Well, here we go. I bought an old, untested CPU board on ebay quite a while ago. It clearly had some battery leakage problems that had been cleaned up by the previous owner, so now I need to fix the rest and get it up and running again.
Step 1: asses the damage.
I noticed some of the copper was missing where the IC fittings used to be, so I first started with measuring the tracks that were still there. A lot of them are broken and not making the right connections. The breaks were too big to fix by covering them with a bit of solder on the component side of the PCB, so I will need to run wires on the solder side of the PCB. But first, I put in the missing IC fittings for the RAM chips.
Next I removed the 24-pin IC fittings for the three 2532 EPROMs. I replaced them with three 28-pin fittings so I can put in two 2764 EPROMs, which are needed to use David Gersic’s Freeplay ROMs and are much easier to come by. Using three fittings still gives me the option to put in the original 2532 EPROMs if needed. This also involves changing the jumpers just above the EPROM IC fittings.
As I am not going to reuse the IC fittings, I just cut them off.
The remaining “feet” were removed with a soldering iron. Once done, I just popped in the new 28-pin IC fittings and soldered them in place.
Now I started measuring the connections on the new fittings and found that more than half were bad. This is going to be a lot of work. I’ve started wiring on the back side of the PCB, but time ran out, so I’m not finished yet. I have to redo connections between both RAM fittings and between the left RAM and right EPROM fittings. There was clearly a lot of damage from the battery leakage.
I will keep you updated on my progress.
I’ve received the new fuse clamps and quickly installed them on the Devil Riders’ power board. No more fiddling with the fuse.
Then I took out the sunken inserts.
After a close inspection I saw that the 3000 hole is in the worst shape. It will need some wood filler and paint touch up.
I’ve reseated the inserts with some superglue, they shouldn’t move anymore.
I also took apart all the targets at the end of the playfield, cleaned them and reinstalled them. Fortunately, the print on them is still in good condition, so I don’t have to replace any of them.
With that done, I started on the popbumpers. They were dirty and one is using a wrong coil that didn’t fit very well and was missing some screws.
All the popbumpers are now clean and ready to be reinstalled. I kept the one with the wrong coil as I’m still looking for a replacement coil.
I also took apart the outhole kicker, cleaned it and reinstalled it. I noticed that the wrong coil used in the popbumper is the same coil that is used for the outhole kicker. So if I ever find a new popbumper coil, I’ll have a spare outhole kicker coil.
Next up the metal parts and the plastics. Then we can start to reassemble!
Oh and my dad came by for a game on the Pinball Champ…
So I put together the little address selector tool that Leon designed on his website (see my links) and used it to test both the Pinball Champ ’82 and Devil Riders driver boards.
I used a LED to test each input individually for the eight possible addresses.
I had to do it on the PC82 since I don’t have an independant 5V power supply anymore… really need to fix that!
The Devil Riders driver board was in perfect condition. This machine is turning out to be just a case of bad switches and dirt, lots of dirt… excellent!
The Pinball Champ driver board was another story. I had intermittend faults on connector CN21 while signal that came for the same SCR’s on CN18 where perfect. After a more thorough visual inspection of the board I noticed this:
As the solder didn’t appear broken and I had a very hard time desoldering this connector, I suspect it had been put together this way 29 years ago… So now it’s reseated properly and I replaced four SCR’s that were faulty, this board is now in tip-top shape as well.
I’ve also tested the Devil Riders CPU board with Leon’s test EPROM and it’s also in perfect condition. I’ve also found some discrepancies in the way the Pinball Champ CPU board functions compared to the Devil Riders board. I’ll get into that next as it’s the only thing that needs to be done before I can really get into the Devil Riders.
Wish me luck!
I’ve been lucky and was able to buy 27 used, untested ZM1550’s. So I’ve made a test setup to get them all sorted out as fast as possible once they arrive.
