VOGONS


Guillemot Maxi Studio Isis DOS support

Topic actions

Reply 40 of 48, by shoyoninja

User metadata
Rank Newbie
Rank
Newbie
Ozzuneoj wrote on 2025-03-24, 07:24:

Not sure if this is of any help, but I was able to get an ISA SAM9707 card working in DOS, despite it having no official DOS support.

See the edit in the OP of this thread for some highlights and what I had to do to get it working:
Dream 97PNP2 - Interesting SAM9707 prototype\sample ISA MIDI card from 2001!

Thank you! This could be useful, The 9707 reference in particular, Ill read it all carefully.

My main problem persists though, the Maestro "frontend". I cant even use the normal bank or firmware loaders for example...

Reply 41 of 48, by keropi

User metadata
Rank l33t++
Rank
l33t++

I am having an issue that the maestro is not always detected by the dos driver, what's up with that? 2 cards and different systems behave the same so it's not a coincidence...?

Edit: to be clear it's the systems not always detecting the card, not just the driver. One time the pnp devices list it the other does not

Last edited by keropi on 2025-03-24, 16:58. Edited 1 time in total.

🎵 🎧 MK1869, PCMIDI MPU , OrpheusII , Action Rewind , Megacard and 🎶GoldLib soundcard website

Reply 42 of 48, by shoyoninja

User metadata
Rank Newbie
Rank
Newbie
envagyok wrote on 2025-03-24, 06:57:

Isnt any news about this card?

I took a break from it because it started to seem pointless. The card does its things in windows, in DOS my initilizer will get the SBPRO mode working. But is quite disappointing at that... with all the issues on the mixer and the FM quality itself.

For GM in DOS its necessary to load soundbanks and find a way to route midi to the 9707, there is no direct path from the Maestro to it. As I mentioned before it should be possible to mod the hardware and connect 9707 midi in to the Maestro midi out.

But bank loading (from dos) is the part to solve next.

Last edited by shoyoninja on 2025-03-24, 17:07. Edited 1 time in total.

Reply 43 of 48, by shoyoninja

User metadata
Rank Newbie
Rank
Newbie
keropi wrote on 2025-03-24, 16:56:

I am having an issue that the maestro is not always detected by the dos driver, what's up with that? 2 cards and different systems behave the same so it's not a coincidence...?

My initiliazer I take it?

Thats strange, I did not experience that on my side. Is there any pattern to it? Ill check the detection code.

Reply 44 of 48, by keropi

User metadata
Rank l33t++
Rank
l33t++
shoyoninja wrote on 2025-03-24, 17:05:
keropi wrote on 2025-03-24, 16:56:

I am having an issue that the maestro is not always detected by the dos driver, what's up with that? 2 cards and different systems behave the same so it's not a coincidence...?

My initiliazer I take it?

Thats strange, I did not experience that on my side. Is there any pattern to it? Ill check the detection code.

nope, it's straight up BIOS/System detection
I mainly tested with p1 MMX systems, I wonder though if the card needs some special setting or something for PCI bus?

🎵 🎧 MK1869, PCMIDI MPU , OrpheusII , Action Rewind , Megacard and 🎶GoldLib soundcard website

Reply 45 of 48, by shoyoninja

User metadata
Rank Newbie
Rank
Newbie

nope, it's straight up BIOS/System detection
I mainly tested with p1 MMX systems, I wonder though if the card needs some special setting or something for PCI bus?

Oh, another feature of the Maestro-2EM?

I used a Pentium 233 MMX with Intel Chipset for testing, I did see some places reporting misbehaviour of the card on different chipsets...

Reply 46 of 48, by shoyoninja

User metadata
Rank Newbie
Rank
Newbie

Some possible paths forward:

I found another linux code for the 9407 on ISA cards, the code seems to have a fully working firmware and wavetable loader, it should be possible to make a loader from it. And then a Hardware jumper from the Maestro MIDI OUT to the 9707 Midi In.

Or: Adapt the code from SBEMU to catch 9707 ports access and route them through maestro. This should in theory allow firmware and bank loaders that already exist to work fine... Had a look at SBEMU code but it seems above what I can do now. Basicly I would need to catch MPU+0, MPU+1, MPU+2, MPU+3 and route it through MAESTRO IDMA. No hardware mod would be neeeded, but it would be a TSR.

