Innovation SSI 2001 re-engineering and/or conclusions

Discussion about old sound cards, MIDI devices and sound related accessories.

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby Tiido » 2018-8-16 @ 11:51

HASL will destroy the gold plating in the slots over time though galvanic corrosion. There is a lot of scientific information on the matter.

Other thing is that HASL plating is typically fair bit thicker than the board itself and from my experience (with game console cartridge slots) this alone is enough to make the connectors unreliable, I can see the gold layer buldozed off after just a few insertions and the exposed materials corrode quite quickly (within days I see discoloration already). After that point the game is likely to crash randomly or not run at all until I wiggle the board around in the slot a bit to dislodge the new corrosion on both the tin plated contact and the damaged original slot connector. The effects are permanent and the connector is never reliable again.
It helps that in a PC you usually leave the thing in the connector effectively permanently, with mechanical welding preventing some of the corrosion related problems happening but how well it lasts still depends on surrounding atmosphere and other factors.

Another thing is to have beveled egde, it will make the insertion cycles much more gentle on the connector. There's lot less downward compression force in the beginning of insertion for the contacts, instead the angled edge helps to push the contacts away so the board can slide in.
User avatar
Tiido
Oldbie
 
Posts: 1066
Joined: 2018-1-14 @ 04:40
Location: Estonia

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby Great Hierophant » 2018-8-16 @ 14:30

I have read that HASL finishes can wear down to the copper in fewer than 100 insertions. A friend of mine and fellow forum member told me about a vintage card he acquired which had HASL finish on the edge connector. The guy who sold him the card had apparently removed the card in a way that the solder was smeared between fingers, requiring cleaning with a soldering iron to avoid issues with the solder bridging pins.

An ENIG and a Hard Gold finish eliminate issues with HASL and only need be applied to the contact pins. The difference between ENIG and Hard Gold is the thickness, in microns, of the gold used to cover the contact. ENIG boards can often withstand up to 1,000 insertions, and I think ENIG is appropriate for hobbyist cards. HASL is not.
http://nerdlypleasures.blogspot.com/ - Nerdly Pleasures - My Retro Gaming, Computing & Tech Blog
User avatar
Great Hierophant
l33t
 
Posts: 2398
Joined: 2003-4-27 @ 08:20

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby Fagear » 2018-8-16 @ 22:17

Tiido wrote:HASL will destroy the gold plating in the slots over time though galvanic corrosion. There is a lot of scientific information on the matter.

Great Hierophant wrote:I have read that HASL finishes can wear down to the copper in fewer than 100 insertions.

Well... I guess those 20 boards have to be scrapped then. :depressed:
Or I can offer those (with discount) to someone who doesn't care about slots on the motherboard. :lol:
If I have to order new boards I have to know how many to order though...

P.S. Do we need to open another thread about DuoSID or stay in the SSI-2001 replica? :confused:
User avatar
Fagear
Member
 
Posts: 268
Joined: 2014-7-10 @ 22:27
Location: Russia, Moscow

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby Fagear » 2018-9-06 @ 18:34

An update.
While working on FMonster I've made several iterations of "building block" designes for parts of SSI-2001/DuoSID circuit.
And while I have to order new boards for production I thought: "what if I'll re-assemble DuoSID from those blocks now?".

Answer is...
DuoSID_0.9_top.png


The board will become significantly smaller!
DuoSID_2018-09-06_01.png


It is still a little bit larger than SSI-2001 replica, but not much! :cool:

Bonus: all jumpers are now at the top of the board for easy access and there is lowpass filter at the output to filter high-frequency harmonics above 20 kHz.
Minor downgrade: there are no more "mono-6-voices/stereo-3-voices" selectors and no line input. First, I think no one will use those features, second, those can be easily replicated by mixing in hardware or software.
User avatar
Fagear
Member
 
Posts: 268
Joined: 2014-7-10 @ 22:27
Location: Russia, Moscow

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby Kaizzer » 2018-9-09 @ 22:13

Beautiful layout!
Kaizzer
Newbie
 
Posts: 2
Joined: 2017-11-26 @ 19:38

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby scg » 2018-9-23 @ 08:03

Stojke wrote:Is there an available working MS-DOS SID player for this card yet?

