Leon Borre Bally Stern CPU test bench

From PinWiki
Revision as of 07:04, 24 April 2013 by Jimpal (talk | contribs) (Created page with "<font size="6" color="#0000FF">'''Complete test bench for Bally -35 -17 CPU boards.'''</font> Image:swash4.gif <font size="4" color="#0000FF">'''Introduction.'''</font> <f...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Complete test bench for Bally -35 -17 CPU boards.

Swash4.gif

Introduction.

The whole test bench is made in three separate parts. A power suply that will suit not only this test bench but also two that will be published later , a test bench for the solenoid driver board and one for the lamp driver board The CPU test bench itself is again made of different parts.I did do this as i think it will be easier to make smaller items one at a time and also because every one can assemble these parts into a cabinet or box that suits him best.

There is no PCB for the power supply, as several types of transformators can be used and will be available , of course you need the right voltages but the dimensions will sometimes be different and would not fit a pre-made PCB, the same goes for some diode bridges and electrolytic condensers we need in the power supply. Luckely the power supply is so simple a PCB is really not needed.

The third part is specialy for the switch test. A normal test bench will use 64 switches, this takes a lot of space on the test gear and is costly. We will use two 8 position rotary switches. With these 8x8 combinations we can also reach the 64 switcch positions. As we need this part only during switch test , we can disconnect this part from J2 and J3 , that is convienient to work if we are not doing the switch test. J2 ad J3 will be liberated.

To make the test bench you need to read the schematics and handle the soldering iron well.

For those who can or will not make the complete test box. You can use the new test eproms to test the U1 and U2 sockets without the test box , just like you used to with the " old" test eprom.

Uses for this test bench.

With this test bench, you will be able to test the CPU board as it were in the pinball machine. The final tests is the same diagnostic test as on the pinball machine. Display test, solenoid test , all lamp test and switch test. As ALL in- and outputs are tested , this will ensure 100% everything is working. To test the CPU, you will need to use 3 different test eproms. You may have to change jumpers to use these test eproms. However, if you do not want to change jumpers, simply burn a copy of the software onto a 2732 and a 2532. Use the 2732 if you have eprom's as game roms, and the 2532 if you have roms as game roms.

Swash1.gif

The power supply.

The power supply has to deliver 4 different voltages.

5 volts ... 12 volts... 190 volts and a pulsating voltage between 20 and 30 volts.

Only the 5 volts and the 12 volts will be regulated, using a classic 7805/7812 voltage regulator. The 5 volts is derived from the 12 volts. This way we need only one bridge rectifier and less capacitors .The 190 volt for the display is not regulated. This will cause the displays to be dimmer if there are more digits lit at once. This dimness is a small trade -off considering that with this design, we will only need a small transformator and a bridge rectifier and one condenser to create high voltage for the display. For safety reasons, we will not use direct 117 or 220 volts AC from the mains to obtain the 190 volts. See the schematic for details on how we will get the voltage using a small transformator with 2x 12 volts AC windings and a single diode. The result is an extremely simple power supply.

The power supply schematic.

Schempow.jpg

The voltage coming into the power supply is 12 volts AC 3 amps. I did not enter with 117/ 220 volts AC since this would have required a large transformator that would not fix in the power supply enclosure. Also , this way, everyone can use an old quartz lamp or car charger transformer that may lying around.

The 12 and 5 is straight forward , a bridge rectifier, ( handles 5 amps.) the condenser and then first the 12 volt regulator followed by a 5 volt regulator , every time followed by a capacitor. Cannot be more easy.

On the upper part of the schematic we find two ( small) transformators. These are indeed very small only 4,0VA and 1,6 VA , Both 220 to 2x 12 volts.

