VOGONS


First post, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t

Well, this is pretty neat! This popped up for sale recently and after talking to the seller for a while he made me an offer I couldn't turn down. According to the seller, this came from an estate sale. The family of the previous owner said that they didn't know much about his old stuff, but he was an electrical engineer.

What it appears to be: Either a prototype or sample of a Windows 9x "studio" oriented MIDI card based on the Dream SAM9707 synthesis chip. It came with a CD and a manual (which appears to have been printed from a document included on the CD). It is an ISA card but seems to have been made sometime after 1998, with the latest dates being in 2001.

What makes it interesting: The SAM9707 is apparently a mostly drop-in replacement for the SAM9407 which is used on many highly regarded wavetable MIDI sound cards... however the 9707 seems to be pretty much unheard of online aside from a couple datasheets (which are also included on the CD). It apparently has some updated features, but I am not sure if they are anything that would matter for the average retro enthusiast. Also, the card has 2MB of onboard sample RAM (not ROM) and allows for a single SIMM up to 16MB, which came with the card, wrapped in a little antistatic baggy.

It seems that the device itself is made by Dream and was manufactured in France. The SIM memory also has French writing on it, so I would imagine that this card, CD and RAM were sent from Dream in France directly to a developer here in the US back in 2001 or so.

20230122_213702 (Custom).jpg
Filename
20230122_213702 (Custom).jpg
File size
1.51 MiB
Views
2182 views
File license
Fair use/fair dealing exception
20230121_213101 (Custom).jpg
Filename
20230121_213101 (Custom).jpg
File size
1.61 MiB
Views
2182 views
File license
Fair use/fair dealing exception
20230121_214212 (Custom).jpg
Filename
20230121_214212 (Custom).jpg
File size
1.38 MiB
Views
2182 views
File license
Fair use/fair dealing exception
20230122_213843 (Custom).jpg
Filename
20230122_213843 (Custom).jpg
File size
1.46 MiB
Views
2182 views
File license
Fair use/fair dealing exception

EDIT: After much trial and error, the card can work easily in DOS and Windows! Here are some highlights from the thread:

The card requires -5v from the motherboard\PSU to produce sound!
MIDI in DOS breakthrough thanks to old post by Elianda
CD Image
DOS Files + Stable Windows Sound Bank Loader
It makes insane music if you use the wrong program to initialize it in DOS
... and here's why.

Thank you to everyone who was interested in this obscure old thing and helped to get it working! Being from 2001, this may be the last "official" ISA MIDI card produced that works in pure DOS. 😀

Last edited by Ozzuneoj on 2023-01-31, 18:14. Edited 4 times in total.

Now for some blitting from the back buffer.

Reply 1 of 72, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t

Here are some more close up pictures of the card.

It is also worth mentioning that the CD contains a lot of very interesting development related data, programs and documents regarding this device. It also contains information and files relating to this device's predecessor: the Dream 94PNP2. As far as I can tell, the 94PNP2 was probably identical to this (since this board actually has 94PNP2 etched into the PCB).

20230121_213107 (Custom).jpg
Filename
20230121_213107 (Custom).jpg
File size
1.36 MiB
Views
2180 views
File license
Fair use/fair dealing exception

I would assume:
94PNP2 = SAM9407
97PNP2 = SAM9707

I went into this expecting the card to be an exclusively Windows oriented device (because the manual only mentions Windows 95 support and drivers), but managed to find some very interesting things on the CD that make me think it may be possible to get it to do more... but it will likely take some work.

Sadly, any further investigation of the card's capabilities will have to wait though because I cannot get the thing to make any sound! When I plug anything into the line-out jack on the card, the volume control on my speakers does that weird thing that happens when you plug speakers into a line-in by mistake... I get a strange fuzzy sound as I turn the knob and I can see my speakers physically moving very slowly. Under Windows 98SE I can install the drivers no problem. At bootup the chart showing the devices installed in the system clearly lists this card, detected as a PNP device (though it is not functional in DOS at this point). Playing back MIDI files in Windows 98SE appears to be working, but there is no audio coming out of the card. As far as I can tell, the diagnostic test applications for the card (on the CD) show that everything passes, but when it plays audio I just get some extremely faint alternating static noises from the Line-Out (sounds more like interference from other activity on the card). The Line-In does not seem to be outputting audio either, and the "surround" port gives a fairly loud repeating scratchy noise.

