MPU FROM A LEAKY NICAD BATTERY. IN THIS ARTICLE, I WON'T GO INTO THE DAMAGE THAT THIS BATTERY CAN CAUSE, EXCEPT WHERE IT PERTAINS TO THE SWITCH MATRIX. WHAT CAN HAPPEN IS THIS - THE ACID GOES UNDER THE J3 CONNECTOR ON THE BOTTOM RIGHT AND BECAUSE PINS 2 AND 3 ARE SHARED STROBE SIGNALS FROM J2, THEY CAN BE SHORTED TO GROUND, OR TO EACH OTHER.

WHEN I BEAD BLAST THE CORROSION IN THIS AREA, I LIFT UP THE COVER OF J3 AND BLAST IN BETWEEN THE PINS. THE SHORT(S) CANNOT BE SEEN OR MEASURED WITH ANY KIND OF EQUIPMENT, SO REPEATED BLASTING / RUNNING SWITCH TEST, IS THE ONLY WAY TO BE SURE THAT SWITCH MATRIX IS CLEAN.

BEFORE ANY OF THIS INFORMATION CAN BE USED, WE MUST DETERMINE IF THE SWITCH PROBLEMS ARE PLAYFIELD RELATED OR MPU RELATED. THIS IS ACCOMPLISHED BY REMOVING THE J2 AND J3 CONNECTORS WHEN A STUCK SWITCH IS INDICATED, (BUT CAN'T BE FOUND). IF THE STUCK SWITCH IS STILL SHOWING, THE MPU IS THE PROBLEM. IF IT GOES AWAY, THE PLAYFIELD / WIRING / SWITCH DIODES CREATED THE PROBLEM.

FINALLY, THERE IS ANOTHER LOCATION WHERE THE SWITCH MATRIX CAN SHORT OR CROSS. LOOK AT THE DIAGRAM FOR THE (OPTION) SLIDE SWITCHES. I HAVE CIRCLED THE AREAS AROUND SWITCHES 1 THRU 3 AS AN EXAMPLE OF WHAT CAN GO WRONG (WITH ANY OF THE 32 SWITCHES). IN THE TOP CIRCLE ARE TWO OF THE SWITCH RETURN LINES (PBO - PB1). THESE LINES GO TO SLIDE SWITCHES S1 AND S2. IF BOTH SWITCHES WERE CLOSED, THEN THE RETURN LINES WILL BE CONNECTED TO THE STROBE LINE (PA5).

THINGS ARE OK TO THIS POINT BECAUSE THE DIODES ARE KEEPING THESE TWO CONNECTIONS SEPARATE. BUT IF ONE OR BOTH OF THE DIODES HAS FAILED, IT CAN ALLOW THE TWO RETURN LINES TO SHORT TOGETHER. THIS WILL CAUSE THE U10 PIA TO MISREAD INFO AND CREATE HAVOC. WHEN YOU SUSPECT THAT THESE SWITCHES ARE CAUSING SHORTS, SIMPLY SHUT THEM ALL OFF,(REMOVING THEM OUT OF THE CIRCUIT). THE DIODES CAN BE TESTED IN THE MPU WITH POWER OFF. NOTE ALSO, THAT THE SWITCHES CAN SHORT TOGETHER INTERNALLY IF THE METAL SWITCH PARTS ARE BROKEN. IF DONE VERY CAREFULLY, YOU CAN REMOVE THE COVERS TO DETERMINE WHICH SWITCH PACKAGE IS BAD.

THIS HAS BEEN THE MOST COMPREHENSIVE ARTICLE THAT I HAVE EVER WRITTEN ON THIS SUBJECT AND I HOPE IT WILL HELP YOU SOLVE YOUR SWITCH PROBLEMS. LATER ON, I WILL DO ONE ON THE WILLIAMS GAMES WHICH USE AN ENTIRELY DIFFERENT SYSTEM.