This is just a little test print on which I soldered three 5 pin connectors and one 4 pin connector and then connect all the segments of both digits together (a with a, b with b, etc.). Then I soldered some wires to connect them to the CN3 connector on the pinball display driver. The shield is connected to pin 1 of CN1 and the decimal points and both anodes are connected to the appropriate pins on CN2 in the same way.
Fortunately, the first ZM1550 module I put in worked, so I know the setup was good.
Unfortunately, not all the modules I put in worked 🙁
Well, I’ve got the fuse back into the power supply and sorted out the previous repairs. I tried testing the CPU but apparantly the 27C64 doesn’t work as well as a 2764. Don’t know why, it should work. So I just popped in the original game EPROMS, connected the power supply and the displays and that worked.
So I connected everything else and here is the result:
I’ve ran the normal diagnostics and except for some lights and switches here and there, everything seems to work. The little motorcycle in the head moves a little bit, but since there still are some wires hanging there connected to nothing, that doesn’t surprise me.
The biggest job is going to be cleaning up… this machine is VERY dirty.
Well, the playfield is almost done!
I’ve cleaned out the cabinet and head.
I’ve installed the new microswitch (#52) that let’s the machine know the ball is on the 2nd level playfield.
I’m still missing a few rubbers, but they are in the mail.
The 20A fuse on the transformer box burned out while I was working on the playfield. I stuck a new one in and that burned out too. So I removed all the lights and measured the wiring to look for shorts. There were none. I stuck the lamps back in one by one and now everything works… Weird.
The guy I bought the ZM1550 display modules from mailed me to say he is abroad for work and will send them next week.
I still have the CPU board problem though… I’m going to swap out IC’s 8, 19, 31 and 37 for good measure. The way the CPU reacts to the DIP switch 4 position looks like it’s unable to read the settings once PROGR is off. When I put the switch back in PROGR mode, I can read all the settings, which to me means the RAM is good.
I’m also writing a extensive memory test program that will write alternative 0x55 and 0xAA’s to every memory position between 0x1800 and 0x1BFF. When a test fails it will show which test failed in which nibble at which location on the Player 1 display. This way I should be sure if the memory is good and I can test with both positions of DIP switch 4.
I’ll keep you updated.
As I stated in an earlier post, the pinball machine will only work when it is in PROGR mode (DIP switch 4 to ON position on the CPU board).
When the machine is in normal mode (DIP switch 4 to OFF position), al the displays show 0’s and that’s it. Sometimes not all displays will light up at once and it will take up to 30 seconds for all of them to turn on.
Lights on the head and playfield will stay on or off at random (changes every time I restart the machine).
When actuating the Test-Advance button in the door, it will go to test #1 and all the displays will test correctly showing 0’s, 1’s, 2’s, etc… up to 9’s and back to 0’s, but after that, there’s nothing I can do to stop that test, go to the next test or whatever.
Also, when I start the machine on normal mode, the error LED on the soundboard lights. Actually it always lights and never turns off. I noticed it takes a little time to turn off when the machine is turned on in PROGR mode, so I guess it’s the CPU not initialising the sound card.
I’ve tested the IC’s 31, 37, 19 and 8, but they all do what they’re supposed to do…
What you can see here is that the PROGR signal enables addresses 1C00 to 1C7F (or 1800 to 187F, since AB10 is not used but the software holds it high for RAM access) for writing when it is LOW. When the switch is put back to the OFF position, this memory range becomes read only.
So what’s next? I’m not really sure. There’s a few options:
- Maybe the TMS40L45-45NP is not a good substitute for the 2114L. I found some people who still have these, so I’ll get me one and see what happens.
- I write a test program than constantly writes to RAM when the CPU board LED is on and constantly reads from RAM when the CPU board LED is off and follow the R/W signal with my logic probe.
- Call Mulder and Scully to find the ghost in my machine…
Anyway, as always I’ll keep you updated.
