VOGONS


First post, by shock__

User metadata
Rank Oldbie
Rank
Oldbie

So let's start the year with something that might possibly be interesting for a few of you guys, as written in my introduction post.

I've recently found someone who is in possession of above card and might even trade/donate/sell the card to me. Considering the card seems to have virtually ceased to exist, with only 2 of them being known to me (apart from a few bbs/newsgroup posts from ~1990 or earlier) I've taken the duty upon me to start redrawing the card, so a scheme can be adapted from it later on.

Here is the current progress, which sadly still has gaps in it, considering I'm doing this based upon photos, with components/chips hiding some traces below them 🙁

Little explaination how to interpret this:
Violet circles: Holes for putting chips/sockets/parts in or VIAs (holes connecting front and back layers)
Blue traces: Front layer traces (sadly incomplete currently)
Red traces: Back layer traces
Violet traces: Front and back layers "overlapping"
White rectangles: chips (with Pin1 being marked and their identifiers)
Components like resistors/resistor network/caps/jacks aren't drawn in yet

Luckily enough, the way the SID-chip, which is pretty much the heart of the soundcard, is glued to the ISA slot can be seen already. Sadly enough this concludes the analog registers of the SID aren't incorporated in the design of the card and the adress decoding and implementation of the joystick port is still obscured. If you wanted to built a compatible on perfboard, one should be able to do so already tho, with a little thought. Once/if I get to finish above redrawing, a scheme will be posted, along with (if there's enough interest) eagle files so one can get this own SSI 2001 clone on a proper, neat PCB.

Funnily enough it seems the card indeed never was produced in larger quanitites, as the 2 cards I've seen pictures of have some minor changes to them. First of all, the card I got pictures of still has a wire thrown on top fixing a missing trace, secondly enough _both_ versions seem to have wrong caps attached to the 6581 SID chip (was 470pF in the C64, is 0.1uF on the SSI 2001), the amplification circuit being minorly different (1K-Ohm/10K-Ohm voltage splitter (which is how it's done in the C64 as well) vs. 10K-Ohm/10K-Ohm splitter) as well as the card I've seen have a "full" ISA slot connector (meaning all "gold fingers" are present) rather than just having "fingers" for the actual lines used.

Hope this might be of interest for one or another user and/or might help DOSBox development at one point, even tho it seems to be implemented just right already.

P.S.: The photos I got will be posted once the guy who currently owns the card has decided if he wants to keep it or not.

Reply 2 of 590, by shock__

User metadata
Rank Oldbie
Rank
Oldbie

Penya's card is the one with the "incomplete" ISA connector and an amplifier based upon the original C64 scheme.

Also before I forget ... if anyone wants to see just another reason why the card never took off ... take a look at "Computer Gaming World Issue 63" which has a review (including a preview of the soundblaster 1.5 still being called "killer card")
(Full scans can be found here -> http://cgw.vintagegaming.org/index.php)

Last edited by shock__ on 2011-01-11, 19:52. Edited 1 time in total.

Reply 3 of 590, by Tetrium

User metadata
Rank l33t++
Rank
l33t++

A similar process was done on vintage-computing.com. They created an 8 bit IDE controller from scratch!

Whats missing in your collections?
My retro rigs (old topic)
Interesting Vogons threads (links to Vogonswiki)
Report spammers here!

Reply 4 of 590, by Zup

User metadata
Rank Oldbie
Rank
Oldbie

How about mixing this design into the card? SID chips are scarce these days... and reliable SID chips are almost inexistant.

They claim that it's a drop in replacement for SIDs, but I guess it can be integrated without a daughterboard. Also, redesigning the board as a PCI card would allow to use it in modern hardware (if you could find SID drivers for Windows 7 xD )

Don't take my comment seriously... I think it's an interesting project, but I'm a Spectrum owner, not a C64 or SID interested person.

I have traveled across the universe and through the years to find Her.
Sometimes going all the way is just a start...

I'm selling some stuff!

Reply 5 of 590, by shock__

User metadata
Rank Oldbie
Rank
Oldbie
Tetrium wrote:

A similar process was done on vintage-computing.com. They created an 8 bit IDE controller from scratch!

Well, this isn't exactly what I'd call from scratch 😉 Just like how doing a page in a coloring book isn't exactly drawing a picture.

