carlostex wrote:

I sent him a message again asking for whatever help he can give to us to this project, so let's hope for the best. He did software at Covox if i'm not mistaken, but maybe he can also give us a hand to maybe figure out faster what that GAL is actually doing, or better yet, get the correct logic and be done with it.

Let's hope that he can help us.

At the moment I see some problems:
1) Not enough information about traces under ICs (the same as was for SSI-2001 replica)
2) PAL/GAL with unknown functionality
3) Mounting bracket, once again. With SSI-2001 I've managed to get out with re-designing the board to be compatible with "ESS AudioDrive-like" mounting bracket, because there was one audio output and DB-15 for joystick. With this board it's a different story. We have one output, but we also have two DB-9 for joysticks and those can not be adapted for DB-15.
Variant A: remove those joystick ports and don't bother with DB-9. But how to mount the board? SSI-2001 is hold in place by DB-15 bolted to bracket. Place on this board DB-15 as well and do not connect it anywhere? A little bit dumb. Put a whole PC-joystick schematics on this board as well to get working DB-15? I think it not worth it and a little bit difficult.
Variant B: take some other standard bracket... like for 2 COM-ports. We will manage to get two DB-9 on place and mount the board. But there is no other hole for audio output on that bracket. Remove external audio jack and get only internal 4-pin header?

Has anyone thought to ask Chuck(G), over at the Vintage-Computer forum, if he might be willing to try decoding the GAL? He may need to borrow the entire Sound Master card, but if that's a concern, I should mention that I loaned him the card that was used to clone the SB2.0 C/MS PAL, and can confirm that I did indeed get it back. 😀

Fagear wrote:

Let's hope that he can help us. […]
Show full quote

Let's hope that he can help us.

At the moment I see some problems:
1) Not enough information about traces under ICs (the same as was for SSI-2001 replica)
2) PAL/GAL with unknown functionality
3) Mounting bracket, once again. With SSI-2001 I've managed to get out with re-designing the board to be compatible with "ESS AudioDrive-like" mounting bracket, because there was one audio output and DB-15 for joystick. With this board it's a different story. We have one output, but we also have two DB-9 for joysticks and those can not be adapted for DB-15.
Variant A: remove those joystick ports and don't bother with DB-9. But how to mount the board? SSI-2001 is hold in place by DB-15 bolted to bracket. Place on this board DB-15 as well and do not connect it anywhere? A little bit dumb. Put a whole PC-joystick schematics on this board as well to get working DB-15? I think it not worth it and a little bit difficult.
Variant B: take some other standard bracket... like for 2 COM-ports. We will manage to get two DB-9 on place and mount the board. But there is no other hole for audio output on that bracket. Remove external audio jack and get only internal 4-pin header?

As far as mounting brackets maybe we can get some with 3D printers. I know that this is an optimistic solution as we don't know the pricing of such service, but i would like at this point at least to try and get all the original functionality of the card.

As it is right now the mounting brackets are the least of our concerns in my opinion. So let's think about it later when we figure the other stuff out.

Cloudschatze wrote:

Has anyone thought to ask Chuck(G), over at the Vintage-Computer forum, if he might be willing to try decoding the GAL? He may need to borrow the entire Sound Master card, but if that's a concern, I should mention that I loaned him the card that was used to clone the SB2.0 C/MS PAL, and can confirm that I did indeed get it back. 😀

I obviously thought about it, but i wasn't going to ask Yvan to borrow his card to Chuck. I don't think Yvan would want to risk shipping his card, as its a very rare card and probably only 2 are know to exist (his and the guy from Idaho right?). So... That's up to him. Chuck did a great job with the SB2.0, and if he had the card he could probably even help in figuring out the hidden traces in board.

In the mean time Yvan has been really nice to us, so let's hope that maybe he can have some more time on his hands to maybe take some better pics. That might help to figure out the hidden traces at least.

And of course, if we can get some help from Nick Skrepetos it would probably help to reach our goal faster.

Well, even with all traces redrawn it would be a useless effort if you can't figure out the GAL (and a full scheme/wiring for it may not be very conclusive at all).

Well we are totally dependant on Yvan, since he is the only person with a card that we could find. If Chuck(G) from VCF would agree to dump the GAL and Yvan would agree to send him the board for this purpose the GAL problem would most likely be solved. Unless of course Yvan can do it himself, which with the limited time he has...
So let's wait on what Yvan thinks of this.

This guy has been experimenting with 3D printed ISA brackets : http://www.lo-tech.co.uk/blog/ However, I think the best solution is to take a 2 x DE-9 bracket and drill a hole in it.

I think that instead of asking Chuck G to decode the GAL, you could use the method he described to try to decode it yourself, and then ask for help if there is something you can't figure out.

