Repair of Gottlieb System 1 CPU Board
Pages proof-read and corrected by Barry Thompson, thanks!
There are a few good sites with technical information about this system 1, do not forget to visit them, here .... and here...
A great disadvantage of system 1 CPU board is that the main IC's forming the CPU, ROM, and output commands are completely obsolete. The board is built around the PPS4/1 Rockwell 4bit processor and his specialized chips. It, along with the Intel 4040 four bit CPU, was the very first CPU I encountered in the late 1970's. The Rockwell chips are the six spider like chips on the board. On this repair page I'll try to establish if these 6 chips are still ok. If they are, we try to control and repair the other regular chips. In general, the driver/buffer chips are the most vulnerable. The repair page is the last attempt to repair this board, before I decide to replace it with one of the new replacement boards available on the market. The Ni-wumph board from the U.S.A. or the PI-1 board of P.Janin from France.
Here you find a good evaluation of both replacement boards.....
Here the description and schematics of the system 1....
A photo of Gottlieb system 1 cpu board.
Most of these CPU boards are victims of the alkali from the leaking battery. We always remove the original battery and replace it with a separate battery holder connected at the board with longer wiring. If the board is corroded by the battery fluids, I STRONGLY recommend replacing the IC's at the bottom of the board. These are placed in the zone where the fluid was active and there is a good possibility that the chips are internally damaged by the alkali. The chips are Z6,Z7,Z8 and Z28. If any of these four IC's have signs of corrosion on the pins, REPLACE them!! Also, all runs that are damaged by the fluid will be checked to insure continuity in the runs. If you find any breaks, sand the bad run and solder a fine wire on it. ESPECIALLY control the runs at the point where they pass from the component side to the solder side. This point is occasionally broken and you have no continuity between the run at the component side and the solder side. Another thing to look out for is the orientation of the IC's. Some have pin 1 notch upwards and some have the notch downwards. If you have a clean board with the corroded chips replaced we are ready to connect it on the bench.
Mounting and connection.
As usual, I take the board out of the machine and connect it at a power source on the bench. The first thing we need is an edge connector. Use a JAMMA type connector as used in arcade video games. They are easy to find and cheap. This type of connector is double sided but we will use only one side. Cut off a piece with 6 pins, just what we need. The power source has to deliver +5 and -12 volts. Use an old P.C. power transformer which delivers these tensions. The connection is made on J1, +5 volts to pin 5 and 6, -12 to pin 2 and ground at pins 3 and 4.
Another connection that we have to make is grounding pin 3 of Z29. Solder a jumper between this pin 3 and pin 7 of the IC ( Z29). This connection replaces the normally closed " slam switch ". Control if the slam switch is considered "closed" by measuring at pin 4 of Z29, you have to find +5 volts there, if not replace Z29. Afterwards this jumper remains in place. This will prevent troubles with the meaningless slam switch.
The last preparation is connecting a test-led ( the + side) at Z16 pin 15. The test-led is normal led in series with a 470 ohm resistor. The - side of the test-led is connected to 0 volts. ( ground)
In the following controls I only describes what you have to measure. If you find something else, you have a first trace of a fault. You will have to work out the rest of the solution, using your knowledge of electronics. In most cases I give some hints or the direction to follow, but it was impossible to explain and trace all the circuitry in detail. So some basic knowledge of electronics will be needed in some cases.
The way we control if the board starts-up ok, is as follows.
If the CPU does not work at all, there will be no steering of the displays. When the displays of this Gottlieb are not " activated" by the CPU this results in the display of all zeros IMMEDIATLY. If they are steered by the CPU at start-up, there will be zeros, BUT only after about 5 seconds. This is totally different from Bally and Williams machines where the displays will stay completely " off" ( black) if the CPU does not start -up.
We have the test-led connected at the output of the display driver. The test led has to go "on" about 5 seconds AFTER we apply the tensions to the board. If that is the case go directly to the chapter " input buffers control". If the led goes "on" immediately after applying the tensions, continue here.
The board connected. Left at J1 the power connector. In the middle under the dip-switches , see the red jumper at Z29 between pins 3 and 7. The test-led connected at the right at pin15 of Z16.
Cut the power and connect a voltmeter at test connector TC1 , you find this white vertical nylon test connector on the left just under the middle of the board, Measure at pin 14, you have to find at power up first -12 volts rapidly changing into +5 vols ( after half a second) . What we are looking at is the RESET signal. If the reset is not working and does not change in +5 volts , first control or replace the active components of the reset circuitry, being Z2 and Q5 and Q6. If that does not help it is always possible the non-active components are bad , certainly both condensers C31 and C32. The transistors can be measured the chip ( Z2) is to be replaced.
The next thing we check are the clock signals. You need to find test connector TC2 ( the right white nylon connector in the middle of the board ) pins 11 and 12. Check there using a scope or a logical test lamp. There should be pulsing signals. Using a universal voltmeter you will find there +0,6 volts at pin 11, and - 0,3 volts at pin 12. If no clock pulses, you have a problem with the Rockwell U1 and the story ends here. (There is a little change the Xtal is bad , but that is not likely).