Zup wrote:

How about mixing this design into the card? SID chips are scarce these days... and reliable SID chips are almost inexistant. They claim that it's a drop in replacement for SIDs, but I guess it can be integrated without a daughterboard.

While the SwinSID is a kick ass project, it's certainly not perfect and IIRC aims to replace the 8580 rather than the 6581. Putting just a socket on the final "SSI 2001 clone" would give the user the power to decide themself what kind of chip to use, even tho the SwinSID might require some voltage adaption (12V vs. 9V). Afterall after the scheme (and possibly pcb files) I never plan to actually produce them in larger quantities, considering the SSI 2001 barely has any software/games which support it, yet I doubt there are too many people caring about developing software for some outdated, proprietary and not even properly implemented piece of hardware, not to forget people actually prefering software rather than hardware emulation/cloning. This "project" is mainly aimed at purists, hardware historians or people looking at a fun soldering project for a rainy weekend ... which might be very, very few, if any.

Zup wrote:

Also, redesigning the board as a PCI card would allow to use it in modern hardware (if you could find SID drivers for Windows 7 xD )

Also if you want a hardware based solution for the SID on a PC, one _should_ always go for the HardSID-Quadro (4 SIDs on a PCI card), the HardSID4U (same as before but USB based) or the Catweasle (Floppy controller for the PC/Amiga supporting all these random floppy formats various homecomputers used, which also has a socket for a SID [or even 2]). Those are really well designed and have good software support, are commercially available (even tho they're not cheap) and professionally made. No need to re-invent the wheel, considering even if you wanted to go with a DIY approach, there's still a SID interface for the parallel port which has some software support one could solder in probably less than an hour or slamming your trusty C64 to the serial port of your PC to get authentic output. Sadly neither the of these solutions are compatible with the SSI 2001 (as the SSI 2001 isn't exactly implementing the SID as it was in the C64 anyways) ... at least to my knowledge.

Zup wrote:

Don't take my comment seriously... I think it's an interesting project, but I'm a Spectrum owner, not a C64 or SID interested person.

Offtopic: The ZX kicks ass (in it's own way), I actually own a DIY clone from russia, incorportating various improvements on the original Spectrum (Google: "Pentagon 1024SL 2.2") ... that one was fun to build and still gets some use here on and off.

Reply 6 of 590, by elianda

User metadata
Rank l33t
Rank
l33t

Well, I would prefer using a software emulation engine as modern SID replacement. Current implementation such as in JSIDPlay2 have made a considerable progress in comparison to earlier emulations. You can even choose on the fly what filter characteristic your emulated 6581 should use.
( http://jsidplay2.sourceforge.net/ )

For using real SIDs there are also some disadvantages:
- since every 6581 sounds different (due to manufacturing) you have to test hear several chips before choosing a suitable sounding 6581.
- In most cases real SIDs are ripped off working C64s which renders the C64 defect. An approach with some serial connection that keeps the C64 in working condition would be much better.

I think SwinSID goes already in the right direction. From my first impression though it seems to have some problems due to the limited computing power of the Atmel AVR.

Reply 7 of 590, by shock__

User metadata
Rank Oldbie
Rank
Oldbie
elianda wrote:

Well, I would prefer using a software emulation engine as modern SID replacement. Current implementation such as in JSIDPlay2 have made a considerable progress in comparison to earlier emulations. You can even choose on the fly what filter characteristic your emulated 6581 should use.
( http://jsidplay2.sourceforge.net/ )

As written above, that pretty is the best approach, especially with all the research being done for the reSID engine (slightly offtopic: weren't you involved in the DIE scanning process of various SIDs?)

elianda wrote:

For using real SIDs there are also some disadvantages:
- since every 6581 sounds different (due to manufacturing) you have to test hear several chips before choosing a suitable sounding 6581.

Well, from my knowledge all SIDs used on the SSI 2001 were 6581R4 chips, dated 0687 ... so that can be ruled out, to a degree that can be neglected ... but as written before they seem to have put wrong filter caps on the SSI 2001

elianda wrote:

- In most cases real SIDs are ripped off working C64s which renders the C64 defect. An approach with some serial connection that keeps the C64 in working condition would be much better.

Agreed there. In fact I actually wonder how Innovation got their SID chips ... did Commodore/MOS ever sell them to 3rd party hardware developers? Tearing SIDs out of (broken) C64s doesn't exactly seem like an efficient way, considering the card sold for $80 I think.

elianda wrote:

I think SwinSID goes already in the right direction. From my first impression though it seems to have some problems due to the limited computing power of the Atmel AVR.

