Jump to content

Quad SSM2164 VCA PCB


seppoman
 Share

Recommended Posts

Hi,

I've just finished the first proto of my new module - a quad VCA based on the SSM2164 IC  8). The 2164 incorporates four high quality VCA channels with logarithmic response and is one of the few highly integrated synth-related ICs which is still in full production (by Analog Devices).

The logarithmic response is kind of unusual for a synth VCA and not very suitable for modular synths (because most envelope generators already output exponential CVs - some reading: http://www.synthmuseum.com/magazine/linexpo.html). But for use as a (final) VCA e.g. of a Midibox SID, the log response is perfect - the CV generated by routing an envelope without tweaking the curve parameters to an AOUT channel is linear so the ssm2164 will give the same change in dB for the same relative value change over the full range.

While not as obviously exciting as a VCF, a VCA is a very useful addition - e.g. it serves as a noise gate for 6581 lovers like me, adds the possibility for true velocity sensitivity. Another great effect in conjunction with a VCF like my SSM2044 module: if you set the VCF to full oscillation with max keytracking, you can use it as a fourth (sine wave) oscillator. Sound demos to follow in the next days :)

The module is 50x80 mm in size, parts cost is about 10-15 Euros (for four channels! :D)

S

4906_SSM2164-1_jpgb23fb2194fb3271d687ee6

4906_SSM2164-1_jpgb23fb2194fb3271d687ee6

Link to comment
Share on other sites

Do both vca and vcf modules work in mbfm's signal path aswell?

The frequency modulation would instanly destroy all ICs  :P No sure, why not? :)

Om nom nom! The SSM cjip is pretty, nice white silkscreen - sure it's not a fake? ;)

yes, I got it directly from AD as a sample some years ago ;)

It's the same size as your SSM2044 pcb - do the mount holes line up as well?

the 2044 module is 90x60 mm, so no they don't line up - do you think it would be cool to make the PCB the same size? I didn't want to waste space and you'll need two VCF modules to feed all 4 channels anyway, so you'd probably place it between the two anyway.

could you post a shematic where the modules (vcf/ vca) wired to core and aout_ng ? for better understanding  ;) do we need a 2nd aout?

Well e.g. for a regular mb6582 configuration, you'd connect the audio out of one sid pair to the audio in of the VCF module, connect the audio out of the VCF module to one audio in pair of the VCA module and the output to your output jack. CV connections: first four CV outs of an NG go to the VCF module and two others go to two channels of the VCA module.

Second half of the VCA module would be free for another core/sid/aout/vcf pair.

I'm still researching if it makes sense to use the two other VCAs to control the SID feedback level. I've tried that out once but got mostly a little variation in level, not very exciting - either I made some mistake or the 6581 simply doesn't give much effect by using feedback - I'll have to try it with 6582s :)

I can smell an SSM2164 bulk order in the air...

hehe, does it already smell that strongly? ;) Well I could imagine something like that happening soon :D

S

Link to comment
Share on other sites

Niice!

If it works, it would be awesome to have a pair of VCAs post filter, and a seperate pair for feedback.

This would mean two add-on boards per core, and no need to move envelopes around the CAN bus.

I'm still researching if it makes sense to use the two other VCAs to control the SID feedback level. I've tried that out once but got mostly a little variation in level, not very exciting - either I made some mistake or the 6581 simply doesn't give much effect by using feedback - I'll have to try it with 6582s :)

Just curious, did you try it post - coupling cap? pre-filter board?

Link to comment
Share on other sites

Niice!

If it works, it would be awesome to have a pair of VCAs post filter, and a seperate pair for feedback.

This would mean two add-on boards per core, and no need to move envelopes around the CAN bus.

well with MBSID V2 you can't move around AOUT channels over the CAN bus anyway so you'll need to use one AOUT module per core :)

Just curious, did you try it post - coupling cap? pre-filter board?

I used the feedback pot connector of the mb6582 board, i.e. put the VCA between audio out and audio in of the SID. No additional parts. do you think a coupling cap would make sense here?

S

Link to comment
Share on other sites

OK I've checked out using the VCA for controlling the SID feedback a bit more in depth - originally I hadn't thought of activating the external input flag ;) so it IS possible to do this (doesn't hurt to adjust the VCAs for a little more gain), but I still find the effect not really exciting. Feedback seems to be an easy way to give the SID's internal filter some more resonance as long as you don't have external VCFs, but the change isn't drastic enough to justify doubling the number of VCA modules and using up the last two remaining CV channels on my AOUT_NGs. If I want mean resonance, I'll continue using my ssm2044 vcfs for that :D

S

Link to comment
Share on other sites

OK I've checked out using the VCA for controlling the SID feedback a bit more in depth - originally I hadn't thought of activating the external input flag ;) so it IS possible to do this (doesn't hurt to adjust the VCAs for a little more gain), but I still find the effect not really exciting. Feedback seems to be an easy way to give the SID's internal filter some more resonance as long as you don't have external VCFs, but the change isn't drastic enough to justify doubling the number of VCA modules and using up the last two remaining CV channels on my AOUT_NGs. If I want mean resonance, I'll continue using my ssm2044 vcfs for that :D

I absolutely don't see the need to add extra VCA modules, but since the chip is a quad anyway, and we can't move envelopes around to use the VCAs for other SID pairs, it seems right to use all 4 VCAs on on SID pair.