20230121_213118 (Custom).jpg
Filename
20230121_213118 (Custom).jpg
File size
1.36 MiB
Views
2180 views
File license
Fair use/fair dealing exception
20230121_213122 (Custom).jpg
Filename
20230121_213122 (Custom).jpg
File size
1.4 MiB
Views
2180 views
File license
Fair use/fair dealing exception
20230121_213221 (Custom).jpg
Filename
20230121_213221 (Custom).jpg
File size
1004.74 KiB
Views
2180 views
File license
Fair use/fair dealing exception
Last edited by Ozzuneoj on 2023-01-23, 04:08. Edited 3 times in total.

Now for some blitting from the back buffer.

Reply 2 of 72, by darry

User metadata
Rank l33t++
Rank
l33t++

This is awesome!

Does the CD contain any stuff that nobody (outside of Dream or someone who might have signed an NDA) has heard of before (soundbanks, datasheets, software, etc ) ?

Reply 3 of 72, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t
darry wrote on 2023-01-23, 03:38:

This is awesome!

Does the CD contain any stuff that nobody (outside of Dream or someone who might have signed an NDA) has heard of before (soundbanks, datasheets, software, etc ) ?

The sound banks may be the same ones available elsewhere, I haven't really been able to do much with it since I can't get sound out of it... but... it seems likely that there are several other files here that are not available anywhere else, including board schematics, datasheets, GERBER and ORCAD files, etc.

I have no idea what the status of this stuff is in 2023, but here is a ~46MB zipped ISO of the CD:

*LINK DOWN* See here.

It's a temporary link, so if anyone has a more reliable place to host it for those interested, please feel free to do so. I will be removing this link once it is down.

If anyone has any suggestions regarding the lack of audio output and the strange scratchy sound coming from my speakers when I adjust volume while connected to this card, I'd love to get it working. The board schematic (97PNP2.BMP) on the CD may help to diagnose it, but I'm not that good with circuits.

Also, there is one small bodge wire change on the card that I can spot... near the PCM3001E by the clock crystal (see the attached picture). As far as I can tell, the small round solder pad near the end of that wire has had a trace severed deliberately, and the trace goes under the PCM3001E diagonally. The bodge wire is soldered to that trace directly... seemingly to avoid connecting it to the tiny round solder pad.

20230122_225559 (Custom).jpg
Filename
20230122_225559 (Custom).jpg
File size
1.31 MiB
Views
2153 views
File license
Fair use/fair dealing exception
Last edited by Ozzuneoj on 2023-01-31, 01:24. Edited 1 time in total.

Now for some blitting from the back buffer.

Reply 4 of 72, by Tiido

User metadata
Rank l33t
Rank
l33t

Very cool find !

That CD is full of very interesting things ~

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 5 of 72, by keropi

User metadata
Rank l33t++
Rank
l33t++

Indeed a cool find!
I would imagine you have to upload a soundbank in ram in order to get sound from this , does the upload function completes successfully?

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

Reply 6 of 72, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t
keropi wrote on 2023-01-23, 06:06:

Indeed a cool find!
I would imagine you have to upload a soundbank in ram in order to get sound from this , does the upload function completes successfully?

Yes, I believe it would require a sound bank, which the Windows drivers most likely load at startup... but the actual line-out seems to be either defective or misconfigured. As mentioned, the line-out responds the way a line-in does if you accidentally connect speakers to it, and it starts doing this as soon as the system is powered on.

Also worth mentioning, for those who are investigating the contents of the CD:
The hardware\PNP\ folder contains some really interesting stuff, but it does not appear to work at all with this particular card. I believe these are left-overs intended exclusively for the 94PNP2, not the 97PNP2. Basically, to configure the 94PNP's resource settings (presumably for DOS support), you have to write the information to a PNP EPROM on the card. The EPROM chip is the tiny square 95MS14 near the ISA connector. To do this, the "94PNP2.PNP" file must be edited with a text editor, then you run a .bat file which executes some external Microsoft Assembler and EXE2BIN applications (had to download these) to assemble and convert that file to 94PNP.BIN (ends up being only ~161 bytes in size). Then you run EEWRITE.EXE to write that file to the 95MS14 EPROM, which is supposed to change the configuration and resource allocation of the card for the next reboot.