TOM CALLAHAN

THE REPAIR CONNECTION

repairc@tiac.net


TYPICAL BALLY SWITCH MATRIX

Ballyjan1.jpg (38758 bytes)

 

SLIDE SWITCH SCHEMATIC

Ballyjan2.jpg (34626 bytes)


 

 

NOVEMBER 1997 WILLIAMS TECH TIP

FIRST PART OF THIS 2 PART SERIES

This article will be our most extensive look at the Williams driver board that was in all games from 1975 to 1984. It is solely responsible for locking up more games than any other board in the system.

We will examine one of the least understood circuits on this driver board. That is the blanking circuit. It originates on the mpu from the 555 or the 556 timer ic depending on level of system. The real purpose behind this logic high is to enable / disable the solenoid and lamp sections for the game up / game over function.

This simple signal can cause major havoc throughout the system as well as throughout the game itself. Whenever a game is discovered with all of its coils burnt, than you can be sure that the blanking signal is responsible.

First off, the blanking signal itself can do no harm, it is when it is shorted to ground by a failed component and their associated chips that this otherwise useful blanking signal causes havoc. It can happen in many ways and I will attempt to describe the most common failures.

The most common section causing the largest damage is the solenoid section. That is because this section comes in contact with a large voltage supply with the transistors controlling 28 volts and the real possibility of this 28 volts getting back through the system.

When a coil is shorted under the playfield or in the backbox or cabinet, and if the transistor wasn’t opened by this direct short, the 28 volts can reach the driver board and take out any other part of the circuit that it chooses to. If the fuse is replaced before the coil has been replaced than this scenario repeats itself until the damage goes even deeper.

This will show up on the driver board as a rather large burnt hole that will spell the end of an otherwise working board. This voltage spike can travel back through the pre-drive transistor and even back to the ic that controls that particular circuit. The amount of damage caused by a spike is best seen when the entire system now locks up and all of the coils will pull in along with the lamps being shorted to ground through those smoking resistors. All of this was caused by a shorted coil / transistor but the poor blanking circuit has made everything worse.

It doesn’t always happen this way. Sometimes just the transistor opens and that’s it! But it is possible that it can go either way. When the opportunity is there, you can the sure that the worst that can happen, will happen. If the minimal damage has been done than the game will still run without that particular coil working. But if the worst case scenario has occurred than the game is DEAD!! You can simulate this by simply taking the blanking signal to ground before starting up the game. The mpu will not come up if the signal is grounded, (or shorted).

This blanking signal that is so important for the starting / operation of the mpu is routed throughout the driver board as the illustration shows. But, as with all complicated circuits, the system designers were not happy with just a plain old signal. They decided to make it more useful by combining it with another signal to control the special solenoid section and the flipper relay. This additional signal comes from the IC 5 PIA chip. When all is well, then this signal will keep the flippers and 6 special solenoids from turning on when it is in "game over" mode.However, the rest of the driver board uses the blanking signal as it comes from the MPU. It travels to the right side of the driver board and goes to the two 7408 chips that control the lamp columns.

Next it continues down to the bottom of the board and heads towards the solenoid section on the bottom left. There it goes to the four 7408 chips that control the 16 transistors, that are in turn, controlled by the MPU. The blanking signal is needed here to shut off both the lamps and the 16 solenoids if the mpu should fail. But, it is usually the other way around.

Let’s look at what can happen if one of the solenoids should fail and take out the IC in that particular circuit. Let’s choose the solenoid number 21 that is connected to Q10 and goes back through IC9 and IC7. If the power surge reaches back through both IC’s then it could easily compromise the blanking signal that goes through a separate gate of IC7. This means that the blanking signal is now stuck low and the mpu will be shut down.


DECEMBER 1997 WILLIAMS TECH TIP

PART 2 OF A 2 PART SERIES
It is always a good idea to check out this IC7 first. Especially if either solenoid number 21 or 22 has failed. But it can get much more complicated. Suppose that this power surge has worked its way back through IC6 or IC7 to one of the PIA’s that are part of this circuit. That means one of the three PIA’s has now got a failed output signal. This output is used only when testing these 6 special solenoids. It is not used during normal game play and would not be missed until the solenoid test is run. Going one step farther, the random spike could have taken out the entire PIA. If this occurs, then another kind of lockup will occur.