This is a little how to on how to “field”-repair a pinball lamp socket. With field repair I mean, you don’t have a new lamp socket handy and you want the thing working NOW. You don’t even have to disassemble the socket to do this, but for this tutorial I’m using pictures from a mounted and unmounted socket repair. This is a trick I learned when I was working as a technician in an arcade hall way back when…
So, your light doesn’t work, but there is power going to the socket. You test the socket and notice that there’s no connection between the “live” lip and the spring inside the socket. This is what you do:
With a piece of sanding paper, sand the top of the spring. Use your soldering iron to put a little dab of tin on the spring.
Take an end of shielded wire and strip one end about half a centimeter. Bend the exposed wire in a little hook and put a bit of tin on it with you soldering iron.
Thread the shielded part through the spring and make sure the hook catches on the spring.
Solder it thight, easy job since the tin is already on the metal. Just heat it up.
Then cut the wire to length, strip the end and solder it to the “live”-lip.
If you’re working on a mounted socket, be sure that you don’t unsolder the other wires.
Now pop in the lamp and if you followed this guide correctly…
Now get to work!
The mailwoman came with goodies this morning; four 3081 chips! So I put them in the CPU board and ran Leon’s test EPROM again. Victory!
So I mounted the board back into the pinball machine, connected everything and switched it on. Not so good… The sound was weird, only a few cracks and screeches would come out of the speaker.
I disconnected the sound board from the CPU board and ran the test and all went well, so clearly there was a problem with the communication between the CPU board and the sound board. I checked the schematics and the only thing that could be wrong was the entry stage behind CN7.
Using my multimeter I immediately found that all the data lines where shorted to the +5V. So only the resistor array ICR6 or the 74LS244 could be the cause. So I started with desoldering the resistor array, but that wasn’t the problem. So I desoldered the 244 and found the culprit!
Off to the store I went. Bought a couple of BF422’s and a 4042 as well for the two display driver boards that needed them and even went round to the hardware store to get some screws, there’s a lot of screws missing in the cabinet and head.
Back home, I put all the components where they should be and succes! I have sound!
The two display driver boards got their components too and now I have 4 working displays.
Next week my ZM1550’s should arive and then I can fix the last display.
And that’s it for the electronics… OR IS IT???
The game still won’t start. No flickering lights, no sound, no highscore showing, nothing… just all 0’s on the displays and that’s it.
When I put DIP switch 4 in PRGM mode, the soundboard now makes the annoying warning sound and I can do all the tests, reset the accounting and set all the options. So I set the options, resetting the ones that already had a good value, clear the accounting and do the tests. All tests are fine. The displays do their thing, the lights go on and off, the solenoids scare me and the sounds are played in order. But then there’s the button test (test #2) that just shows 18. And whatever button I press, it doesn’t change. I guess I’ll have to go through the wire harness and measure what signals I do get.
Guess that’s my queue to end this article and get back to work!
I went back to the power supply board today for two reasons; I still had that little POWER FAIL circuit to fix and I had trouble finding the reason why my controlled playfield lights weren’t lighting up. So I found that the 5V was missing from CN2.
After some testing I found that it was the fuse and fuseholder were oxidised. Some cleaning took care of that and now the 5V is back.
It occured to me that now that this 5V was back, the POWER FAIL circuit had to be revisited, and low and behold, the POWER FAIL signal is gone. So the circuit wasn’t failing, I had a missing voltage and had completely missed it. DOH! /facepalm
I reconnected the POWER FAIL signal on the CN9 connector on the CPU board and everything still boots up 🙂
After that was all over with, I started disassembling the playfield, documenting everything with a lot of photo’s since I want to be able to put it all back together again too…
Once I removed everything from the top of the playfield, I flipped it over and got to work on the top drop target bank.
I removed the broken drop target, replaced it with the one I got from David from zaccaria-pinball.com (see my links) and flipped the playfield over again.
So with that done, I started cleaning. I haven’t done a lot yet, but here’s a picture for you to see how much difference it makes already.
And notice the working lights, what a big difference 5V makes 🙂