It looks to me like the EPROM on this card has been resoldered, possibly by hand, so it could be a slightly different version from the one on the 94PNP2 or it could just have been replaced at some point. And it is worth mentioning that when experimenting with the above process to see if the card could actually initialize and be detected in pure DOS, it appears that the EEWRITE application does not interact with this card in any way. It prints ???? on the screen, says "erasing"(yeah, I actually felt pretty dumb\scared when I saw that), but never proceeds past that point and requires a reboot. Rebooting the system shows that the card is still detected as a PNP device and it is still detected in Windows, so it appears to not be able to read or modify the EPROM on this card at all.

After looking around a bit more, I realized that the 97PNP2.BIN file (in the root directory of the CD) which is nearly 1000x the size of the 94PNP2.BIN (in the PNP directory) is described in the manual as being the firmware for the card. Sadly, there doesn't appear to be any firmware editing or flashing tools\instructions for this newer version. Also, the manual says there should be a Firmware folder within the Hardware directory, but it does not exist on this CD... 😢

Any attempt to get the card working in DOS is either going to require messing with some software from other devices that have a similar setup (Dream SAM9x07 and PNP chip of some kind) or figuring out some way to modify the assembled 97PNP2.BIN file directly... which I highly doubt anyone with these skills will want to do, since this may be the only one of these that anyone will ever try to do this with.

But first, I need to figure out the line out issues.

EDIT: I just found this in the "whatsnew" file on the root of the CD:

" * The firmware updates at March 15, 2001 (compatible with 97x3Flsh)"

Googling 97x3Flsh actually turns up several references to the program written in Atmel\Dream datasheets, so it is possible that this program exists somewhere else, if it ends up being useful for configuring the card for DOS support.

Now for some blitting from the back buffer.

Reply 7 of 72, by keropi

User metadata
Rank l33t++
Rank
l33t++
Ozzuneoj wrote on 2023-01-23, 06:55:

Yes, I believe it would require a sound bank, which the Windows drivers most likely load at startup...[...]

I would not be so sure about that, is there a tool to manually load a bank?

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

Reply 8 of 72, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t
keropi wrote on 2023-01-23, 07:34:
Ozzuneoj wrote on 2023-01-23, 06:55:

Yes, I believe it would require a sound bank, which the Windows drivers most likely load at startup...[...]

I would not be so sure about that, is there a tool to manually load a bank?

In the SNDEDIT folder there is an installer for the sound bank editor. The installer appears to be broken (at least on my system) so it deletes it's own temporary files from the windows Temp folder before it can install them, so after the installer extracts the files I had to quickly copy them from to another folder before they were deleted... THEN I was able to properly install the bank and instrument editor.

The bank editor does provide a program to load and change banks, and when I opened it it said it currently had a 1MB bank loaded. That bank file is located in the install directory when installing the drivers, so I assume it just loads that when the Windows driver initializes. I loaded the Roland 4MB GS samples and it took about one second and they were loaded with no errors.

Like I said, it "plays" music as well as digital effects in Windows without complaining but I simply can't hear anything due to the screwed up line out.

Even when I run the DOS sound setup program for games in a DOS box (98SE), it detects devices and tests them without crashing... I just never hear anything. Curiously, when I run the Daggerfall sound setup in Windows (a good test because it supports so many sound devices) it auto-detects the digital sound device as a Roland RAP-10 at port 330 with DMA 5, and it doesn't crash when I tell it to test digital sound. Even dxdiag sound test goes through all the various tests without errors or crashing.

So, this is a full fledged soundcard that works in Windows 98... I'm just not getting any audio from it.

I also found some other very cool things on the CD:

*A 1997 letter from Dream to their customers regarding the Roland lawsuit being resolved and what that means for customers. Also includes 1MB and 4MB GM and GS banks.

*Some Dream advertising "Demos" for upcoming products... one of which is a Dream SAM9777, which is supposed to be a full fledged PCI sound card chip. Googling it turns up a datasheet, but I don't believe this ever made it to market?

*Some source code for something... I don't know what exactly. It's in the Software folder.

Now for some blitting from the back buffer.

Reply 9 of 72, by keropi

User metadata
Rank l33t++
Rank
l33t++

well... I'm out of ideas now
maybe the card was damaged in the past

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

Reply 10 of 72, by Rawit

User metadata
Rank Oldbie
Rank
Oldbie

Awesome discovery. Neat to see a mention of the Roland/Dream conflict in the PDF of the soundbanks.