You can get DOS player here: https://stdio.ru/download/ssi-2001/osp2001-0.2.zip
And sources: https://stdio.ru/download/ssi-2001/osp2001-0.2-src.zip
scg
Newbie
 
Posts: 1
Joined: 2018-9-23 @ 07:54

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby Fagear » 2018-10-03 @ 11:26

While I'm preparing new version of the DuoSID, I've made a demo video:
https://www.youtube.com/watch?v=xQChS2w_CDk
User avatar
Fagear
Member
 
Posts: 268
Joined: 2014-7-10 @ 22:27
Location: Russia, Moscow

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby matze79 » 2018-10-03 @ 13:33

Wow you do a nice job, making a PCB for every block etc. :)
https://dosreloaded.de - The German Retro DOS PC Community
https://www.retroianer.de
matze79
Oldbie
 
Posts: 1579
Joined: 2014-12-12 @ 14:25
Location: Germany, Frankonia

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby Thermalwrong » 2018-10-08 @ 13:01

Thanks for the video of them playing, the 'pseudo stereo' effect of the duoSID works really nicely and it sounds very clean :)
Thermalwrong
Member
 
Posts: 304
Joined: 2018-3-18 @ 19:24
Location: England

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby Fagear » 2019-2-03 @ 21:58

Here is a new version of DuoSID that is assembled from FMonster blocks.

2019-02-04 00.45.07.jpg
New version of DuoSID
User avatar
Fagear
Member
 
Posts: 268
Joined: 2014-7-10 @ 22:27
Location: Russia, Moscow

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby AlphaC » 2019-2-18 @ 15:25

Thumbs up!
User avatar
AlphaC
Newbie
 
Posts: 30
Joined: 2014-3-02 @ 07:30
Location: Norway

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby Fagear » 2019-2-18 @ 22:01

I've assembled one new board, but didn't test it yet.
IMG_8333.jpg
New DuoSID

Also I've assembled two SwinSID boards (PCB layout is mine based on Nano SwinSID schematic):
IMG_8345.jpg
DuoSID + SwinSID

Old version on the left, new one on the right. New PCB is smaller.
IMG_8350.jpg
DuoSID: old vs. new
IMG_8355.jpg
DuoSID mounting bracket

All available settings are now on the top edge of the board (as new version is made of FMonster blocks that all have that feature).
Also separate volume control is now available (to balance any differences between SIDs).
IMG_8357.jpg
DuoSID settings
User avatar
Fagear
Member
 
Posts: 268
Joined: 2014-7-10 @ 22:27
Location: Russia, Moscow

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby elianda » 2019-2-19 @ 00:22

Does it also fit the FPGA-SID ?
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
User avatar
elianda
l33t
 
Posts: 2423
Joined: 2006-4-21 @ 16:56
Location: Halle / Germany

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby AlphaC » 2019-2-19 @ 05:47

Nice soldering Fagear =)
User avatar
AlphaC
Newbie
 
Posts: 30
Joined: 2014-3-02 @ 07:30
Location: Norway

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby Fagear » 2019-2-19 @ 11:37

AlphaC wrote:Nice soldering Fagear =)

Thanks!

elianda wrote:Does it also fit the FPGA-SID ?

First time that I hear about this project...
As it looks now - it probably will not fit. Width of the board is not the problem, but a part with connector hangs over capacitors and jumpers on my board. Probably, if use longer pins for socket, it will fit fine. But it would be better, if FPGA-SID would have board rotated 180 degrees. There is plenty of space under the SID on my board.
User avatar
Fagear
Member
 
Posts: 268
Joined: 2014-7-10 @ 22:27
Location: Russia, Moscow

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby matze79 » 2019-2-20 @ 15:44

The SwinSID is now obsolete :)

Uses ARM CPU with DSP and FPU.

armsid.jpg


Also works on SSI very well, and Digital Sound works fine.

http://dzi.n.cz/8bit/armsid/index_en.php
https://dosreloaded.de - The German Retro DOS PC Community
https://www.retroianer.de
matze79
Oldbie
 
Posts: 1579
Joined: 2014-12-12 @ 14:25
Location: Germany, Frankonia

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby Fagear » 2019-4-24 @ 21:11

matze79 wrote:The SwinSID is now obsolete :)

Yes it is. I have to buy ARMSID and FPGASID to make a thorough comparison. One thing: FPGASID probably will not fit onto DuoSID because of oversized board and its shape.

In the meantime I've tested DuoSID and it works great!