The approach certainly is good ... just like the SIDStick, which is basically a portable SID player based upon emulation on a parallax propeller microcontroller, which should have enough power for faithful SID emulation. I'm gonna build one of these for a friend some time soon, so I can make recordings before giving it to him, so the SID-ophiles can comment on the quality of emulation 😉

Reply 8 of 590, by elianda

User metadata
Rank l33t
Rank
l33t
shock__ wrote:
elianda wrote:

Well, I would prefer using a software emulation engine as modern SID replacement. Current implementation such as in JSIDPlay2 have made a considerable progress in comparison to earlier emulations. You can even choose on the fly what filter characteristic your emulated 6581 should use.
( http://jsidplay2.sourceforge.net/ )

As written above, that pretty is the best approach, especially with all the research being done for the reSID engine (slightly offtopic: weren't you involved in the DIE scanning process of various SIDs?)

(slightly offtopic: Yes, I did this stuff http://mail.lipsia.de/~enigma/neu/6581.html )

elianda wrote:

For using real SIDs there are also some disadvantages:
- since every 6581 sounds different (due to manufacturing) you have to test hear several chips before choosing a suitable sounding 6581.

Well, from my knowledge all SIDs used on the SSI 2001 were 6581R4 chips, dated 0687 ... so that can be ruled out, to a degree that can be neglected ... but as written before they seem to have put wrong filter caps on the SSI 2001

Wrong filter caps ofcourse influence the filter characteristic. But sticking to the revision of the 6581 to distinguish certain filter characteristics even with correct filter cap values is an urban legend. The 6581 revision does not tell you how the filter curve goes. Just take a look at the curve profiles in the configuration dialog of jsidplay2 that are measured from real SIDs.

Luckily there is still an active community working on C64 emulation / reverse engineering. Though f.e. I don't really get feedback which commercial companies took a look on the SID dies on my page for developing emulation engines. I know that there are some VST plugins around that claim to have the 'best' SID emulation available...

Reply 9 of 590, by Tetrium

User metadata
Rank l33t++
Rank
l33t++
shock__ wrote:

Well, this isn't exactly what I'd call from scratch 😉 Just like how doing a page in a coloring book isn't exactly drawing a picture.

So what did they base it on? This IDE controller supposedly works with much larger harddrives then any typical IDE drives.

Reply 10 of 590, by Zup

User metadata
Rank Oldbie
Rank
Oldbie

Why not? IDE controllers are really dumb... the 8 bit IDE controller for Spectrum only uses a TTL chip. The complexity of IDE interface is put into the software and the hard drive.

Do you want to use IDE drivers larger than 528 Mb? OK, change the software. Hardware is not an issue here.

I have traveled across the universe and through the years to find Her.
Sometimes going all the way is just a start...

I'm selling some stuff!

Reply 11 of 590, by Tetrium

User metadata
Rank l33t++
Rank
l33t++

Ah! Now I see, cheers 😉

Whats missing in your collections?
My retro rigs (old topic)
Interesting Vogons threads (links to Vogonswiki)
Report spammers here!

Reply 12 of 590, by shock__

User metadata
Rank Oldbie
Rank
Oldbie

Since the guy I got in contact with apparently wants to turn the card into $$$ here are the shots I've got. Minor editing has been done to it, as the original photographs were slightly curved. Be warned, the pictures are 1MB each.

Enjoy

Front
Back

Also the CGW Article

Reply 13 of 590, by h-a-l-9000

User metadata
Rank DOSBox Author
Rank
DOSBox Author

When reading this thread I remembered having some MOS chips somewhere. Turns out three 6581 are amongst them, two 6581 CBM from week 6 1985 and one 6581R3 week 7 1986. Well, so it wouldn't be an exact copy of the Innovation card but my project won't be anyway.

A week or so later the prototype was operational. For pictures, see
http://archiv.to/GET/FILEB81F2EA23D995 (1,97 MB, three jpeg files).

One of the CBM chips turned out to be bad.

I've also recorded some samples:
http://archiv.to/GET/FILEBDFA7CC7DC6A5 (27,74 MB, wav files of U6, MechRevenge, and F-19).

There are some potential problems with the clock/wait state/access time synchronitation cirquit, though I've tried a few I/O clock and waitstate settings, couldn't hear anything obviosuly wrong.