Coming off the 220 volts AC winding, we feed that into a bidge retifier that can handle 300mA and 300 volts. We follow this with a 220uF capacitor. The non-regulated voltage can rise as high as 230 volts which is to much for the display to use. Therefore, we regulate the voltage to 190 volts by using a zener diode in series with a 1Kohm resistor. On the other hand if all digits are used, the voltage will drop to around 165 volts. This will still work, it will just make the display dimmer. During normal display test, I found that all digits where still visible. It will fade slowly as the voltage is depleted from the 220uF capacitor. If you think you would like to use a larger transformator for the 190 volts, then you will need a 5 watt zener diode and 500 ohm resistor to regulate the power due to it rising even higher when the display is blank. I recommend using a smaller transformator as the advantages outweight the disadvantages.

To create the pulsating voltage as well as 12 volts AC, needed on the lamp driver board tester, we will use another transfo and tap off of the 220 volts from the first transformator. That way, it will work in all countries regardless if their main voltage of either 117 ot 220 volts AC. The pulsating voltage is generated with a single diode in series with the 24 volts. We use one winding to have the 12 volts AC.

All elements used in this power supply are very common and available anywhere.The 190volt zener diode can be several zener's in series as long as the added result is about 190 volt in total. The bridge rectifier for the 190 volt circuit must be able to handle 330 volts 300mA , the same for the 220uF capacitor he must have a work-tension of 400 volts.

The output 12 volt AC will not be used in this CPU test box. It is foreseen for the lampdriver board test box.

Material

Material.jpg

The capacitors are bigger than the tranformers. The small rectifiers are on the mounting board on wich we will assemble these pieces into a power supply. Everything needs to fit inside the black plastic box. The output will be fed to a molex 6 pin connector. Both 7812 and the 7805 regulators are mounted on a piece of aluminum.

The transformators are really small. The bigger first transformator is only 4,0 VA ( type, ERA E32TR4 made in Belgium) it will deliver about 0,5A. on the secondary winding. The second transformator for the display high voltage is only 1,6VA . ( ERA E24 TR2 )

Boxbuit.jpg

Here is the finished box. The fuse in the right corner and in the middle the on/off switch, next to on/off LED ( red arrow). I did not use the black power cord to connect with the 12 volts AC first transformator . This could be confusing so I used a couple of ( blue) female banana buss connectors to feed the power box with the 12 volt AC.

Powpriniefaf.jpg

here are both power supplies. The upper delivering 12, 5 and a pulsating 24 volts, coming from the green diode (at the red arrow). The lower is the 190 volts supply. I used two zener diodes in series to get 190 volt since I could find only zener diodes with a max of 180 volts. I added a 10 volts zener in series with the 180 volts zener.

Powniefaf.jpg

Both circuit boards in the power supply enclosure.

Boxaf.jpg

The power supply box finished. On top, the colors of the connector wires and their voltages.

Ultimate question; Why are you using a 12 volt regulator when the 12 volts is directly coming from the rectifier? I use this 12 volt regulator to lower the input voltage to the 5 volt regulator. That way there is less heat and I can then use a small aluminum heat sink plate.

Swash1.gif

The test box = test bench.

The whole test box ( bench) consists of one printed circuit . The connection points on the printed circuit are connected to the cables going the cpu. We have an extra cable from the display going directly to J1 and some wires into the test box. Photo's and drawings makes things and connections clear.

On the test box is push button wired to ground, the other side of the push button is connected to a wire ending at a small grip, this grip will be connected during test at pin 1 of J3.

Power is coming in by a 6 plug connector adapted to the one coming from the power box.

Pin.jpg

I could not find a suitable box for the test equipment. So I decided to build my own. It consists of a piece of wood, with a small box build onto the left side. Inside the box on the left, I mounted the display and the rotary switch wich are used to choose between 1 of the 4 players or the credit/match display output. A push button is also mounted here to launch the normal game tests. This box contains the printed circuit board, wich holds a bunch of LEDs. These LEDs are coming trought a piece of white plastic that I use for a cover. The dimensions of the bottom board is 50x39 cm. The small box build on top is 19cm large

Prischem.jpg

The control LED's and how the are connected ,on the same print is the "power on" LED and the" test-LED".

Printtek.jpg

Printje.jpg

The simple print.

Bestukt.jpg

All LED's in place. The "power on " LED is yellow and the "Test LED" orange. The red LED's are for the strobe/select outputs. The greens for the data /address outputs.