Is there any hint of audio coming out of the card or only "computer thinking noises"?

Edit: This manual shows jumper positions: https://archive.org/details/manualzilla-id-57 … ge/n13/mode/2up

This doesn't seem to match with the config on your card.

YouTube

Reply 11 of 72, by Bancho

User metadata
Rank Oldbie
Rank
Oldbie

This is really awesome. Like a Dream Goldfinch! I've downloaded the ISO as it may help with something I've wanted to try and do!

I have this board which I picked up off eBay a long time ago. I don't know anything about it but (maybe out of a keyboard or synthesizer) but I'd love to try and turn it into a Wavetable board of some sort. It has a Cleanwave Rom and two 9707s!

7RQF9QIh.jpg

Reply 12 of 72, by dionb

User metadata
Rank l33t++
Rank
l33t++
Ozzuneoj wrote on 2023-01-23, 06:55:

[...]

the line-out responds the way a line-in does if you accidentally connect speakers to it, and it starts doing this as soon as the system is powered on.

Really stupid but simple thing to check:

Is it possible that the ports are mislabeled and you are actually sticking speakers into line-in?
What happens if you try the other 3.5mm options?

Reply 13 of 72, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t
Rawit wrote on 2023-01-23, 10:11:
Awesome discovery. Neat to see a mention of the Roland/Dream conflict in the PDF of the soundbanks. […]
Show full quote

Awesome discovery. Neat to see a mention of the Roland/Dream conflict in the PDF of the soundbanks.

Is there any hint of audio coming out of the card or only "computer thinking noises"?

Edit: This manual shows jumper positions: https://archive.org/details/manualzilla-id-57 … ge/n13/mode/2up

This doesn't seem to match with the config on your card.

I did realize last night that if I turn the volume on my speakers way up and click the volume adjuster in Windows I can hear a very faint and raspy hint of that "ding!" sound... so something is coming through, but it's like there is a really bad connection or short somewhere.

As for the jumpers, those are just to select the memory configuration. I had tried it with and without the 16MB SIMM installed, so I have changed the jumpers back and forth for that. Good suggestion though, and I'm surprised that that manual actually exists online! That seems to be very similar to the manual on the CD I received. It's interesting that it still refers to all of the development stuff though. I think it's safe to say this wasn't a design that made it into consumer channels and was only ever sent to developers.

dionb wrote on 2023-01-23, 19:17:
Really stupid but simple thing to check: […]
Show full quote
Ozzuneoj wrote on 2023-01-23, 06:55:

[...]

the line-out responds the way a line-in does if you accidentally connect speakers to it, and it starts doing this as soon as the system is powered on.

Really stupid but simple thing to check:

Is it possible that the ports are mislabeled and you are actually sticking speakers into line-in?
What happens if you try the other 3.5mm options?

That was actually the first thing I checked, since they actually aren't labeled on the back plate of the card... but there wasn't clean output from any of them.

I have also tried plugging a jack into the line-in while using line-out for speakers, because I know most of these types of circuits have some kind of jack-detection which changes the circuit when a device is connected. Didn't seem to make a difference either, but it still may be somewhat related.

I'm wondering now if maybe the line-out jack is actually bad? I have one of those big Intergraph keyboards with the built-in speakers that I tried to diagnose for a while before realizing it was just the jack itself that was bad. I replaced that and it has worked fine.

Bancho wrote on 2023-01-23, 18:30:

This is really awesome. Like a Dream Goldfinch! I've downloaded the ISO as it may help with something I've wanted to try and do!

I have this board which I picked up off eBay a long time ago. I don't know anything about it but (maybe out of a keyboard or synthesizer) but I'd love to try and turn it into a Wavetable board of some sort. It has a Cleanwave Rom and two 9707s!

Wow, that is so cool! There are definitely some possibilities there.

With the SAM9707 chips you could very well make a couple of sound cards for yourself. I'm not going to say anymore as to what may be possible with the files on this CD, but... well, yeah... you can probably figure that out. 🤣

I also thought that it was like a Dream based Goldfinch at first... but the SAM9407\9707 surprisingly seem to have extensive digital\wave playback and mixing capabilities as well. I don't think they were generally used for that on sound cards though because they were almost always paired with some other chip that handled all of that.