Whenever one of the PIA’s fail on the driver board, it will always lockup the MPU because they share address and data signals through the 40 pin connector that joins them together. Now things can get really complicated! How can you determine what has occurred on this driver board? There is a really simple test to determine if the PIA’s are involved in the collapse of the MPU. Looking at the diagram, there are two small sections of the blanking signal that have a "cut" notation next to the trace. This is where a fine cut can be made to separate the blanking line from the rest of the board. Two cuts are needed to cover both sections of the board. With the blanking signal now out of the circuit, if the PIA’s are OK then the MPU will come up after a short delay. This delay however is bad news for both lamps and coils. So it is advised that all connections are removed from driver board before turning on the system. This will preventive any overloads from not having the blanking signal present.

If the MPU is still locked up after severing the blanking signal then one of the pia’s is bad. All that has to be done now, is to determine which one of the three is the bad one. Obviously, the IC5 is the most likely, because we have been dealing with burnt coils and shorted transistors! But here is the problem with being that sure. If the solenoid in question is connected to one of the other pia’s, ( remember the solenoid test?), than it could be another pia that has been zapped. It would have been nice if all of these chips were still socketed. Now you have no choice but to start with IC5 an continue on across the driver board until you have found the bad PIA. After having found the bad unit, then you will be delighted to see the MPU come up again!

But after you have rejoined the cut traces on the blanking line, the MPU now locks up again! It seems as though there is more than one IC involved in this failure. Go back again and open up one side of the blanking line and see if the MPU will now come up. By separating the left and right sides of the line it will be easier to figure out which IC is holding down the blanking signal. This is strictly trial and error, however, it is possible to cut down the size of the area your going to search by making more cuts in the blanking trace that you are following. I am sure that by now you can appreciate why we charge $69.00 to repair this complicated board.

If it had been the lamp section that had failed, then it would have made sense to be looking at that part of the driver board and all of the IC’s that makeup this circuit. After having done hundreds of these boards I still get one in on occasion that has ALL OF THE PIA’S and MOST OF THE IC’S blown! This board was in a game that was fused over and over again, trying to get it to start. When this occurs, the problem can spread throughout the entire driver board. I have seen games with ALL OF THE COILS burnt to a crisp because someone had repeatedly replaced the blown 21/2 amp slow blow with a much larger fuse which allowed the board to fry what few parts that were not yet cooked enough!!

This should not be considered an invitation to go deep into this system. I really don’t want to see any driver boards in here for repair that show signs of previous repair attempts because they read this article. It really is a polite warning that only a pro should attempt repairs on this complicated driver board. As it is now, most of these boards that show up here have already been worked on many times over the years but that doesn’t mean that one more attempt won’t hurt. Finally, let’s look at the two smaller diagrams on the page. The so-called game up / game over circuit is there for the special solenoid section has we have stated. Sometimes we get a call from someone who’s game is always alive when in game over mode. This can occur if the blanking circuit gets stuck inside either IC7 or IC9. Another sign of a failure in this area is the "flippers always on" symptom that can occur when this stuck signal reaches IC8 which controls the turning on of the flipper relay. This IC8 can fail if some solenoid it is connected to sends a spike back through the circuit. Finally, there is no substitute for clean precise workmanship. If you decide to work on this driver board yourself then at least use the best machined pin sockets that are available under ANY IC that is being replaced! This is a rugged board but it will someday be in short supply and only the properly serviced ones will survive. Good luck!

Tom Callahan

The Repair Connection


Return to tech tip of the month

Return to THE REPAIR CONNECTION home page


Copyright ©1997 THE REPAIR CONNECTION

Sent Pinball questions to: repairc@tiac.net

Send questions or comments about this web site to webmaster@repairconnection.com

Revised: February 02, 1998.

Designed and maintained exclusively for THE REPAIR CONNECTION by

All rights reserved.

Best viewed with