Hello, community. Member for a long time, and have used the search with good results multiple times. This time, however, no luck. Let me give you a quick rundown on what's happening.

This is a standard Williams sound board, regular mods completed a long time ago. For comparison, I have 2 non working, and 1 that works great (it's in my TriZone machine). This board works great in Flash, so the problem is limited to the sound board itself. Power at IC11 (7805, 5 v regulator) is stable for a very long time (stopped after about 20 minutes, but it was perfect). 12v unregulated coming in is stable, measured after R27, and going into IC11.

2 boards have 6802 processors, TriZone has a 6808. The TriZone board works correctly, in the Flash machine, with all 3 processors, and all 3 ROMs. Caps have been replaced as a normal first step. I've even tried moving caps and part form the TriZone board to the Flash board, to verify they were still OK.

All 3 board tests produces the normal, correct sounds in both TriZone and Flash.

After initial power on, I can ground the sound pins (10J3, 2-5 and 7) and it will make the correct sounds, but only for about 30 seconds. It *feels* like there's something either not getting correct power, or failing. I'm at a loss which component to check next. I'm leaning towards the 6820 PIA, or the 1408 D/A converter.

Asking the wisdom of the community to guide me.

I really only have a DMM for testing, but I'm pretty good with soldering. I have no issues removing a 40 pin IC and replacing it - if I can find the replacement.

Thanks, much!!
Scott

SCANBE, no. Both the CPU and ROM sockets have been changed. The PIA is labeled with the Motorola "M" icon and is soldered to the board.

Once it fades, the test still produces the correct test sound, and I can still adjust the volume, so it's not the amplifier section of the board. I think it's the version 1 board - the one with 2 fuses. See the attachment.

Thanks for any ideas.

SB_Schematic (resized).jpg

Thank you for the reply.

Is there some easy way to check if a pin goes high? Really, I only have a DMM. It's not fast enough to show me that changing. Granted it's a good one - Fluke - but it's still just a DMM.

I'll start with IC4. I'll report back when done. Thanks.

Upon inspection, pin 7 looks like it has corrosion on it. Didn't notice this before. Hmmm...

IC4 removed, socket installed, chip replaced. Same, same. On to IC8.

Checked pin 13 on IC4 doesn't seem to go high at all. What does that tell me?

Duh...was checking the wrong pin. It does go high. 5 volts.

/sigh. IC8 removed, socketed, replaced. Sound seemed to last a few more seconds, but that's about it. This is pretty depressing.

Went ahead and bought/replaced the ROM socket, anyway. No change.

Any ideas are welcome.

Just found this thread. Going through it now. Hope I find some solution. Thanks for the inputs so far.

(it ain't working yet, so I'm not done T/S...)

https://pinside.com/pinball/forum/topic/williams-type-1-sound-card

Thanks for all the inputs so far. I've been working all week, and had precious little time to dedicate to my passion. Plus, I still have husband duties, taking care of the wife duties, dog duties, etc. But, so far...

- The speaker is fine. The test sound exercises that, and the amp (among other things). I did find the jumper from internal sound production to the internal amp (the remote volume control one) had "issues" (jiggle or touch it, and sound would disappear then come back), so I removed it, and used old style telephone winding wire to ensure that connection is solid. It ain't pretty, but it works, and can remove with a pair of pliers and about 10 seconds.

- I spent time removing, socketing and installing chips, and I'm actually in worse shape now. Nothing at all. The beginning sounds (sounds for like 15 seconds) have stopped. So now I'm removing sockets, and going back to on-board soldering. I missed some connections, and can't seem to find them, so doing away with the farce altogether.

- Yes, I have 3 boards. 2 for Flash, and 1 (the working one) from my TriZone. I'm not willing to part out a working board, but I have used it to test the rest of the Flash machine. The ROM is the same, so I tested the 3 ROMs - they all work in the operational board. I tested all 3 CPUs, 6808 and 2 6802s, and they all run the TriZone board (it originally had the 6808). Those are the only chips I can move around with de/soldering. The Flash machine (MPU/driver board) is sending all the correct signals to the card, so it's definitely the card itself.

Where I'm at right now: Ordering:
- 2* MC14068B (B input NAND gate),
- 2* MC1408L6 (D/A converter, this one is harder, because it's a 6 bit chip - lots of 8 bit chips are available - can we use an 8 bit? All my boards have 6 bit),
- 4* MC14050B (buffer chip)
...to replace all those in both non-working boards. I've done all the caps on the board that had *some* sound, but not the other. So I'm getting a replacement set of those for that board, too. It would be nice to have a spare board. As (I think) I said above, I'm somewhat of a purist, and I prefer to have all original parts. Hey, they worked for 40 years, right? I should get another few out of them, and I won't make it for 40, so I'm good.

I'll update again when I get - and install - all of that. Hope it will be fairly quick. It's all from within the US. I have time.

@G-P-E, @schwaggs, thanks for the idea, but the amp is not failing. The test runs for hours. If the amp was failing, that would stop at some point.

So tracing the diagram, when I ground 10J3-2, it sends IC8, pin 11 from 5v to ground, that works,
IC8, internally, sends pin 12 to ground, that works,
IC8 pin 12 sends IC4 pin 10 to ground, and that works.
IC8 pin 12 is also tied to IC3 pin 11. That seems OK.

IC4 (from what I understand), at this point, should send pin 13 high (5v). That doesn't work. Pin 13 is always high.
IC4 pin 7 is tied to ground, pin 14 is tied to C24, C24 is also tied to ground on the other side. It appears to measure correctly.