Swash1.gif

Construction of the test box.

Solspan.JPG

The power connector ( male) ends on a small board on wich it will be easy to solder the different 5 volt and ground connections. ( blue arrow)

Bak1.jpg

The test box made of wood , just a smal compartement ( box) that will house the print and switches. On top of the compartiment comes the white plastic.

Kasopen.jpg

This way the print and the switch the push button will be mounted inside the compartiment.On the right side is the place to put the CPU under test. Of course the display comes at the outside on top of the plastic.

Kastoe.jpg

Fnished and closed

Kastekst.jpg

The finishing touch.

Swash1.gif

The display

In the final solution, i did use the normal pinball display instead of the LED display I built in the prototype. This way, we can use the display connection to test other displays if we need to. For the same reason, I left the display out of the test box. It is more convenient to plug in another display for testing. This solution results in a connection cable that has half going to the CPU ( connector J1) and half going to the display ( J1). . A photo makes things clear.

Here the schematic how the special display cable connections are made.

Dispcab.jpg

Cabdisp.JPG

The wires that are not going directly from J1 display to ( red arrow) J1 CPU are the 190 volts, ground, 5 volts. The wire between pin 15 of the display connector goes to a 5-position rotary switch ( oragne arrow) in the test box. This is used to select between the 4 players and the credit/ball display outputs.

Swash1.gif

Solenoid control circuitry.

When passing a solenoid test , the number of the solenoid that is activated will be on display, but the real command to the solenoid is coming from U10 in a binary combination. This results in 8 outputs at J4. As the selections and decodings is done one the solenoid board, we will check here if the outputs are active during the test . The solenoid output LED's are connected as follows.

J4.jpg

Swash1.gif

Lamp control circuitry.

The controled lamps are again steered by a combination of address-lines and data-lines signals. Again we check if these output pins are active. The control LED's are connected this way.

J1.jpg

Swash1.gif

The switch circuitry.

Switch controle.

To perform the switch test we use a separate switch box ,simulating all the playfield switches. It is made of two rotary switches with 8 positions.

Switchen.jpg

There are no two J 2 connectors, this is the same connector , this way of representation gives a clear drawing.

This is what we need.

Switchmat.jpg

If you are lucky like i was, you have some old cables that fit..Or make some using new connectors.

Swiklaar.jpg

Switches and cables mounted , ready for soldering.

Switvolaf.jpg

Switch box finished .

J3pin.jpg

At J3 pin 1 is a pin sticking out. We have to connect here a grip coming from the test box. See why and how in the chapter " Start Testing"

Swash4.gif

Start testing.

Here follows the description how to use the test box. The description gives us what happens if the board is ok. It is under FAULT HANDLING that you will find what to do if you discover something that goes wrong.

Keep in in mind that all the output signals come from U10 and U11. The output for the controlled lamps and the output for the switch matrix ( the strobes) plus the output of the display ALL are coming from the same pins of U10!! They are separated by timing and the display is mixed with the signal CA2 of U10. I did place LED's on all the output pins to check if the outputs are ok during the different game test. BUT when you use the basic test eprom ALL the outputs of U10 and U11 are activated, this way from the beginning we will know if they are all ok or not. So if you do the basic test it is not longer needed to pass the (game ), lamp or solenoid tests..the outputs are already checked. The switch test is always recommended as during the basic test all pins of U10 and U11 are tested as outputs and during the switch test some of these pins will be tested as inputs.

Swash1.gif

Connect the CPU board on the test box.

Put the CPU board on the test box. Connect J4 and J1 . Connect the BLACK grip coming from the push button of the test box to pin 1 of J3. Connect the RED grip from the test led to pin 15 of U9.

Swash1.gif

During next tests we use some test eproms . When used on a CPU board the type of eprom used is dependent of the excisting jumpering of the board. When on your board you have eproms use 2732's ( as test eproms) when on the board are masked roms use 2532's ( as test ep^roms) . If you do not wish to have two sets of test eproms it is of course possible to adapt the jumpers on the board and to use always the 2732 test eproms, but this means sometimes more soldering on the CPU board under test.

