exxos's Atari, Amiga & retro forum

OK, so the sound circuit isn't overly complicated:


There is an M14050B Hex buffer fed by the 6821, and this then heads off to an M14529B which is a 4 channel analog data selector. The inputs determine what channel is selected as below:


Unfortunately that means the only culprits can be either the 6821, the M14050B or the M14529B. Since I tried a known good 6821 with no change, that leaves the pool of suspects somewhat smaller, so I have ordered one of each.

Oddly my CocoSDC doesn't seem to work with the Coco (For now), so I picked up a game cart to test with.

Since the video is working as expected, and the audio is wired to where it should be, pin 3 of the modulator, then no sound leaves very few components that can be faulty.

I guess the only way to find this is to have something constantly playing sounds and then trace back through the circuit until a signal appears.

stephen_usher wrote: 26 Feb 2024 12:19 I guess the only way to find this is to have something constantly playing sounds and then trace back through the circuit until a signal appears.

Yes, that's what I've been trying to do, unsuccessfully I might add!!! There is a SOUND command, so SOUND 77,255 gives a pitch of 77 for a delay of 255, which equates to 20 seconds.

At present though, zero output.

So now I have a better idea of where things are coming from and going to that will help track it with the scope.

And I can run this to help keep sound going whilst I test:


So I shall do that when I next have the chance.

One oddity is that pin 3 of the modulator has one connection from pin 11 of the 6821 directly for something called 'single bit sound', but when I run the sound command I see nothing at that pin, suggesting that it's something else entirely and that the main sound comes from pin 10 of the 14529 provided it's selected.

Is there anything on the data lines going to the D2A chip?

stephen_usher wrote: 26 Feb 2024 14:02 Is there anything on the data lines going to the D2A chip?

So I have scoped the inputs and outputs of the 4050 and there is input and output. I don't think the 4050 is an issue.

I can see nice traces on the scope on the inputs and outputs, varying with the pitch I apply to the sound command. I think the base functionality is there, in that the sound command works.

I also see output @TP7, however, this is where I hit a snag. I also looked at the group of resistors, R10-15 and I see nothing, but there is another group, R80-85, and they are right next to the 4050, and I do see signals there, so it seems the schematic I have doesn't quite match with the machine on my bench.

It looks more like TP4 is connected to pin 14 of the 14529 (It buzzes out fine) but all I see that do is go high when I play a sound, and I should see the output of the 4050.

There is a TP7 over by the cartridge port and interestingly, when I run the sound command I do see output there, but I don't think it is anything to do with sound since it looks more like data, so probably just related pulses when I run the command.

I think I need to trace from the D/A output up to the 14529, since the SNDEN line from the 6821 is going high as expected. What doesn't make sense are lines 6&7 of the 14529 as they don't seem to change if I use a CSAVE or CLOAD command, though for normal sound output they should both be low, which they are.

So tracing the output from the 4050 up the board I arrive at R39, a 66K resistor- again not on my schematic. It's the last leg as the other side connects to pin 14 of the 14529. I get a signal at the input end of the resistor, and at the other end, well it just goes high.

So I know where the sound is stopping, just not sure why, and I think I need to find a new schematic that matches my revision board. Not that I can tell that either since it's not printed anywhere on the PCB!!

This is what the scope looks like:


Yellow is the 4050 end of the resistor and blue is the other side, which just goes high...

There's probably a pull-up on the output side of the resistor, though 66K seems a very high value.

stephen_usher wrote: 26 Feb 2024 16:37 There's probably a pull-up on the output side of the resistor, though 66K seems a very high value.

There is a pull up on the line, according to the schematic, R72, but it simply doesn't match my machine on which it is R39. It's kind of in the right place in the circuit though. R72 is marked as 100K.

However on a whim I decided to pull the 14529 since it's in a socket, and lo and behold I can see the signal both sides of the resistor, which suggests to me that the IC is either marginal or faulty and that's where the issue lies.

Testing at pin 14 of the socket, and on TP4, the signal is clear as a bell. These are all in a direct path to that side of the resistor.

There is literally nothing between the output of R39 and pin 14 of the 14529, so I can only hope that when the replacement arrives, it just works.

If it doesn't, then I can only assume there is some sort of other logic fault I've yet to diagnose.

OK, that chip is the problem. There's obviously an internal short to VCC.

stephen_usher wrote: 26 Feb 2024 19:52 OK, that chip is the problem. There's obviously an internal short to VCC.

Yeah, looks very much that way, hopefully the replacement will be here quite quickly.

Quick test before work this morning:


The green clip lead is hooked up on the left to TP4 (Labelled TP7 on the schematic), which connects to the input, pin 14, of the 14529 which we now know is dead.

The right hand end is connected to the end of the audio output lead that is hooked up to pin 3 of the modulator input. This pin goes back to pin 10 of the 14529, which is the normal sound output.

Now when I run the SOUND command I get output through my mini TV, so that's yet a further nail in the coffin of the 14529. I also get saving noises if I use the CSAVE "" command.

I'm sure that the sound output should be redirected by the dead IC over to the cassette port, but it is reassuring to know that CSAVE does actually do something.

I also wanted to test the circuit path from the output of the 14529, pin 10, over to pin 3 of the modulator end of things, so I hooked up TP4 to pin 10 of the IC socket, and sure enough, that section of the circuit is also good.

So, whilst I wait for the replacements to arrive, I may have a look to see what memory is required to take this to 64KB. As it's a later machine, it has a set of jumpers that are either set to 16KB or 64KB, but given the current 16KB RAM is effectively 4116, it uses 3 different voltages, so I need to figure out what the jumpers do before blindly fitting 4164's.