When all that is checked, we take a look to see if there is any activity on the Rockwell chips. To do so we look at test connector TC1, pins 1 to 12 . Using a scope or electronic test lamp we look for pulses at these pins. It is not necessary that you have activity on all pins. On a few is enough. If there are no pulses at any pin, we can stop here and state that there is a problem with the Rockwell chips. If you find pulses, we proceed controlling the input buffers.
Control input/output buffers.
In order to check these buffer chips, we will activate their inputs and look if there is a response at the outputs. The input buffers are, Z29,Z27,Z9 and Z8. Use a test-rod connected at 0 volts ( ground) to activate the input and control the response at the output pin using a logical test lamp or voltmeter.
Z29: activate pin 1 ( applying 0 volt) and control at pin 2 ( =output) = +5 volt. and so on still on Z29, activate pin 5 out = pin 6; activate pin 11 out = 10; activate pin 9 out = pin 8.
Z27: activate pin 1 out = pin 2; activate pin 3 out = pin 4; activate pin 5 out = pin 6; activate pin 9 out = pin 8, activate pin 11 out = pin 10, activate pin 13 out = pin 12.
Z9: the same pins as Z27.
Z28: Here only two outputs, activate pin 3 out = pin 4, activate pin 11 out =pin 10.
After these controls we are sure all input buffers are ok. If these outputs did not respond, of course the chip was bad, and needs to be replaced.
Concerning the outputs , the only ones we are sure that have to be " alive" in these circumstances ( board on the bench) and with all type of games are the ones of the switch matrix. Using a logical test lamp look at J7 pins 2,3,4,5,6 and 7. If there are pulses missing at one of these pins look at U5 pins 3,4,5,6,7 and 8. If you have the pulses there then Z8 is bad. If no pulses are there ( at U5), two possibilities .... First the bad Z8 put the signal coming from U5 to ground , so replace it. If there are no pulses after replacement, you have a special Rockwell chip that is faulty, and the story ends.
It is impossible to control all the outputs of U4 at pins 1,2,3,4,5,6,7 and 8 with the board on the bench. But if we control some of them, we can be pretty sure the troublesome U4 will work ok for all of them. We take advantage of the fact the machine plays a tune when the money-switch is activated. If we can simulate this command, the outputs to the chimes or sound board ( depending on what model machine we have) and coming from U4 will give some pulses. We proceed as follows:
Solder a short wire ( 2 inch) at pin 3 of J7 and with the other end touch for a brief moment pin 12 of J7. This simulates the switch matrix input from the coin mechanism. The result is that the outputs at pins 7, 8 and 10 at J5 will be activated. These outputs are normally connected to the chimes or sound board, the outputs coming from Rockwell chip U4, passing the buffers Z6 . We control the outputs using test led , or three at the same time . The test led(s) will light up in the rhythm of the sounds. If you use three at the same time you have a small light-organ.. Otherwise you do one at the time , starting every time by briefly touching pin l2 at J7. The single test-led is connected with the - side to ground and the + side to J5 Pin 7/8 or 10. You can connect all three pins using three test-leds.
ATTENTION !! Dip switch 23 must be set at "on" This is the 7 the dip switch of the most right block of 8 dip switches. This allows sound when putting in coins.
If test led does not light up look directly at pin 4,5 and 6 of U4 and follow the signal trough his buffer chip Z6 , pins 6,8 and 10. If there is no output at all at U4, again replace the buffer Z6 perhaps he holds the signal " low" . If Z6 is replaced and still no pulses, again end of story because of a bad U4 Rockwell chip.
Three test-leds connected at J5 pins 7,8 and 10. You can use one test-led and connect it at one output pin after another. The blue wire ( under left) to activate the switch-matrix.
More outputs to find are driving the displays. These are Z18,Z19,Z20 and Z21 for the display digits, others for the segments, divided into separated groups are, Z11,Z12,Z13,Z14,Z15,Z16 and Z17. As we know that at power up there are all 000's on the displays we know which ones have to be active. The outputs that are active to give us these zero's are to find at J2 an J3 and should be verified.
J2 pins 1 to 19 everywhere pulses except pins 8,9 and 10.
J3; pulses from pin 6 to pin 21.
If there are some missing signals , take out the schematic and replace the buffer chip where that signal is coming from.
Normally you should now have a working CPU board. If you place the board back in the machine, check the arriving tensions +5 and -12 on the CPU board at C16 and C17, at the two electrolytic condensers at the left of the board. If there are abnormal reactions of the machine, check the connectors. Mostly there are oxidate pins or pins that have lost their tension. You can replace the connectors by JAMMA type connectors, which you can adapt in length bij cutting them back or by gluing some together, as described in the pages " Repair of the Gottlieb driver board system 1". Also ground connections are to be controlled. Keep in mind the CPU board has only one pin to ground for the +5 and one for the -12 volts. The driver board only has one grounding pin ( J1 pin 22 ). These are weak points of the system 1. Take a look at the pages I link to at the beginning of this article. There are very good hints about these matters and how to solve the problems.