Test eprom number 1 ... [balmemtest.zip Pinaqua.gif]

The first test we run is with the test eprom described on my page " Renewed and simplified repair method for Bally /Stern CPU boards" Plug in the test eprom ( test eprom 1 ) in U6 and power up the test box. The led on the cpu board will blink. Using the test-led ( deconnect it from U9) we can control the outputs of both U10 and U11 chips from pin 2 to pin 17 pins 18,19,39 and 40. Touching these pins will blink the test-led . At the same time ALL the LED's ( EXCEPT J4/11) that are connected to the output pins have to blink. This proves that the output of U10 adn U11 is really coming on the connector pins, remember there is a resistor between the pins of U10/U11 and the pin on the connector. Afther this test push the push button on top of the cpu board , this will launch the memory test. Afther stopping a while the blinking of the CPU led has to restart, meaning the memory test is ok. If the test led stays on or off a memory chip is bad.This ends the basic tests.

Swash1.gif

Testing U1 and U2 sockets and their selection.

Here we use two new test eproms number 2, 3

Take out the test eprom 1 and replace it by test eprom 2 in U6 , also place test eprom 3 in socket U2. At power up the led on the cpu board will blink, together with the test led on the test box

The test used is the same as the primary test but is now located in the U2 eprom.. So the blinking CPU LED indicate that the eprom in socket 2 was found and did respond to the test program. If you have normally a game rom in socket U1 , we will test that socket to. Change the test eprom from U2 to U1 , power up again, and push the button on the CPU board. The blinking of the test led and the CPU led will restart. Meaning the test program found U1, indicating all went well and you can trust the contacts and the selection circuitry of both socket U1 and U2. Of course when no game rom is used in U1 this last test is useless.This ends the second test phase. If U1 or U2 are not found because one or more address lines or datalines cannot reach the eprom the CPU LED will not blink, the LED can be "on" or "off" meaning you have a problem in socket U2 or/and U1.

Images of the test eproms for ..U6.. [newu6.zip Pinaqua.gif] U2 / U1 .. [u1plusu2.zip Pinaqua.gif] Burn these in a 2732 and /or 2532.

Swash1.gif

Final testing using the game roms.

We change the test eprom again and now use the game roms. It is a good idea to have some " know good" game roms at hand, if you are not shure about the ones that came with the board under test.

At power up you have the normal behavior of the CPU board as in the pinball machine. Starting with 7 flashes of the led on the CPU board. The display will show alternating a 00 and the high score to date. The high score to date can be nothing ( blank display) due to the reset of the memory as we worked on the board. When we push now the red push button on the test box , the CPU board passes in test mode, starting with the lamp test, next push will lead us to the display test , next to the solenoid test and the last one to the switch test.

Swash1.gif

All lamp test

The LED' s on the test box representing the output pins to the lamp driver board ( J1 LEDs will be "on" from the moment you power up the board , this because the outputs are common with the switches. Anyway the lamp test is not needed , as the outputs are all tested during basic test.

Swash1.gif

Display test

Next push on the red button brings us to the display test, the display will cycle with all 111111 /222222/333333/ and so on, this on all digits, using the display rotary switch on the test box you can choose any player or the credit/match display outputs.

Swash1.gif

Solenoid test .

We push again the red button, bringing us to the solenoid test. The display wil cycle from 1 to 15 or from 1 to 30 depending on wich type of game roms you are using; The games with no connector at J5 will in general cycle to 15, the games with a more elaborated soundboard connected at J5 will cycle to 30. The LED's at the output pins J1 LEDs ,will be on or blink , some can blink very slowly if you have a board cycling from 01 to 30. We do not need to spend time here as the outputs are all tested during the basic test.

Swash1.gif

The switch test.