The best solution for the bracket might be to simply cut those on a CNC and then use a sheet metal brake for the bends required. My CNC is almost up and running and although I didn't plan on cutting metal with it, I could give it a try. Even if it takes 20 passes to go through the metal, it doesn't matter since I'm not being charged by the minute to use my own machine. I'll build a tiny sheet metal brake too. I have enough leftover PC brackets to test on the real sheet metal.

I was being lazy and didn't want to go get my Sound Master 1 to see how the bracket is done, so I was searching online when I saw this image:
http://mail.lipsia.de/~enigma/neu/pics/covox_sm_plus.jpg

You'll notice that once again, there is a GAL16V8-25LNC on the card, right next to the FM/DAC address jumpers. I'm getting pretty confident that Covox used GALs for address decoding and nothing more.

As for the card itself, I'll see what I can do to help complete the traces and confirm part numbers and resistor values.

For the GAL, I haven't had time to check into it yet, but as I said above, it looks like it might only be used for address decoding, which would simplify the whole thing. Because if that's the case, we don't need to figure out how the GAL is programmed but simply add logic ICs to do the same job. If you look at the ISA bus pinout you'll see that the pins connected to the GAL on the SM+ are the address bits 1 to 9.

Sorry for the lack of time, but right now I'm dedicating my free time to the CNC mentioned above, which it seems we'll need for the SM1 clones (and other cards if you guys need other custom ISA brackets) - but only if it can cut metal correctly.

The other route would be to order custom-cut brackets, I have a manufacturer which I've dealt with. They can laser-cut parts and bend them with a (digital?) sheet metal brake. The parts-run starting fee is a bit high though, I'm not sure how much you guys are willing to per per brackets and how many we'd need. Maybe we could order generic brackets which we could use for all ISA projects and then use a CNC to cut the holes depending on the project.

Yvan256 wrote:

For the GAL, I haven't had time to check into it yet, but as I said above, it looks like it might only be used for address decoding, which would simplify the whole thing. Because if that's the case, we don't need to figure out how the GAL is programmed but simply add logic ICs to do the same job. If you look at the ISA bus pinout you'll see that the pins connected to the GAL on the SM+ are the address bits 1 to 9.

From what i could see from Fagear's PCB drawing and looking at the card pictures myself only 2 GAL pins connect to the ISA address bus pins, A31 and A28 which are address 0 and 3 respectively. Quite a lot less than the Sound Master + board. Other pins connect to the logic 74H**** IC's and one seems to pass through one AY8930 pin.

As far as adding other logic chips to substitute the GAL that could possibly involve moving traces and chips around to get everything to fit and work. I don't know how complicated that is but looking at the board design i think it would make the already messy original board design even messier.

In my opinion the next step is at least try to get all the traces in the board and then maybe we can solve the GAL issue, if nobody from Covox appears to give us a hand.

If a full layout has been drawn one can revert the schematic from that (knowing the GAL's function isn't critical to that) - allowing one to largely redesign the layout with ease (I've done that before ... it's cumbersome, but nice if you want to build expansions on a known working design where the schematic has been lost and only PCBs exist - which I've done for this http://i.imgur.com/RqHIbqI.png ).
If the function of the GAL is known, replacing it with standard logic isn't too hard - even tho I've got to admit I don't really see the point, except for eliminating the need of a GAL programmer (most EPROM burners have features for that).

shock__ wrote:

Nice effort!
Don't expect too much tho, I managed to find the guy who designed the SSI-2001 who actually promised me schemes & docs ... and possibly a card. After a year of "I didn't get to check yet" I stopped asking. That was a bit less than 4 years ago.

Glad to hear I'm not the only crazy who does things like this. Considering I was doing this sort of thing in 2001 and earlier, I like to picture myself as "ahead of the pack" 😁

Question: what was his name again? 😉

carlostex wrote:

Good news! […]
Show full quote

Good news!

I contacted Nick Skrepetos on facebook a few months ago. I was not sure he was the Nick Skrepetos who worked for Covox but since this person was from Eugene, Oregon i thought it was worth the shot.

Well he replied to me yesterday, confirming he is the Nick Skrepetos who worked for Covox. I sent him a message again asking for whatever help he can give to us to this project, so let's hope for the best. He did software at Covox if i'm not mistaken, but maybe he can also give us a hand to maybe figure out faster what that GAL is actually doing, or better yet, get the correct logic and be done with it.

Stay tuned.

Very curious to see if anything comes of this. 😀

Stiletto wrote:

shock__ wrote:

Nice effort!
Don't expect too much tho, I managed to find the guy who designed the SSI-2001 who actually promised me schemes & docs ... and possibly a card. After a year of "I didn't get to check yet" I stopped asking. That was a bit less than 4 years ago.

