I had doubts the issue was in the audio circuit itself, since I would expect either some audio output even in just one channel. So I decided to see if the DAC itself was a the problem.

I initially tested a working Roland SCC-1 to use as a benchmark. I took voltage readings and signal readings on my oscilloscope for clock and input signals. I then did the same for the faulty card.

Between the two cards, the only difference showed up on pin 15 of the DAC, which is the signal in that runs from pin 13 on the Roland PCM sound generator (IC12).

On the good card, the signals look like this:

On the bad card, they look like this:

Basically, when a MIDI file is playing there is no stream of data being fed into the DAC, hence no audio output.

This means the issue must be further up the chain and likely somewhere in the sound generation circuit.

Next step I decided to test the MIDI input from the MIDI circuit into the sound generation circuit.

It enters by way of a signal fed to the sound generation control chip (IC10) via pin 67.

Probing this pin on the bad card revealed no activity when playback was stopped, but a signal when playback was started. Haven't tested the good card for comparison, but at first glance it does look like a signal is being fed into the sound generation circuit.

So, this leaves what to test next. There are 9 different ICs in the sound generation circuit:

IC12 - PCM sound generation
1C10 - sound generation control
IC13 - gate array
IC14 - program ROM
IC11, IC15 - RAM
IC17, 18, 19 - sound ROM

I've done some initial testing of voltages on each chip. Everything is approximately ~5V. No shorts that I can find. No overheating chips. And no loose legs.

I've also tested other components in this sound generation circuit like caps (no shorts) and nearby resistor arrays (all measured fine).

I'm thinking of plugging the working SCC-1 back in and then maybe systematically checking each IC along data paths to see what inputs/output signals look like, then comparing to the faulty card. I'm hoping I can narrow it down to whichever chip may be faulty and then try to obtain a replacement.

If anyone has any other suggestions for things to try, I'm all ears.

After a bit of a hiatus, I resumed trying to diagnose and repair this card.

Starting with IC10 (main CPU controller) I did some more comparative readings between a working and non-working SCC-1.

Pin 67 (MIDI input) with the working card while playing a MIDI:

Pin 67 (MIDI input) with the non-working card while playing a MIDI:

On both cards the activity indicates there is MIDI input going into IC10.

Then I did some testing on the crystal connections on pins 69 and 70.

Pin 69 (working card):

Pin 70 (working card):

Pin 69 (non-working card):

Pin 70 (n0n-working card):

Both cards have a working crystal with a relatively stable 20MHz signal.

I also tested one of the data lines on IC10 (pin 13).

This is where a difference in the signal shows up:

Pin 13 (working card):

Pin 13 (non-working card):

There is a clear difference in signal between the two cards. The signal also changes when playing a MIDI versus not playing one on the working card. But on the non-working the signal doesn't change.

The data lines run between various ICs including the CPU, Gate Array, PCM generator and RAM. I don't know if this means the problem is directly with IC10 or possibly with one of the other ICs.

Hope you can figure it out. It's such a wonderful card.

Tempus wrote on 2023-12-27, 23:59:

Hope you can figure it out. It's such a wonderful card.

Thank you! I hope so too!

And even if I can't ultimately repair it, I'd at least like to figure out what the fault is.

Maybe it can serve as an MPU-401? I mean if you can't figure it out.

Tempus wrote on 2023-12-28, 09:11:

Maybe it can serve as an MPU-401? I mean if you can't figure it out.

Yup, it does work perfectly as an MPU-401 interface. So if all else fails it's still useful for that.

Continued to do some reading and testing related to the main CPU (IC10) chip.

I've been trying to figure out what the various pinouts do and determining what normal operational behaviors would look like. There is a test point (TP1) on the card connected to pin 66 on IC10.

I was able to obtain the data sheet for this chip and it turns out it's the serial data transmit pin.

Probing with the oscilloscope didn't give me much. With the working card, it seems to generate activity when playing a MIDI. With the non-working card, when starting a MIDI file it seems to trigger some sort of output signal, but not sustained in the same manner as the working card.

Given this is digital serial data I have no idea what it's actually transmitting or how to interpret it on the scope. I imagine this was probably intended to be hooked up to some sort of serial interface with proper testing equipment for this chip.

I also checked pin 72 which outputs a System Clock. It reads 10 MHz on the oscilloscope (non-working card). Haven't tested it on the working card yet, but I'm assuming that is a normal reading.

Shponglefan wrote on 2023-12-28, 16:31:

Tempus wrote on 2023-12-28, 09:11:

Maybe it can serve as an MPU-401? I mean if you can't figure it out.