To execute the switch test . Leave the game roms in place and connect the switch test box . Plug in J2 and J3 . With the pin sticking out at pin 1 J3 it is still possible to hook on the connection ( black grip ) coming from the test box. After power up , push the red push button of the test box 4 times , you will pass lamp test ,display test and solenoid test to come at last to switch test. Place rotary switch 1 in the second position and scan all other positions of rotary switch 2. The display will indicate the switch numbers from 1 to 8, then change rotary switch 1 to the next position , again scan all the positions of rotary switch 2, now you wil simulate switches 9 to 17. Continue all other positions of rotary switch 1 and scan all other switches. If all switches are tested you can remove the switch box. The number of the switch circuit under test will be displayed. Be shure that at position one of rotary switch 1 the display stay's blank.

Swash4.gif

Fault handling.

Swash1.gif

Start with the basic test-eprom on the CPU board. If you have any doubts about the jumper setting control that first !!!

To run the basic test all we need is the power on the cpu board J4 connected, and the test led connected at pin 15 of the CPU chip U9. If the test LED blinks the test is running.That is for the moment the ONLY result we are looking for.

The basic test eprom uses a minimum of the possibility's of the CPU board to run . Of course that is still a lot of the circuitry on the board.. The functions that are needed to run the basic test are, a good working CPU chip, the selection of U6 , the reset circuitry, not a lot at first glimp but the good working of the CPU IC can be altered by circuits and IC's that are not under test but disturbing the address and/or databus lines or selection lines. That is the reason we will take out from the board all what's possible . Start with taking out the both memory chips U7 and U8 also both PIA chips U10 and U11.With the test eprom in place power up. If you have a blinking test led go to " Test running". If you no blinking test led , first measure the signals at the U9 CPU chip.

U9

Start at pin 40 you need 5 volts there, if ok ground this pin for a brief moment, this simulates a reset, if the reset circuit is bad this will start the test. If the test starts now go to the reset circuit here below. If you did not have 5 volts at pin 40 from the start , go to reset circuit here below. If you did have 5 and afther grounding the test sill is not started go to " U9 two"

Reset cirquit.

1) The test run after grounding pin 40.

This means everything is ok up to now except the reset cicuit. This reset circuit consists of transistors Q1, Q5 and resistors around it, start measuring if you have 0,6 volts at the base of Q1 and 4 volts at the base of Q5 this will tell you if the resistor circuit ahead the base of Q1 is ok or not , if at the base of Q5 it is wrong that can be Q1 faulty if both ok and no 5 at the collector of Q5 then Q5 is bad.

2 )No 5 from the start at pin 40.

This means that you have no 5 at the collector of Q5. Replace Q5 still not ok measure the base of Q5. If the base tension is not 4 volt , replace Q1, still wrong . Must be one of the resistors in the circuit measure them.

U9 two.

The test is not working but perhaps the CPU chip is ok but "running" without control of the progam. We can see that directly by measuring pin 5 ( U9) if you have 2,5 volt pulses here the cpu is running and U9 is ok. If that is the case go to "U9 running " If you have no pulses at pin 5 continue here below.

The tension is 5 at pin 40 that is ok, now we measure pin 3 and 37 here are the clock signals; best is to use a logic probe. The clock are pulses about 2 volts.. If you have no clocks replace U16 first still not ok replace U15 in general it is U16 that dies first.. if you did replace both IC's and still no go, control the resistors of this clock circuitry..

Now we have pin 40, 3 and 37 ok. We look at pin 2 we need 5 volt there ,normally this can't go wrong the pin is via resistor R135 at 5 volts.The same for pin 4 IRQ also with a resistor R134 to 5 volts. This was the last pin to measure at U9, if all that is ok and still no pulses at pin 5 measure for safety between pin 1 and pin 8 to see of the power is coming to U9, you never know, at last measure pin 6 ,there to a steady 5 volts if not your push button on the cpu board is stuck! If all that is ok and no pulses at pin 5 , replace U9.

U9 running.

The cpu chip is running wild, meaning he does not follow the instructions contained in the test eprom at U6. The main reasons are, address lines or data lines disturbed, selection of U6 not functional.

