|
LAST UPDATE: 7/6/01- typos fixed. | ||||||||||||||||||||||||||||||||||||
| SYSTEM 80 PINBALL CPU TEST NOTES. Use these tips to help isolate a CPU problem of almost any system 80 game. Just got started, so come back for more updates! Did you do the startup guide to see where your CPU has problems? GO HERE |
|||||||||||||||||||||||||||||||||||||
| Before you start working on any pinball games, go here: CLAY'S PINBALL REPAIR GUIDES These repair guides are the foremost complete guides on the internet. I am only adding some of my tip here to help identify problems that are not covered by these guides directly. That being said, let's get down to the boring technical stuff!! |
|||||||||||||||||||||||||||||||||||||
| GOTTLIEB SYSTEM 80 (ANY 80, 80A, OR 80B) CPU TROUBLESHOOTING. Let's start with the basics. System 80 had 4 different CPU types: Early type system 80 with no plated through ground hole (see below) that took two game ROMS in Prom 1 and Prom 2 sockets. These boards were used in Spiderman (#653) to Star Race (#657). These boards can be converted to use in later games Later type system 80 with the plated through ground hole (shown below) that took one game ROM in Prom 1 socket. These boards were used in James Bond (#658) to Haunted House (#669). System 80A was a revised board that used better book keeping and 7 digit displays. These boards were the same as system 80 except that U2 and U3 are now different. System 80B had many changes. None were board design related, but the use of chips was reduced. The game ROM was bumped up to 2, but the startup ROM U2 and U3 were removed wrapped into the game ROM. System 80B is shown below with the daughter board over U2 and U3 and lack of display decoding chips. All the display data was decoded at the 20 digit alpha display. It would take much work to make a system 80 or 80B work in either place. The reset board is not required, and can be used in any system 80 game. |
|||||||||||||||||||||||||||||||||||||
![]() |
|||||||||||||||||||||||||||||||||||||
| GOTTLIEB SYSTEM 80 TROUBLESHOOTING These guides will best be approached by situations that I have seen and tried to fix. Most can cover many CPU problems that are encountered. EMAIL if you would like to add a situation or none of the situations cover your problem. |
|||||||||||||||||||||||||||||||||||||
| Only Power and displays connected (J1, J2, and J3 if you have one), but nothing happens on power up except general lighting on the playfield. This can be many different problems. The first thing to check is the proper power supply voltage to the CPU. Make sure you have a stable and clean +5 volts DC. This can be checked across capacitor C1 next to connector J1. Next thing to check is the display power. No display power means no way to see if the CPU is doing anything. Check fuses and schematics to find out where to measure the display power voltage. Get the easy stuff out of the way. See HERE for system 80 and 80A display voltage checks. The 60 and 42 volt is used for display power. Do you have the original CPU, or did you buy one and are trying to get it to work? This can be a problems since many system 80 CPUs suffered from acid death. Make sure you have the proper CPU (see above), proper game ROMs, and such. A quick way to tell if your CPU is original is to look on the board for the serial number. Serial numbers can be found on almost every board and every display not to mention the head and cabinet. A CPU with the same serial number as the cabinet is a pretty good sign that you have the original. More on this later. OK, you are sure you have display power, the right CPU, and the right game ROM, but the thing still doesn't have any displays. Maybe the game worked and it doesn't anymore. This is the hardest problem to solve in a document like this. I will soon have a CPU test page that will go over what you have to do to figure out what is going on, but now you are going to have to settle for "your CPU is dead-for now". GO HERE if you can wait that long. |
|||||||||||||||||||||||||||||||||||||
| Power and displays connected (J1, J2, and J3-if you have one), and the game powers up with displays working. The CPU doesn't act like the slam switch is closed when J5 is connected. Acts like J5 wasn't even connected. Things to check: Slam switch is not making contact or the signal is not getting to the board. Check and clean slam switch (on coin door near the lock). Check the connector J5 pin 10 (the left most one, hell, check them all) for good pin connection with no corrosion. This switch sends a ground signal to the board, so try putting a jumper on the slam switch (both blades of the switch) to a good ground and see if that cures it. Turn game off, remove the J5 connector, put a jumper from capacitor C1 (the negative side of it) to J5 pin 10. This will "ground" the slam input and simulate a closed slam switch. Turn game on and see if that helps. Suspect bad connectors, wires, etc...if this cures the problem. If this doesn't solve the problem, look into the following: Z26 (7404) pin 12 and 13 buffers the slam signal into U5 RIOT pin 15. These are the only chips that deal with this signal. A bad RIOT and you would probably see messed up displays too, so Z26 might be your bad chip. The way this works is the slam switch brings the gate of Z26 "low" or to ground and causes the output at pin 12 to go "high" or +5 volts since a 7404 chip is a simple inverter. This +5 volts goes into the RIOT U5 at pin 15. You can measure this voltage with a good voltmeter when the slam is closed. A bad Z26 and the output will never go high. A bad U5 and the CPU can never read the signal, but the displays would be messed up too (safe bet, but not always true). Don't rule out a bad coin door connector or a damaged harness at the coin door. |
|||||||||||||||||||||||||||||||||||||