I'm attaching the schematic here, which should resemble the "prototype" to like 95%. I hope there are no mistakes in there 😀 as I've made the board with partial hand-drawn sketches first and then created this schematic mostly from my head.

So now, what to do with it? xD
These few games that support it don't match my interest, the sound has a certain charme though.

Edit: updated schematic, was not finished.

Attachments

  • Filename
    SIDPC_schematic_02.png
    File size
    65.37 KiB
    Downloads
    415 downloads
    File license
    Fair use/fair dealing exception

1+1=10

Reply 14 of 590, by elianda

User metadata
Rank l33t
Rank
l33t

Hmm why did you used 1nF caps instead of 470 pF ?

As for the signal timing, wasn't there some common design to convert from Motorola to Intel?

Retronn.de - Vintage Hardware Gallery, Drivers, Guides, Videos. Now with file search
Youtube Channel
FTP Server - Driver Archive and more
DVI2PCIe alignment and 2D image quality measurement tool

Reply 15 of 590, by h-a-l-9000

User metadata
Rank DOSBox Author
Rank
DOSBox Author

The 6581 datasheet says 1nF, shock__ says 0.1µF, which seems a bit much, so I settled for 1nF. Maybe it doesn't even play a role for the 3 or 4 games that support the card.

This chip doesn't have the typical Motorola interface, the Ø2 signal is both enable and clock, so more effort is required.

1+1=10

Reply 16 of 590, by shock__

User metadata
Rank Oldbie
Rank
Oldbie

So ... another exhausting session of google research has been done.

- I've managed to track down a programmer at Spectrum Holobyte who wrote in a 2007 interview that he programmed for this card (even described it a little) ... maybe he has some more infos on it.
- Found the President of American Innovations which seems to be the current name of "Innovation Computer Corp." Maybe he has some docs or possibly such a card left.
- Tracked down Penya, who borrowed his card to a friend, who might help me out drawing all the missing traces into the image in my original post.

So let's hold our thumbs this turns into something.

Reply 17 of 590, by shock__

User metadata
Rank Oldbie
Rank
Oldbie

So ... another minor update:
- Video of the SSI-2001 in action by Penya's friend DemonBob -> http://www.youtube.com/watch?v=hZHYBS6xAK4&fe … player_embedded
- Spoke to Kevin Seghetti of Spectrum Holobyte (who wrote the driver for said card for the Falcon A.T. game). He sadly doesn't have such a card anymore, but mentioned that Innovation was supposedly working on a sound card with dual SIDs (just like the SoundBlaster Pro has 2 OPL2 chips). Trying to verify that now.
- I'm currently in contact with the designer of the SSI-2001, but didn't get any answers yet, as he was on a business trip til last monday. Quite excited what this might uncover. Already prepared some kind of interview, covering various topics that might be interesting from a retro-enthusiasts view ^^

Reply 18 of 590, by DBob

User metadata
Rank Newbie
Rank
Newbie

Some sample:

Ultima VI - intro, menu, story, gameplay
F-19 - intro

Tracked down Penya, who borrowed his card to a friend, who might help me out drawing all the missing traces into the image in my original post.

What kind of information do you need exactly? PM me with the details pls.

Reply 19 of 590, by Great Hierophant

User metadata
Rank l33t
Rank
l33t

I can see that the Innovation card does not have a clock crystal and therefore the 6581 must use the OSC signal, which runs at 14.318180MHz. In the Commodore 64, the CPU and SID both run at 1.022MHz and that machine uses the same clock crystal in its NTSC models. You can get a 1.022MHz clock by dividing the 14.318MHz signal by 14. With the limited logic on the Innovation board, which must be shared with the Gameport, can this be done?

Second, there are three sets of jumpers on the card. The group of four above the right 138 must be to control the addressing of the SID, starting at 280 and working its way up to 2A0, 2C0 or 2E0. I would guess that the lone jumper above the left would be to enable or disable the gameport. The group of three jumpers above the 558 would presumably be to change the gameport speed, which I recall was a feature of this particular card.

If the CAP1A-1B values were 470pF in the C64 and 0.1uF (100,000pF) on the SSI 2001, what kind of effect would that have on the filter output?

http://nerdlypleasures.blogspot.com/ - Nerdly Pleasures - My Retro Gaming, Computing & Tech Blog