Great job to those who made the Gravis clone work. I admire. I arrived too late to the party, otherwise I could also been help to clear up about the data about the AMD chip and such. About a year ago I thought myself too about this, but I do not search in a net so much.

Those who are handy can build themself a good expandable and experimental programmer so you get past the LPT pin port toggling. Imagine that you have several of the LPT ports to toggle like 4 for output and 4 for input?
The simplest steps first:
Based on this idea here which I found
https://abhishekdutta.org/interfacing-with-the-isa-bus/
You can get a 8bit simple ISA card.
All components easy to get hands on!

(I developed on as a 16 bit ISA card version though with 1WS and can work with undocumented 0WS for a write operation on 486). For this I/O board I made on top myself a programmer/dumpreader. Mine is a 8x8 port version and I also tested the russian chips like K555ID7 for adress decoders and it works also great with K155 series. 😉

it can be good start for any other good things even for many debuging/testing purpose and development tool, but also elementary as a custom flash programmer.

Finally, it has gone from the rats' nest on the previous page to this (my apologies to the actual electronics guys if my attempt at layout and routing makes your brain hurt):

After the hours spent chasing odd behaviour, it turns out that most of the problems really were caused by that solderless monstrosity; even my soldering job on the flash chip doesn't seem to have been part of the problem. The basic software design seems to have been sound all along. I still have to refine the timing and add support for the chip erase and program command sequences, but that should be much less troublesome, now that I can be sure the signals are actually reaching the flash IC intact.

While interesting, I don't see how moving this work to the ISA bus would advance anything, for the moment.

good job 😀 nice to see the project advance!

Any news from shock__ on the cards themselves?

hard1k wrote:

Any news from shock__ on the cards themselves?

If that question was directed at me, I don't know anything more than you do. There was some mention of looking into in-circuit flashing and possible revisions to the board, but I am not sure if he made a decision yet. It could be that he was waiting to see if this flash experimentation showed any signs of success before putting more effort into the board layout, or maybe he just needed a little time away from staring at that board and removing and re-soldering flash chips. Until very recently, I didn't have any results that would have justified bothering him.

As of last night, chip-erase and word-program are functional. With a little more test code, I'll be able to write a full ROM file to the chip, verify and dump the contents. When that's ready, I think it will be time to start discussing what we plan to do with it.

Even with the basics already working, getting the rest of the flash programming functionality working properly took some time. The file programming operations were inexplicably failing. Sometimes, it would work as expected, and other times there would be varying amounts of corruption. It wasn't until I decided to write some more targeted test code that I isolated the symptoms: small files were fine, and there were more failures and more corruption as the files grew. The reason: I was wrong; my soldering job on the flash chip wasn't good enough to be trouble-free. The pin for the most significant bit of the address wasn't so well soldered. Now that I've rectified that, I am able to write a full file to the chip, verify it and dump the contents back to file.

As a test, I programmed and verified a full 4 MiB instrument ROM file. It would be nice to see what that sounds like, and to see what IWINIT, IWSBOS and company have to say about it, but that will have to wait.

In case we decide to use a similar design directly, I wanted it to be as simple to build as possible. That meant no programmable devices used in the design, which had a bit of an unpleasant effect on performance. Given that my test chip is only word-programmable and the command sequence must be sent for every word, the result isn't exactly pretty: a complete program and verify took around 8 minutes, 43 seconds. There is a certain amount of code clean-up I can do, but I don't expect any miracles. A complete dump to file takes about 1 minute, 40 seconds; that code can benefit significantly from a little optimisation.

I rather suspect we'll settle on a different flash chip, quite possibly the one shock__ has been using all along. I guess we'll have to make a decision soon enough.

I have to apologize for not reading the hole thread, but is this newly made gus a gus-only Card,
or does it have an additional SB/Adlib core to make it a perfect "one-card-solution"

Thx
Doc

It's an InterWave-only card (like GUS PnP). FM and Sound Blaster compatibility are offered through the old Gravis/AMD software only (slightly hardware-assisted by the InterWave); there is no other hardware added for compatibility purposes.

There was a GUS (extreme) that had additional Chips for SB/FM

Yes, this was discussed earlier in the thread, but shock__ decided that it wasn't worth the extra design effort or expense to integrate additional hardware for that purpose.

Interestingly, it turns out that the AMD 29F series uses the same command sequences (JEDEC standard, presumably) as the SST part that I used as a guinea pig. So, I adjusted the timing in my test software slightly, and modified the board to use an old 29F400 I salvaged. Here, the chip erase function is significantly slower, but because it's only a 512KiB flash, a full program takes a little over a minute, and a complete dump to file takes about 17 seconds. Since the old chip is a 120 ns part, it will be useless for the ARGUS, but it was an interesting test. A 70 or 55 ns 29F800 (or even the rather expensive 29F160) should work similarly, only a little more quickly for individual read/write cycles.

With the same basic code able to program either chip, in-circuit flashing is starting to look even more viable. Whether we will actually want to try implementing it is a different matter altogether.