| Power and displays connected (J1, J2, and J3 if you have one), and the game powers up with displays working. The slam switch works, but the coin or tilt switches don't work like they should. The playfield switches don't work either. This is a switch problem if not a connector problem. Check the connector pins at J5. Don't rule out a bad coin door connector or a damaged harness at the coin door. Go to the switch problems listed on it's own page HERE. |
|||||||||||||||||||||||||||||||||||||
| CPU is powered with (J1, J2, and J3 if you have one) and the displays are not showing proper zeros or are displaying weird characters. No matter what is connected, the displays are acting strange. This is a better sign that you think. At least the CPU is running. You could have display trouble other than what is going on (missing segments, segments stuck on, etc..), but getting the displays to show something useful is the first step. Try disconnecting all but one display (with power off) at a time if you have a system 80 or 80A game. This will help eliminate a bad display from freaking out the others. Replace the bad display if that is the problem (or go HERE to see how to fix it) Double check the connectors on the CPU. Are they corroded? missing pins? Fix if this is the case. Are you just missing digits or segments? Is the displays acting like they are trying to talk a different language? Keep reading. The problem doesn't go away when doing the above, so you probably have a CPU problem. I would first suspect U5 RIOT since this only controls displays and the slam switch. Can you swap U4,U5, or U6 with each other? Most of the time these are not socketed, but you might be lucky. I would then suspect the mass of buffer chips (system 80 and 80A only) on the right side of the board. There is a page dedicated to display problems. GO HERE. system 80B is a special case since the display itself decodes the data out of the CPU, but GO HERE anyway. |
|||||||||||||||||||||||||||||||||||||
| CPU is powered with (J1, J2, and J3 if you have one) and the displays are showing zeros in the credit display, or solid zeros in some player displays. These numbers are usually brighter than normal numbers when the game is working. Displays showing solid numbers mean the CPU is probably locked up. The brighter numbers indicate that the number is not "strobing" or being multiplexed by the CPU. Chances are the CPU is not starting up right. I have had these problems related to not jumping the board right when doing a conversion. The CPU is running, but the proper data is not going where it should. Is the board a new one you are trying out? Are you sure it is the right one? See top for the difference. Do you have any solder splats on the board from some work you did? Shorting out the address lines can cause this. If your game just started doing this, maybe the game ROM or U2 and U3 ROMs are bad. You will have to check the microprocessor to make sure it is running by probing the address lines with a scope. Game ROM could be toast too. More on this type of CPU trouble to come. |
|||||||||||||||||||||||||||||||||||||
| I go through the startup sequence as shown HERE, and get through the coin door test. The problem comes up when I try to get the switch connector (J6) working. Some switches don't work or multiple switches are showing up. Switches start at U4 RIOT and go through a 7404 inverter Z11 and Z12. The strobes leave the board and go to the switch matrix as shown in your manual. You do have a manual, don't you? Better get one. Test all switches and check off the number that is shown in the manual for every switch you test. Make sure each switch only shows one number. The "strobes" are sent out to the coin door (J5) for the tilt and coin switches along with the playfield (J6). Playfield strobes go through pins 1 through 8. When a switch closes, the stobe is sent through that switch and back to the CPU. The returns are pins 10 through 17. The strobe then goes through Z13 and Z14 gates which then feeds data back to the U4 RIOT. One missing switch may be just a dirty contact which can be cleaned with a business card. It may also be a bad isolation diode. The diodes keep each switch from being seen as another or whole row/column. A open diode will make the switch invisable. A shorted diode can make a whole row or column be seen by the CPU. Turn the game off and test each diode in the diode isolation blocks found all over the under side of the PF and cabinet. Diodes should test .5-.7V in one direction and nothing in the other with the meter set to diode test (you should know this by now- right?). A missing row can be a bad connector pin at J6 pin 10-17. This probably means the strobe is getting to the switch (strobes are columns, returns are rows), but the connector is not letting the strobe back to the board (return). A missing column could be a bad connector at J6 pin 1-8. This probably means the strobe is not leaving the CPU because of a connector problem. Switches not working at all might be U4 RIOT. When these go bad all sorts of switch problems pop up. Sometimes one gate of Z11,12,13, or 14 (and even Z15) can go bad causing problems. Don't forget some acid damage that might have got under the DIP switches. These can cause shorts which will really screw up the CPU switches. MORE ON SWITCH PROBLEMS HERE |
|||||||||||||||||||||||||||||||||||||