Glad to hear I'm not the only crazy who does things like this. Considering I was doing this sort of thing in 2001 and earlier, I like to picture myself as "ahead of the pack" 😁

Question: what was his name again? 😉

Jim Voigt (pbi@igc.org).
Don't exactly see the craziness in that tho 😉 It's all public info and pretty much has the best chances to get successful. Tried the same approach recently with a demotape from a band I like/liked ... got me 3 unreleased and 1 pre-release track 😀

Stiletto wrote:

Very curious to see if anything comes of this. 😀

Most likely nothing will.

Great Hierophant wrote:

This guy has been experimenting with 3D printed ISA brackets : http://www.lo-tech.co.uk/blog/ However, I think the best solution is to take a 2 x DE-9 bracket and drill a hole in it.

I think that instead of asking Chuck G to decode the GAL, you could use the method he described to try to decode it yourself, and then ask for help if there is something you can't figure out.

He's working on a clone of the Roland MPU-401 interface card, so tallying the diy clones there's adlib, innovation, sound master, covox, disney, mpu card, & cms for sb2. Need to do a pas sbpro combo card:)

awgamer wrote:

Great Hierophant wrote:

This guy has been experimenting with 3D printed ISA brackets : http://www.lo-tech.co.uk/blog/ However, I think the best solution is to take a 2 x DE-9 bracket and drill a hole in it.

I think that instead of asking Chuck G to decode the GAL, you could use the method he described to try to decode it yourself, and then ask for help if there is something you can't figure out.

He's working on a clone of the Roland MPU-401 interface card, so tallying the diy clones there's adlib, innovation, sound master, covox, disney, mpu card, & cms for sb2. Need to do a pas sbpro combo card:)

Problem with a PAS clone would be the considerable amount of custom chips - while not impossible to reverse, that would require a level of expertise I've not seen on this forum so far 😒

Which reminds me ... I've seen a bunch of unused Interwave chips on eBay a few years ago (and a bunch of chip brokers still have larger quantities of the chip available) ... considering GUS cards are quite expensive these days, how about designing a new card around them? Except for 1 PAL chip the GUS PnP seems very straight forward and could be used as a reference design (there are even more simple GUS PnP compatible cards around like the PCA761AW [Interwave + support chip + 3 74-logic ICs]). A datasheet for the Am78C201 chip also seems to be available.

The original Pro Audio Spectrum would be possible to reverse if John Neary had and could give us all the necessary information. Like shock said, a lot of custom chips in there and with quite a few GAL's as well. With no help, the amount of work would be too great. An AdLib Gold clone would be interesting if the Yamaha chips are readily available. My guess is that these chips aren't. The surround module would be easier to clone though as Cloudschatze owns one.

That said i think at this point a better retro project would be to design a Tandy 1000 Card, not PSSJ but only the 3 voice generator. We would need someone well versed in x86 assembly to try and get a way to trap the necessary ports. EMM386 is not an option.

carlostex wrote:

The original Pro Audio Spectrum would be possible to reverse if John Neary had and could give us all the necessary information. Like shock said, a lot of custom chips in there and with quite a few GAL's as well. With no help, the amount of work would be too great. An AdLib Gold clone would be interesting if the Yamaha chips are readily available. My guess is that these chips aren't. The surround module would be easier to clone though as Cloudschatze owns one.

I think that "GOLD CTRL" IC might be a custom job one may not easily get one's hand on - so that's pretty much a no-go. The card might also have more custom logic (can't find a proper high-res photo).
The surround module on the other hand is a piece of cake, also no custom chips there (just a Yamaha IC that may not be commonly available, but certainly nothing one can't find).

EDIT: Wow ... never realized the GUSPnP was a 4 layer card.

For pas I was thinking more along the lines of emulation using a microcontroller, where it emulates the card and dac interface for both pas & sbpro, otherwise it's the two opls that do all the heavy lifting. Also, has there been a solution to cms identification? Seems the cms could use the diy card treatment itself because of that. As for tandy, I thought many games expect jr/tandy video, sound, and memory mapping to work properly, much more of a pita.

carlostex wrote:

The original Pro Audio Spectrum would be possible to reverse if John Neary had and could give us all the necessary information. Like shock said, a lot of custom chips in there and with quite a few GAL's as well. With no help, the amount of work would be too great. An AdLib Gold clone would be interesting if the Yamaha chips are readily available. My guess is that these chips aren't. The surround module would be easier to clone though as Cloudschatze owns one.

That said i think at this point a better retro project would be to design a Tandy 1000 Card, not PSSJ but only the 3 voice generator. We would need someone well versed in x86 assembly to try and get a way to trap the necessary ports. EMM386 is not an option.

The original pas would be the only one worth doing as it's the only one where it's the best sound output option for a collection of games. By the time the pas16 was supported, you had your pick of the litter of other card support so it's superfluous.

If gal is not registred it's possible to read it, even if it's protected .