Reply 47 of 48, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t
shoyoninja wrote on 2025-03-28, 16:54:

Some possible paths forward:

I found another linux code for the 9407 on ISA cards, the code seems to have a fully working firmware and wavetable loader, it should be possible to make a loader from it. And then a Hardware jumper from the Maestro MIDI OUT to the 9707 Midi In.

Or: Adapt the code from SBEMU to catch 9707 ports access and route them through maestro. This should in theory allow firmware and bank loaders that already exist to work fine... Had a look at SBEMU code but it seems above what I can do now. Basicly I would need to catch MPU+0, MPU+1, MPU+2, MPU+3 and route it through MAESTRO IDMA. No hardware mod would be neeeded, but it would be a TSR.

Just curious, I know there is probably a big difference between these cards (especially being PCI vs ISA) but have you tried the DOS initializer and firmware in this post?
Re: Dream 97PNP2 - Interesting SAM9707 prototype\sample ISA MIDI card from 2001!

Mainly the 94DINIT.EXE (36kB version) and 97PNP2.BIN firmware. It probably won't work, but it'd be interesting if it did.

Now for some blitting from the back buffer.

Reply 48 of 48, by shoyoninja

User metadata
Rank Newbie
Rank
Newbie
Ozzuneoj wrote on 2025-03-29, 04:48:
Just curious, I know there is probably a big difference between these cards (especially being PCI vs ISA) but have you tried the […]
Show full quote
shoyoninja wrote on 2025-03-28, 16:54:

Some possible paths forward:

I found another linux code for the 9407 on ISA cards, the code seems to have a fully working firmware and wavetable loader, it should be possible to make a loader from it. And then a Hardware jumper from the Maestro MIDI OUT to the 9707 Midi In.

Or: Adapt the code from SBEMU to catch 9707 ports access and route them through maestro. This should in theory allow firmware and bank loaders that already exist to work fine... Had a look at SBEMU code but it seems above what I can do now. Basicly I would need to catch MPU+0, MPU+1, MPU+2, MPU+3 and route it through MAESTRO IDMA. No hardware mod would be neeeded, but it would be a TSR.

Just curious, I know there is probably a big difference between these cards (especially being PCI vs ISA) but have you tried the DOS initializer and firmware in this post?
Re: Dream 97PNP2 - Interesting SAM9707 prototype\sample ISA MIDI card from 2001!

Mainly the 94DINIT.EXE (36kB version) and 97PNP2.BIN firmware. It probably won't work, but it'd be interesting if it did.

I did not, I honestly have no hopes the loader will work, its not just a slot difference (if that was the only issue it probably would work). These chips, 9407 and 9707 are designed to interface on ISA bus, and you interface them on 330h and 331h (for "normal" commands), and 332h 333h for 16 bit memory transfers. 330h is the base port.

On the ISIS on 330h you have the Maestro MPU, in fact the only thing you can talk to is the Maestro, there is no direct connection to the SAM. To even reach the 9707 you need to issue a command to Maestro, to set its GPIO to the command you want the SAM to see, the Maestro is working as the ISA slot for the SAM9707.

These commands to control the GPIO interface need to be implemented using the Maestro 2EM registers (that have nothing to do with the MPU port), so it would be really, really strange that any loader designed for the normal ISA operation can deal with it.

An example, the command 3F sets the 9707 on UART mode.

Normally on the ISA card, to set it you just have to write 0x3F to the port 330h+1 (base +1 is control).

On the Isis you write first 0x1 (that's the equivalent of 330h+1 through the ISA slot, this translation is done outside the chip on ISA implementations) to the port Maestro Base + 0x44 (Maestro GPIO Address), setting the GPIO address to the SAM9707 control register, then you write 0x3F to the port Maestro Base + 0x46 (Maestro GPIO Data).

In short you need one extra write operation to accomplish it.

But I will try it just for the sake of it... This firmware though, I can certainly load it using my initializer and and see what happens 😀. Though there is a chance it will fail due to differences on the microcode each these initializers load prior to the firmware.