IC4, Pin 13 is tied to IC3 (the PIA) pin 18. It's always high.

I've changed IC4 (twice), but I don't know if the 5v is being generated by it, or if the PIA is messed up, and it's putting out 5 on pin 18. If it is messed up, that would explain why nothing works. That's a chip I don't have to swap and test.

It's no longer producing any regular sounds, except the test. I don't know where I went wrong, but something is even more amiss.

Originally, that 30 second blurt, was regular sounds, if I grounded an input pin. It would produce that sound, and the whole thing lasted about 30 seconds, if I kept grounding pins, then no more game sounds. If I hit the test button at that point, it would still produce test sounds.

Maybe I wasn't clear enough, originally - but it no longer matters. Something happened, and I'm not getting anything but test sounds now.

For anyone wanting to use a PC power supply to bench power these boards, it's pretty easy to do.

Any ATX power supply will do.
- Any yellow wire is +12, that will go to 10J1, pin 1 or 2,
- Any black is ground, that will go to 10J1, pin 5 or 6,
- There's only one -12v you can use, and that's on the mainboard connector. Typically, it's the blue wire, but if you're looking at the connector, wires away from you, with the connector hold key on the right, it's the second down on the right side, pin 12 or 14, depending on your connector. That connects to 10J1, pin 8 or 9.
- To force the power supply on, you short pins 14 to 13, or 16 to 15, depending on your connector.

If you did it right, you'll get normal operation from the board, and you can short the ground plane (any mount hole is the easiest to clamp to) to the input sounds selects on 10J3, pins 2, 3, 4, 5, 6. Those are the basic sound points.

See the pic for details.

Hope this helps someone.

atx-pinout (resized).png

Well, it's been a while, but I got some new chips, and made some changes. Spent some serious time examining and following the schematic (or logic, depending on what your doc is called), I found...

Bear in mind, this is an ATX powered board, with no sound pins grounded.

- All pins on 10J3 (Sound Select inputs) should read 5v at all times,
- Pins 2,3,4 and 5 run through IC8 (4050 buffer),
- Pins 5,6,7,8 and 9 run through IC9 (4050 buffer),
- IC4 (4068 NAND gate) will have 5V on 2, 3, 4, 5, 9, 10, and 11

If you ground any input pin, it will send the corresponding ping on the 4050 to 0, which in turn sends the corresponding pin on IC4 to 0. That operation sends pin 13 on IC4 high (5V) to pin 18 on IC3 (the 6820 PIA).

The only chip I needed to change was IC9. That was messed up, and holding all 4 inputs low. So now, I have the (supposedly) correct signals going to the PIA. If I ground a sound select pin, the correct corresponding 4068 pin goes low, and pin 13 goes high. I haven't verified the correct PIA pin goes low, but as the trace runs across the pin, I can't see how it wouldn't. I've verified the solder joints on the PIA look correct, and re-flowed them. Nothing seems cross-connected with solder.

The test sound still works (don't ask the wife - she hates it, but I like to know it's working).

No sound output on an input pin ground, not even those initial few seconds on first power on.

So what does the group think? Failed PIA? It's the only thing I can think of. I've had some great suggestions above (@Tuukka, and others). Voltages all seem to be correct. My next step is to remove the PIA, zip socket the slot, and try a different chip. Don't have one of those hanging around, though.

If you buy strips to socket your board, be sure the pitch (space between the pins) is 2.54mm, or they won't fit correctly.
Got these. Seems perfect to socket these things. Can cut to length for every chip.
https://www.amazon.com/gp/product/B012ACSO4Y/ref=ppx_yo_dt_b_asin_title_o00_s00

Looking forward to more suggestions...
Scott

All cards are working.

Spent some time detailing the power readings I should be getting on IC3, IC4, IC6, IC8 and IC9; voltages in various places on the board; which pins go to 0 and which ones go high when sound select inputs are grounded; what changes on the data bus I should see (D0 through D7 and PA0 through PA7 on the PIA), the inputs PB0 through PB7, outputs PA0 through PA7, grounds, resets, all that jazz.

I've already proven the CPUs and ROM chips, so didn't see the need for those. Followed various traces on the board and the schematic to see what I should have and where. Found the IC data sheets online (mostly for the pictures), wrote down all the power values from the working board, then compared against the bad boards until I found differences. All in all, it didn't actually take that long to do, once I got into it. Just took a long time to get into it.

First bad card: IC8 and IC9 changed, IC3 - PIA, Pins 39 & 40 weren't connected to each other, and 37 & 38 appeared to not be - so made sure they were. Had 5v on 40, but not 39. No idea why it worked for those 15 seconds or so - I don't think it should have - but it did. As I removed it, socketed it, then cleaned up and reseated the same chip (yeah, I know, just wasted time), but maybe the act of installing the sockets fixed other connections I didn't get when I re-flowed the chip. No idea, but it's hard to flow the top when the chip is in the way. It was easy to verify and fix connections with the socket strips installed.

Second bad card: R41 and IC9, 4050 buffer chip failed - that was an easy fix, once I found the issues.

GRUMPY Thanks...I did try to cool a few chips, but that made no difference. It was worth a shot, though, and I'll remember for next time.
Tuukka Thanks for the suggestions. It started me down the path. Guess I just needed to get into it more. Really helps to have one working example you can test on.

Anyone got dead cards they want to get rid of? This is getting fun.