Yup, it does work perfectly as an MPU-401 interface. So if all else fails it's still useful for that.

regarding the Intelligent MPU-401 chip, maybe this post I made will be of interest to you:

https://github.com/schlae/mpu-imc/issues/1#is … ment-1832851222

because SCC-1 card like MPU-401N has basically single chip that integrates 3 chips to handle the Intelligent MPU-401. that chip is (micro-controller + SRAM + ROM) in a single chip. you can check if that chip is running in the following way:

Re: Roland LAPC-N / I question

i.e. get the version of the running firmware inside it.

On the repair - what about making dumps of the ROMs and see if none of them is not broken, you can find SCC-1 ROM dumps here:

Re: About Roland Virtual Sound Canvas 3

I guess most suspicious is the ProgramROM as it's running the show - but I guess that is also the easiest to replace.

mattw wrote on 2023-12-29, 03:01:

regarding the Intelligent MPU-401 chip, maybe this post I made will be of interest to you: […]
Show full quote

regarding the Intelligent MPU-401 chip, maybe this post I made will be of interest to you:

https://github.com/schlae/mpu-imc/issues/1#is … ment-1832851222

because SCC-1 card like MPU-401N has basically single chip that integrates 3 chips to handle the Intelligent MPU-401. that chip is (micro-controller + SRAM + ROM) in a single chip. you can check if that chip is running in the following way:

Re: Roland LAPC-N / I question

i.e. get the version of the running firmware inside it.

I had previously tested the MPU-401 functionality both with Roland's utility and playing MIDI via external modules (via MIDI out). That I can confirm works.

I'll run the debug commands; it would be interesting to document the firmware version for that chip.

On the repair - what about making dumps of the ROMs and see if none of them is not broken, you can find SCC-1 ROM dumps here:

Re: About Roland Virtual Sound Canvas 3

I guess most suspicious is the ProgramROM as it's running the show - but I guess that is also the easiest to replace.

Thanks for the link! I snagged those ROM dumps.

My gut feeling is that if any of the ROMs were bad, the main program ROM would be more likely than the instrument ROMs. I assumed that incorrect instrument ROM data would result in different symptoms as opposed to no output whatsoever. That may not be a correct assumption, but it's what I'm working with for now.

I was also thinking of removing the main RAM chip and seeing if the signals still behave in the same way, and then replacing it as a starting point.

Then maybe moving to the main program ROM next.

I also have replacement ICs for both the gate array and PCM generator. So unless it's the main CPU itself, I figure I have a decent shot of fixing this card.

Shponglefan wrote on 2023-12-29, 05:50:

Then maybe moving to the main program ROM next.

yeah, I also think if any of the 3 WaveROMs has a problem it will cause corrupted sound rather than no sound - but who knows, maybe the ProgramROM has CRC check for the WaveROMs and if some of the WaveROMs is bad/corrupted/missing it just stops producing any sound, i.e. it errors out.

the ProgramROM - in the forum link I sent you - we already replaced and experimented with it many times - that is how key parameters in the table here:

https://docs.google.com/spreadsheets/d/13LyKT … #gid=1739698529

were found, i.e. Patch the ProgramROM, burn new one and test to see the difference in sound (in the RAR archive from the forum link there is even a picture from a programmer doing new ProgramROM chip and what type was used). However, in SC-55 unit it's a little easier to burn and replace new chip and replace that ProgramROM, because it's DIP-32 chip compared to SOP-32 chip on SCC-1 card. Anyway, 1st is to try and dump your ProgramROM - maybe it will confirm it's good and you don't need a replacement and then it's s something else. (if the ProgramROM is confirmed as bad, you will need to find EPROM chip in SOP-32 package that is equivalent to '27C020' - I haven't searched for one and I don't know what are the compatible options).

P.S.
there are SOP-32 SMD Sockets, they are Yamaichi brand, but they are not really affordable - one socket price is kind of OK, maybe 10 bucks (and it will help such repair instead to do desolder/solder several times of those SOP-32 chips), but usually the minimum order quantity is 5 or 10 pieces, which makes it unaffordable and not economically viable. Using Socket for the ProgramROM would be useful even for working card, because then you can easily swap between any of the 3 ProgramROM versions available - V2, V3 or V4 - they are all in the RAR archive from the forum link I posted in my previous post.

I’ve had a SCC-1 with one of the wave ROMs being faulty, and the instruments from that ROM made clicks and noise. The rest sounded fine. Music that didn’t use any of those instruments were indistinguishable from playing the same music on my SC-55 which was in 100% working condition.