It seems I've reached a bit of an impasse, and will have to suspend my efforts to contribute to this project. Everything that I need to do to make additional progress depends on access to the hardware, which I simply cannot afford at the moment. I spent a few minutes poking on eBay, just to see what the prices are like, and even placed a few bids. It became clear within minutes that, accounting for import fees, taxes and the exchange rate, all of the listings exceed my very limited budget, and this for cards that may not work to begin with, and that I would have to re-work immediately after initial testing (which, though unlikely, may also damage the card). Being an incorrigible perfectionist, this isn't the result that I would have hoped for, but it can't be helped. I can only offer my apologies to those who may be disappointed by this.

640K!enough wrote:

Everything that I need to do to make additional progress depends on access to the hardware, which I simply cannot afford at the moment.

What hardware exactly? Where in Canada are you?

the Goat wrote:

What hardware exactly? Where in Canada are you?

At this point, physical access to an InterWave-based sound card is basically essential. I have been trying for months to find one at a price that I consider reasonable, but have found nothing. Because there is a chance the card won't survive, borrowing one isn't practical.

I am in southern Quebec.

a pity you are in Canada, i live in europe so shipping costs would be really expensive, especially if customs find out the price of the boards...

As it turns out, I may not be able to do the things that need access to the hardware, but I am finding out the hard way that it will bother me to no end if I try to leave everything as it is now. Given that none of my options are particularly pleasant, I have tentatively decided to pursue the least objectionable one: I will continue to do what I can with the ROMMaker, the parallel port flash software and FFF parser. Unless something changes, the problems with MIDI playback will remain, because trouble-shooting without hardware will be impractical, and in-circuit flashing will stay on the wish list, rather than the list of implemented features.

640K!enough wrote:

As it turns out, I may not be able to do the things that need access to the hardware, but I am finding out the hard way that it will bother me to no end if I try to leave everything as it is now. Given that none of my options are particularly pleasant, I have tentatively decided to pursue the least objectionable one: I will continue to do what I can with the ROMMaker, the parallel port flash software and FFF parser. Unless something changes, the problems with MIDI playback will remain, because trouble-shooting without hardware will be impractical, and in-circuit flashing will stay on the wish list, rather than the list of implemented features.

I'm in the USA, so I'll keep my eye out for an additional Interwave card. If I had more than one, I would probably be willing to sell one at a not so absurd price. The one I have I got for $75 about a month ago on eBay.

cyclone3d wrote:

I'm in the USA, so I'll keep my eye out for an additional Interwave card. If I had more than one, I would probably be willing to sell one at a not so absurd price. The one I have I got for $75 about a month ago on eBay.

I don't know if the items you're seeing aren't available to Canadian buyers or if I just keep missing them, but I haven't found any near the price you paid. There is one I was watching, but it is steadily increasing and will probably be rather expensive by the time bidding ends. Then, of course, there is a certain Hungarian guy who always has a couple listed, but those prices are absolutely out of the question.

In other news, the FFF parser is now about as complete as I need it to be, and I have a few more ideas of things I can try to fix the MIDI playback. I haven't found a way to get that done without hardware; TiMidity++ uses the .PAT files, rather than FFF banks, and doesn't have the problem to begin with. Unless I find another way to do it, it will have to be a process of generating a ROM, installing it, listening to a series of MIDI files (under Windows, DOS and IWSBOS/MegaEm emulation, modifying the ROM and instrument settings, generating another ROM, etc., until the problems are resolved.

For now, I guess I can go back to further refining the ROMMaker and enhancing the flash programming software.

640K!enough wrote:

I don't know if the items you're seeing aren't available to Canadian buyers or if I just keep missing them, but I haven't found […]
Show full quote

cyclone3d wrote:

I'm in the USA, so I'll keep my eye out for an additional Interwave card. If I had more than one, I would probably be willing to sell one at a not so absurd price. The one I have I got for $75 about a month ago on eBay.

I don't know if the items you're seeing aren't available to Canadian buyers or if I just keep missing them, but I haven't found any near the price you paid. There is one I was watching, but it is steadily increasing and will probably be rather expensive by the time bidding ends. Then, of course, there is a certain Hungarian guy who always has a couple listed, but those prices are absolutely out of the question.

In other news, the FFF parser is now about as complete as I need it to be, and I have a few more ideas of things I can try to fix the MIDI playback. I haven't found a way to get that done without hardware; TiMidity++ uses the .PAT files, rather than FFF banks, and doesn't have the problem to begin with. Unless I find another way to do it, it will have to be a process of generating a ROM, installing it, listening to a series of MIDI files (under Windows, DOS and IWSBOS/MegaEm emulation, modifying the ROM and instrument settings, generating another ROM, etc., until the problems are resolved.

For now, I guess I can go back to further refining the ROMMaker and enhancing the flash programming software.

Yeah, the one I bought I must have found right after it was posted. The price was BIN and I bought it immediately. Same way I got a TB Tropez. Freaked out when I saw it for a crazy low price.

There is a recycling center I go to every once in a while as well. They generally don't have a bunch of old stuff, but I will check there as well in a few days. Last time I was there, I cleaned them out of the few sound and video cards they did have that were interesting.. except maybe a Geforce 7800 which I may get if they still have it.

cyclone3d wrote:

Yeah, the one I bought I must have found right after it was posted. The price was BIN and I bought it immediately. Same way I got a TB Tropez. Freaked out when I saw it for a crazy low price.

Sometimes you get lucky. I'm sure there are many people who check ebay regularly for deals like that -- many in this forum and thread!