IMG_8770.jpg
DuoSID with two SwinSID nanos

Some minor errors were found though:
- I was going to try light a LED when each SID is addressed... turned out that I've connected trace from LED not to /CS pin but to an 02 pin, so it just always glows... Oh, well. :-/
- I've swapped opamps order in the output stage. Nothing bad while "LEVEL" trimmers are set to max. But if you turn those down, not only level drops but high frequencies too (accidental low-pass filter).

To fix the last thing I had to swap opamps to correct order and to do it I had to cut some traces and solder some mod wires. Nothing difficult or ugly, but now it works properly.

IMG_8760.jpg
DuoSID: swapped opamps fix

Also I've tested SwinSID nano on board of my layout. Also not without errors: I've rotated power supply to crystal generator 180 degrees. :smug: So I've had to rotate it again and get output signal from the other side via mod wire.

IMG_8752.jpg
SwinSID generator fix

I've assembled two slightly different SwinSIDs: one on ATmega88 and other on ATmega88PA. Both work overclocked (32 MHz is an OC above "maximum" 20 MHz) but ATmega88PA version runs cooler. Also I have to note that SwinSID output is not truly analog - it is just PWM on 125 kHz. So filtering must happen on some stage: in the soundcard ADC (when recording) or speaker (when used in C64 with stock schematic).
Luckily on the DuoSID there is 3rd order lowpass filter that deals with it just great, outputting fine noiseless analog signal! :cool:

DS1Z_02.png
SwinSID: PWM output

Talking about noise... It is less than -70 dB! Pretty clean output even on very noisy motherboard.

DuoSID_noise.png
DuoSID: noise level


Also didn't noticed some errors and glitches of SwinSID that were shown in some YouTube videos, maybe I have one of the later firmwares with some bugfixes. But there are still moments when SwinSID gets music wrong compared to real 8580R5.

DuoSID with one SwinSID was featured in my video (SID+AY in Bad Blood): https://youtu.be/1fpP_-TrkMY
User avatar
Fagear
Member
 
Posts: 268
Joined: 2014-7-10 @ 22:27
Location: Russia, Moscow

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby elianda » 2019-4-25 @ 08:48

The card looks really versatile. Though I got a question how timing precise your card is. There are quite a few songs on the C64 that play samples or use oscillator feedback to produce reverb effects. For these the timing of the writes and reads is critical to produce the correct output.

From our experience the PC SID solutions usually fail in this aspect.

On C64 the CPU and SID run on the same clock, how close can you get with the ISA card?
How do the reads behave, if I issue a burst read on e.g. Osci3?
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
User avatar
elianda
l33t
 
Posts: 2423
Joined: 2006-4-21 @ 16:56
Location: Halle / Germany

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby matze79 » 2019-4-25 @ 09:43

Yeah i had some issues with swinsid in the past when setting the voltage divider Jumper on swinsid nano, the mcu was dying in some C64 models or stuttering. Others were fine.
5k6 Ohms parallel to 100nf Capacitor is not very good.

i think fpgasid is way too over engineered, the armsid is equal to real SID.
But i cant tune the Filters yet or upload new Firmware as there is no Dos Utility and i have no C64 on my Desk, they all stored away.

I can`t hear any difference. My Friend is a SID junkie and did not notice it at all.
So even a C64 Mastermind can hear the difference.
They will see it only if you open up the Case ;)
https://dosreloaded.de - The German Retro DOS PC Community
https://www.retroianer.de
matze79
Oldbie
 
Posts: 1579
Joined: 2014-12-12 @ 14:25
Location: Germany, Frankonia

Re: Innovation SSI 2001 re-engineering and/or conclusions

Postby matze79 » 2019-4-25 @ 09:48

elianda wrote:On C64 the CPU and SID run on the same clock, how close can you get with the ISA card?
How do the reads behave, if I issue a burst read on e.g. Osci3?


The SSI-2001 never was meant to resemble the C64 Soundsystem on PC.
If you want this you can build yourself a HardSID ISA.
Attachments
prototype.jpg
isa_sid_parts.txt
(3.67 KiB) Downloaded 4 times
isa_sid_top.png
isa_sid_placing.png
isa_sid_schematics.png
https://dosreloaded.de - The German Retro DOS PC Community
https://www.retroianer.de
matze79
Oldbie
 
Posts: 1579
Joined: 2014-12-12 @ 14:25
Location: Germany, Frankonia

PreviousNext

Return to Sound

Who is online

Users browsing this forum: Tiido and 4 guests