Take out the test eprom and run the board without any eprom. Measure the address and data lines pins, address lines are pins 9 to 24 , data lines are pins 26 to 33. You have to find on all of them pulses , not one must be stuck high or low. If one of these is stuck , bend up the pin to measure, if , with the pin bended and that way not connected to anything ,you have pulses now. If no pulses replace U9 . If ok you have to find what causes the line stuck high or low. You can do that by cutting temporally some runs to disconnect parts of the IC's that are connected to that line and find out at last wich one is the troublemaker..

The last possible cause for a non running test is the selection and signals on the U6 test eprom. Reseat the test eprom in U6. Measure again with the logic probe on the U6 pins if you have pulses on all the data and address lines here and on the selection pins , pin 20 and pin 18. If an address or data line is missing suspect the socket , if a selection signal is missing or stuck suspect the socket or one of the chips handling the selection, follow with the schematic where the pins 20 and 18 ( depending on the jumpers) are connected and see if you find an IC with pulses at his input and no pulses at his output.

Normally you did find the trouble if you have followed these guidelines. Go to the next chapter " Test running".

Test running.

Put the test eprom back in U6 reseat the U10 and U11 chips. If the test do not run anymore one of the PIA's ( U10 or U11 ) is bad. If the test runs all control LED's will blink. If that is not the case chech the outputs of U10 and U11 directly at the output pins , 2 to 17and pins 19 39 all should go up and down between 0 and 5 volt . If a pin does not follow that rythm bend upward the pin that is stuck and measure again. Still bad replace that PIA. Ok ? Then you have a short on that output pin , follow the run to see what's connected and eliminate the short. If you have ALL signals at both PIA's stuck , replace U10 first. The replaced PIA works, replace the second to... If after replacement of the first PIA both are still stuck , there must be a fault at the incoming signals. Start with U10 this PIA has now been replaced and has less change to have bad contacts in the socket as it is far away from the battery and leakkage of the battery hardly get so high on the board.

Check with the logic probe for pulses at pins 22 ,24, 25,35,36, the signals arriving on these five pins involved at the selection where in one or another way already checked during this procedure ( it are address signals) , therefore if one of these is missing suspect first the socket or follow the signal to see where a run is cutted , use the schematic. Pin 23 a selection signal to, is coming from U17 pin 6 check the inputs coming from U18 pin 15 ( 10 for Stern) ) and pin 12, at one of these three is the fault. Last to check is pin 21 a signal directly coming from the CPU U9. ATTENTION ; the signals at pin 21 and 23 are EXTREMELY short pulses , you will only see these using a scope or a sensible logic tester. The Radio-Shack probe is just ok but barely to see , the self made tester of this site is not sensible enought , (sorry for that..).

If afther finding something here on U10 and U11 still does not work replace it to . Sill no go at U11 repeat the same checks as you have done at U10.

Now U10 and U11 are ok we continue with some other circuits. BUT first check if all LED's on the test box connected to the output pins are blinking. If not there is a bad resistor between the output pin of U10/U11 and the connector output pin , or a connector pin that has a bad solderring.

Zero Crossing.

Attention the indications between (.) are for Stern boards.

Use the logic probe to check that circuit. The pulses start from U14 and finally arrive at Pin 18 U10. Start at U10 pin 18 if you have pulses there all is ok and proceed with " Interrupt generator" . If nothing at U10 pin 18 go to U14 pin 14 (U14 pin 6) If you have pulses there and you will, because they come from the power source , replace U14.

Interrupt generator.

Check U11 Pin 40 ( pulses) if ok go to " Memory test ". If no pulses here , check at pin 3 of U12 , if pulses here the connection between this pin and U11 pin 40 is bad, perhaps the socket . If no pulses at U12 pin 3 replace U12, if still no pulses check the resistors around U12.

Memory test.

To start the memory test leave the test eprom 1 on the board , just push the push button on top of the CPU board.

Although Bally and Stern cpu boards can be interchanged, there is a major difference: Stern boards type -200 have one extra memory chip at U13. To make both types of boards use the same test, I've had to use some tricks in programming the test chip..

If I had 'nt done this, and you're testing a Stern board without U13 (or with a bad selection of this chip), it would be reported as ok, because a regular Bally board doesn't contain a U13 chip.....