Tempus wrote on 2023-12-29, 10:26:

and the instruments from that ROM made clicks and noise.

thank you, that is good to know and it makes perfect sense. so, @Shponglefan can rule out all 3 WaveROM chips - they all cannot be faulty and focus on the rest...

Tempus wrote on 2023-12-29, 10:26:

I’ve had a SCC-1 with one of the wave ROMs being faulty, and the instruments from that ROM made clicks and noise. The rest sounded fine. Music that didn’t use any of those instruments were indistinguishable from playing the same music on my SC-55 which was in 100% working condition.

That's good to know. I suppose I won't know if the instrument ROMs on my card are working until I can get audio output, but at least I know it's probably not the source of the current issue.

mattw wrote on 2023-12-29, 09:19:

Anyway, 1st is to try and dump your ProgramROM - maybe it will confirm it's good and you don't need a replacement and then it's s something else. (if the ProgramROM is confirmed as bad, you will need to find EPROM chip in SOP-32 package that is equivalent to '27C020' - I haven't searched for one and I don't know what are the compatible options).

Unfortunately at the moment I don't have a way of adapting the chip to my reader. I've ordered a SOP-32 adapter, along with replacement memory chips and a couple Holtek HT27C020-70 EPROM chips just in case.

Once I get these items in, I can continue trying to diagnose the fault.

there are SOP-32 SMD Sockets, they are Yamaichi brand, but they are not really affordable - one socket price is kind of OK, maybe 10 bucks (and it will help such repair instead to do desolder/solder several times of those SOP-32 chips), but usually the minimum order quantity is 5 or 10 pieces, which makes it unaffordable and not economically viable. Using Socket for the ProgramROM would be useful even for working card, because then you can easily swap between any of the 3 ProgramROM versions available - V2, V3 or V4 - they are all in the RAR archive from the forum link I posted in my previous post.

I'll think about that. Part of me does want to keep the SCC-1 looking stock (especially if I ever decide to sell them).

But having at least one with sockets would make for easier testing and troubleshooting. I could possibly socket the RAM as well, if I go that route.

Shponglefan wrote on 2023-12-30, 05:01:

Holtek HT27C020-70 EPROM chips

those are One-Time-Programmable, but still you're in luck that you found any, because SOP32 '27C020' PROMs seem to be very rare now to find and it's better than having nothing.

So, I see your board is with chip marked with"A2", i.e. it's V2 ProgramROM. that's why I suggest in order not to waste those 'HT27C020', just burn one of them with "SCC1_R15279812 GSS2_v3_(program).BIN" - that will be an upgrade from V2 (no matter if it fixes the card or not - it will still be an upgrade when it's fixed) and V3 is the last known that is 100% SC55-compatible, V4 PorgramROM looks more like SC55mkII ProgramROM and it's from the latest revision of SCC-1A cards. I still think V4 operates as SC55-compatible mode, but we don't know (at least I don't know) and thus better as I said not waste those precious 'HT27C020', because V4 structure is SC55mkII-like.

Shponglefan wrote on 2023-12-30, 05:01:

I'll think about that. Part of me does want to keep the SCC-1 looking stock (especially if I ever decide to sell them).

But having at least one with sockets would make for easier testing and troubleshooting.

IMHO, it depends, i.e. I guess on the potential buyer, because I think having socket for the ProgramROM is a nice upgrade, especially if you bundle it together with one additional chip, for example 'HT27C020' with V3 and the original chip (if it's working) or something like that. I mean how many people with SCC-1 cards currently can change their ProgramROM - even those that can solder well and thus are not afraid to desolder the chip, probably will struggle to find something like 'HT27C020', etc. that's why I think is an upgrade, but currently more important is if the ProgramROM is the fault or not....

mattw wrote on 2023-12-30, 07:08:

So, I see your board is with chip marked with"A2", i.e. it's V2 ProgramROM. that's why I suggest in order not to waste those 'HT27C020', just burn one of them with "SCC1_R15279812 GSS2_v3_(program).BIN" - that will be an upgrade from V2 (no matter if it fixes the card or not - it will still be an upgrade when it's fixed) and V3 is the last known that is 100% SC55-compatible, V4 PorgramROM looks more like SC55mkII ProgramROM and it's from the latest revision of SCC-1A cards. I still think V4 operates as SC55-compatible mode, but we don't know (at least I don't know) and thus better as I said not waste those precious 'HT27C020', because V4 structure is SC55mkII-like.

Is it possible to switch program versions without issue?

Since there is also a ROM embedded in the CPU, the service manual states both should be replaced at the same time.