Do you have an EPROM reader? Maybe you could dump those two ROMs to find out what that board came from. Could have been a keyboard or maybe a jukebox\karaoke type of thing.

Now for some blitting from the back buffer.

Reply 14 of 72, by Kamerat

User metadata
Rank Oldbie
Rank
Oldbie

Just a quick question to Ozzuneoj: Does your PSU supply -5V? Looks from the picture of the back side that the card have the -5V line connected and it might be the cause for a silent card if it's missing.

DOS Sound Blaster compatibility: PCI sound cards vs. PCI chipsets
YouTube channel

Reply 15 of 72, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t
Kamerat wrote on 2023-01-23, 23:21:

Just a quick question to Ozzuneoj: Does your PSU supply -5V? Looks from the picture of the back side that the card have the -5V line connected and it might be the cause for a silent card if it's missing.

That is SO FUNNY! I just found that myself!! I was just going to post here that I'd fixed it less than 5 minutes ago by adding a -5V power adapter to my power supply!

I was looking at the schematic and saw that the TL072C Op-Amp had a -5v pin, and sure enough it traced all the way back to the -5v pin on the ISA connector. That seems strange now in terms of "retro" hardware because -5v was more common on much much older cards (PAS 16 and SB 2.0 come to mind), but at the time when this board was designed in 1996-1998 and even any of the later revisions in ~2000, it was still perfectly normal to expect a -5v line to be available on a power supply.

Anyway, this thing sounds great! Digital sound, MIDI... it's all there! I will tinker with it some more and post my findings.

Now for some blitting from the back buffer.

Reply 16 of 72, by Kamerat

User metadata
Rank Oldbie
Rank
Oldbie
Ozzuneoj wrote on 2023-01-23, 23:29:

That is SO FUNNY! I just found that myself!! I was just going to post here that I'd fixed it less than 5 minutes ago by adding a -5V power adapter to my power supply!

🤣 Glad you got it working. 😁

DOS Sound Blaster compatibility: PCI sound cards vs. PCI chipsets
YouTube channel

Reply 17 of 72, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t
Kamerat wrote on 2023-01-23, 23:38:
Ozzuneoj wrote on 2023-01-23, 23:29:

That is SO FUNNY! I just found that myself!! I was just going to post here that I'd fixed it less than 5 minutes ago by adding a -5V power adapter to my power supply!

🤣 Glad you got it working. 😁

Thank you for the suggestion though. If I hadn't thought of that at exactly the same time as you, then it would have been solved after reading your post. 😁

Now, on to the next quirk...

I just tried loading different GS and GM banks to the card and most of the time it crashes Windows when doing this...? I can load the 1MB bank easily, remove it, add it again, remove it, add it again, over and over. But so far I've only managed to get the 4MB GS or 4MB GM bank to load once out of maybe 5-6 tries. Last night I tried to load it once and it worked without a hitch so I figured it was fine.

EDIT: Okay, I guess it's only certain banks that are having problems? gmbk9732.94b will not load without crashing the system in some way. Same with GSSBK080.94B. And yet GSSBK320.94B seems to work just fine? Even if I load it more than once into the RAM.

After the blue screen it will usually go back to the desktop and pop up an error window that says:

"Error in loading bloc in card at address fc000" or "...address 74000" etc. ... then the system hard locks.

I tried a 4MB stick of RAM and tried it with and without changing the memory size jumper and it still crashed. Interestingly, it reported the correct size even without the memory size jumper changed from 16MB to 4MB, so I'm not sure exactly what that jumper does. EDIT: Oh, Windows seems to freeze at a blank screen with a flashing white cursor if the jumper is set to 4MB with a 16MB SIMM installed. Good to know.

Now for some blitting from the back buffer.

Reply 18 of 72, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t

Well... it seems to be somewhat unpredictable as to what sound banks it will allow without crashing. I just tried one from the 94b sound bank collection called "analogs.94b" and it crashed on that one too. Very strange.

Now for some blitting from the back buffer.

Reply 19 of 72, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t

I found an interesting Korean page with some Hoontech cards that are listed as 97PNP... though that may just be a coincidence as they are quite different cards overall.

https://m-blog-naver-com.translate.goog/spc99 … &_x_tr_pto=wapp

Now for some blitting from the back buffer.