| I go through the startup sequence as shown HERE, and get through the coin door test, display tests, switch test, but end at the lamp/coil test. Lamp, coil, and sound problems are all related. You can break down problems from the start: Lamps, coils, and sound (L,C,&S) are screwed up, don't work, or don't work when they should. I assume you have ruled out the driver board for problems. This means that you tested the transistors according to Clay's guide and you think the driver board is OK. The U6 could be toast or partly working. Have another driver to try just in case? How about a CPU to driver connector? You are lucky if you do. COILS start at U6 RIOT and go to Z27 for coils and sound. The coils then go to Z28, Z29, and Z30. These chips can go bad and screw up some of the coil fires. I would suspect this if you have some coils out of sequence, but your sound is good. SOUND starts at U6 RIOT, go through Z27, then break off and go to Z31. From there the data goes through the CPU to driver connector, to Z13 on the driver, and then to the sound board. MORE ON SOUND HERE LAMPS start at U6 RIOT and go to Z32, 33, 34, and Z35. They then go to the driver board. I would suspect U6 if you have strange coil fires, lamps not sequencing properly, or strange sound. I have had U6 strobe lamps very strange and lock some lamps and coils on. Don't suspect U6 if sound and coils work, but not lamps. Don't rule it out either. Isolated gate chips other than U6 can very well cause L, C & S problems. Testing the data from these chips will be described later. Don't rule out bad CPU to driver connections. Don't rule out bad driver board problems. Figure out what is/is not working before diving into a bunch of chip replacements. |
|||||||||||||||||||||||||||||||||||||
| I go through the startup sequence as shown HERE, and get through the coin door test, display tests, switch test, lamp test, coil test, but I still don't have sound. Sound problems can be a pain to find. I have broken down most sound problems HERE. |
|||||||||||||||||||||||||||||||||||||
| I go through all the tests and everything seems to work fine. I play a game and everything is fine for a while and then the game resets. What gives? I would suspect that the capacitor in the bottom of the cabinet is getting tired. This often causes resets at random. See Clay's guide. After that, suspect loose connections from the power supply or possibly unstable voltages. These are more of a power supply issue and gone over HERE. Does the game reset during the same type of thing while playing? Strange resets can happen when you have a bad diode or missing diode on a coil. These diodes are there to help stop spikes from getting back to the electronics. A bad diode will always seem to reset when a kicker fires, but not always. Suspect any diode on any coil. Did you just do some board repairs? Suspect bad/loose solder joints, bad sockets, broken pins on chips, bent over pins on chips making partial contact, etc... Are your connectors good? These are always a pain, so don't rule them out. |
|||||||||||||||||||||||||||||||||||||
| I'm trying to set the DIP switch settings on my CPU, but the game isn't "seeing" them. Sometimes they work, sometimes they don't. Strange things happen- credit display goes out, sounds keep playing, credits are not right when coin is dropped, or features during a game don't work the same way. DIP switches might be to blame here. I have had boards with dirty dip switches that work and then don't. This will cause all kinds of problems depending on how they are set. The most troublesome are the red color rocker type switches used in the "Mars- God of War" era of boards, but could happen to any type or year. To test, take out an ohm meter and flip the board over on a bench. open all switches and measure across each switch. Do the same with all of them closed. Test all diodes below each DIP with the diode function of your meter. Each diode should measure 0.45 to 0.75 volts in one direction and none in the other. You can do both tests at once by closing all switches and measuring all diodes through the DIP switch. Do this with the dip switches open too. Compare all the readings to see if any are out of range. Most of the time it is a bad DIP or open diode. |
|||||||||||||||||||||||||||||||||||||
| I turn my game on and now the high score is 770,000. This never changes and the score to beat is different. None of the book keeping data makes sense. What is going on? The CPU uses a battery to keep the 5101 memory chip from forgetting what scores and basic book keeping info were the last time the game was turned off. This data is normally kept in check with the ni-cad battery on the board. You might want to do yourself a favor and get rid of this battery since it will kill your board sooner or later. Options for this are gone over by clay, but I prefer a 1mF memory cap. Clean looking and no wires hanging out for batteries. This will not leak undesirables on your board either. A loss of memory will cause the game to "default" to standard high scores and scores to beat. The book keeping data will be garbled and not useful. All signs of a bad battery or bad 5101. Often the HGTD is defaulted to 770,000 or 550,000. You can check the 5101 in the book keeping test #20. Sometimes the test brings back 5101 in the player 1 display, but the 5101 is not really bad. Often this chip is GOING bad, but will still provide the info you need. Hit the game start button and run the test a few more times to see if it goes away. I would replace the 5101 if your problems do not go away with a new battery (or cap). You can run without a 5101or battery, but you will never get to keep your high score. You want everyone to know that you scored 945,000 on black hole, don't you? Be careful, sometimes Gottlieb games reset the HSTD when your score goes over a particular value, but not over 1,000,000. Anything over 990,000 on "Mars-God of War", for example, will reset the score to 770,000. You can see why I stopped my big game (with 2 more balls left) on my MGOW in the pics of the backglass HERE. |
|||||||||||||||||||||||||||||||||||||
| Any other situations I can put here? EMAIL ME and I will add some. |
|||||||||||||||||||||||||||||||||||||
| TO THE REPAIR PAGE | |||||||||||||||||||||||||||||||||||||
| TO THE PINBALL PAGE | |||||||||||||||||||||||||||||||||||||