So what you'll find is this...

The control LED will dim and after one second it'll light a first time. This means the test of U7 has succeeded. Again a second later the LED will light again. The test has run on U8 and U13 and has succeeded. If the LED lights a second time after 5 to 6 seconds, this means the test has succeeded but only for U8 !!

So on a Bally board both flashes will be about 4 to 5 seconds away. On a Stern 200 board the second flash has to be one second after the first. If it comes 5 to 6 seconds later, then U8 is fine but U13 is not ! (you can test this by removing U13) This way we can make a difference between cpu boards with and without U13. This was the only way to do it, as the only indication available is a LED and timing. So when testing, if U7 or U8 is defective there'll be no or only 1 flash. If you have a Stern 200 boards and you find 2 flashes with 5 seconds between them, then U13 is not ok or bad selected. Remove U8, and the test will keep on running on the U8 U13 test. Now it's easy to find out where the problem lies, by checking the selection and data signals.

Tests to do when a memory chip is bad

Swash1.gif

With a defect or bad selection of the memory chips, the test will keep on running on this one chip, by always selecting him again. This select is at the highest address of the chip, which results in all data and selection signals present and easy to measure. When U7 is defective (no flash at all) you have to find:

  • pin 1 = 0 volt
  • pin 2 to 11 = 2 to 3 volt
  • pin 12 = selection signal, don't measure this with a normal voltmeter but use a logic-probe or scope.
  • pin 13 = 2 to 3 volt
  • pin 14, 15 and 16 = 3,5 volt
  • pin 17 and 18 = 4,5 volt
  • pin 19 = 3,5 volt
  • pin 20 to 23 = 2 to 3 volt
  • pin 24 = 5 volt

For U8 and U13 you should measure this:

  • pin 1 to 4 = 2 to 2,5 volt
  • pin 5 = 1,5 volt
  • pin 6 = 4,5volt
  • pin 7 = 2,5 volt
  • pin 8 = 0 volt
  • pin 9 to 16 = 2 to 2,5 volt
  • pin 17 = 5 volt
  • pin 18 = selection signal, difficult to measure with normal voltmeter, use a logic-probe or scope.
  • pin 19 and 20 = 3,5 to 4 volt
  • pin 21 = 2 volt
  • pin 22 = 5 volt

When signals are missing or different, then you've found the problem. Follow this signal back to it's source using the schematics, until you find where it's blocked. If the signals themselves are ok then the memorychip is bad.

Swash4.gif

This concludes the repair method using the first and basic test eprom. Before we proceed with the game roms and execute de normal game tests on the board we make shure the sockets and the circuitry for the other game roms is ok.

Swash4.gif

Test of U2 and U1 ( if present).

Remove the test eprom 1 and place test eprom 2 in U6 , at the same time test eprom 3 in U2 The only extra thing we check with these two test eproms is if the test eproms in U2 and U1 respond ok.

At power up the first eprom checked is the one in U2. If all is ok the test LED still connected at pin 15 of U9 will blink , the LED on the CPU board will blink to. If the LED is stuck this means the program does not find U2. The only possibility is , bad contact in the socket or a missing signal. Check, with the logic probe using the schematic all pins , most likely there is a selection signal missing at pins 18/20 or 21. Look where it comes from because that depends on the jumpering..

If ok, we check U1 now, Take the test eprom out of socket U2 and put it in socket U1.Power up again and push the button on the CPU board , this selects the test eprom at U1. If ok the LED blinks. If not ok follow the same method as for U2. ( The check of U1 is only needed if there is a game rom used in U1).

Swash4.gif

Game tests.

Put the game roms on the board.Your board is booting ok now , with 7 flashes. If the display selection is at one of the 4 players it will display 00 alternating with the highest score to date. ATTENTION the highest score can be 00 or nothing ( blank display) as we worked on the board . To start the game tests we push the red push-button on the test box.( the push button brings briefly "ground to pin 1 of J3, simulating the normal test button at the inside of the game door). This start the first test wich is:

All lamp test

The control LED's on the test box connected at J1 will all light, they are "on" not only during the lamp test but also during the other tests. This because the outputs of U10 , where the lamps are connected ,are also active during the other tests.

Swash1.gif

Display test.

Push the push button on the test box once more. The display test starts. As we are shure the display of the test box is ok, al I treat here is concerning the CPU signals. If you use the test box to test a bad display go here .. Pinaqua.gif ( under construction - never was a valid link). Measure using the logic probe. To control alle display outputs , just turn the rotating switch on player 1 to 4 and the credit/ball position.

No display at all. Check first the "blanking circuit". Blanking starts at U10 pin 39, if missing there replace U10 , if ok there , follow it trought, U20 pin 12, U14 pin 12, U14 pin 13 . Where you lose it check that out , only 5 components involved, see schematic. If all that seems ok , follow the lines for " Missing a player"

Missing segments on all the digits, = missing segment BCD data , these come from U10 pins 6,7,8,9 If ok there check resistors R 80,81,82,83.

Missing some digits, = these come from U11 pins, 3,4,5,6,7,8,9. Must be ok at the output of U11 ( already checked) . Check resistors R 90, 91, 92, 93, 94, 95, 96.

Missing a player , = Latch strobe missing, these come from U10 pins 2,3,4,5. Check at U20 the 4 players inputs are at pins 3,6,1,10, the latch is at pin12 of the same U20. If you have the inputs and the latch , replace U20.

Missing the credit/match display. Here the signal comes from U11 pin 2, going to U20 pin 13 where it is mixed again with the latch signal . Check the in- and output.

Swash1.gif

Solenoid test.

Push the red button on the test box again. The display will cycle from 01 to 15, or from 01 to 30 , depending on the type of game . All the outputs coming from U11 are checked already. We need no time to waste here.

Swash1.gif

Switch test.

The normal readings using the switch box are.

S1 position 0= nothing, S2 can be in any position.

S1 position 1. Placing S2 at respectively 1,2,3... will display 01,02,03...

S1 position 2. Placing S2 at respectively 1,2,3.. will display 09,10,11...

S1 position 3. Placing S2 respectively at 1,2,3.. will display 17,18,19...

S1 position 4. Placing S2 respectively at 1,2,3.. will display 25,26,27..

S1 position 5. Placing S2 respectively at 1,2,3.. will display 33,34,35..

S1 position 6. Placing S2 respectively at 1,2,3.. will display 01,02,03..

S1 position 7. Placing S2 respectively at 1,2,3.. will display 09,10,11..

Take the switch control box and connect it at J2 and J3 of the CPU board. Power up and by pushing the button on the test box cycle trought lamp test/ display test /solenoid test to switch test . As the strobes for the switch matrix are coming from U10 and the outputs of U10 are already checked during the lamp test , we can be shure these are ok. Only the returns can be involved. If there are some returns not working correctly , there are two possibilities. The return is stuck "Hight". At position 0 of S1, you will directly read wich return is involved.

For display 01 it is pin 10 of U10 for display 02 it is pin 11 of U10 and so on till display 8 for pin 17. All that with switch S1 in position 0.

If the return is stucked " low" then for any position of S1 EXCEPT position 0, you will read ,turning S2, all normal numbers until you come on the bad return , afther that no more display... An example S1 in position 1, turning switch 2 we read for the first position 01 for the second 02 for the third 03 for the fourth " nothing" = pin 13 of U10 stucked low.

The positions of S2 are corresponding as follows,

Position 1= U10 pin 10... Position 2 is U10 pin 11...Position 3 is U10 pin 12... Position 4 is U10 pin 13..Position 5 is U10 pin 14... Position 6 is U10 pin 15... Position 7 is U10 pin 16... Position 8 is U10 pin 17.

A bad reading involves the replacement of U10 or a shorted input pin. The short can come from the capacotors at the inputs C19 to C26.

This ends the complete control of the CPU board . If not working in the pinball , look for bad connections at the CPU board connectors, this is the only thing that can go wrong if you have the right tensions at J4, and all you need are these tensions to be shure to have the 7 flashes .

.Swash1.gif