The bigmech page about the feedback loop

AFAIK the feedback does certainly help with the originally whimpy resonance of the SID filter, but if you bypass all oscillators, and the EXT input, (which you probably would when using the much tastier SSM filter) then you could get some more gritty, noisy almost FM-ish sounds. I imagine the confuration as basically SID as oscillator bank, with feedback, then to SSM Filter, then SSM VCA. Kinda like this:

[tt]

{ext in}-+---{SID}----+-{SSM Filter}-+-{SSM VCA}---{Output}

          \-{SSM VCA}-/

        (feedback loop)

[/tt]

I used the feedback pot connector of the mb6582 board, i.e. put the VCA between audio out and audio in of the SID. No additional parts. do you think a coupling cap would make sense here?

Don't know if you need to ADD one, just saying that since the SID is DC coupled (single power rail +9v/GND), and the VCA looks to be AC coupled (+V/0/-V) this might reduce the effect if not properly coupled, but if you used the 3-pin header (J3/23_SIDx on Wilba's 6582 board) it should be fine. Just make sure all OSCs, and EXT IN are set to bypass the SID filter.

Link to comment
Share on other sites

I absolutely don't see the need to add extra VCA modules, but since the chip is a quad anyway, and we can't move envelopes around to use the VCAs for other SID pairs, it seems right to use all 4 VCAs on on SID pair.

well you don't need to move envelopes around on the CAN bus to use the other two VCAs for another SID pair - just go old-skool and simly use wires ;) nobody stops us from connecting the other two audio and CV connections to another SID/AOUT :) so in fact doing VCA feedback doubles the number of modules needed for a full mb6582 from two to four.

AFAIK the feedback does certainly help with the originally whimpy resonance of the SID filter, but if you bypass all oscillators, and the EXT input, (which you probably would when using the much tastier SSM filter) then you could get some more gritty, noisy almost FM-ish sounds. I imagine the confuration as basically SID as oscillator bank, with feedback, then to SSM Filter, then SSM VCA. Kinda like this:

[tt]

{ext in}-+---{SID}----+-{SSM Filter}-+-{SSM VCA}---{Output}

          \-{SSM VCA}-/

         (feedback loop)

[/tt]

yes that diagram is what I'm doing. What do you mean with "bypass all oscillators"? I can only switch them on or off, and with off the only thing left would be a sine as long as the ssm filter is set to oscillate.

Don't know if you need to ADD one, just saying that since the SID is DC coupled (single power rail +9v/GND), and the VCA looks to be AC coupled (+V/0/-V) this might reduce the effect if not properly coupled, but if you used the 3-pin header (J3/23_SIDx on Wilba's 6582 board) it should be fine. Just make sure all OSCs, and EXT IN are set to bypass the SID filter.

uhm why would I want to not use the SID filter at all and still need the feedback loop? I thought the whole intent of doing feedback at all was to increase the VCF's resonance, no?

about DC/AC coupling - yes I'm using the J3/23 connectors for feedback. I already noticed there might be an issue with this... the SID probably works on a virtual GND scheme and the Sid module (and the mb6582 pcb as well) use an electrolytic cap on the output to AC couple this. So the output is probably going about 0V, but the mb6582's audio input AGAIN does AC coupling which is probably not necessary in the feedback situation. with or without this second cap, the ext in of the SID is getting an AC coupled signal which has a huge negative DC offset compared to what the sid's output is giving. So I guess to get good effect of the feedback it would make sense to not use the intended connector but grab the output signal from before that cap and feed it back after the input cap? Guess I'll try out if that helps to improve performance :)

S

Link to comment
Share on other sites

well you don't need to move envelopes around on the CAN bus to use the other two VCAs for another SID pair - just go old-skool and simly use wires ;) nobody stops us from connecting the other two audio and CV connections to another SID/AOUT :) so in fact doing VCA feedback doubles the number of modules needed for a full mb6582 from two to four.

Ok, You got me. Don't know how, but I just hadn't seen it that way... This is definitely the most obvious way to do this.  :-[

yes that diagram is what I'm doing. What do you mean with "bypass all oscillators"? I can only switch them on or off, and with off the only thing left would be a sine as long as the ssm filter is set to oscillate.

Off hand, I can't recall how the MBSID handles it (TK?), but the SID chip will supposedly do it. Register 0x17. 4 bits for filter resonance, and 4 bits for each of EXT, OSC3, OSC2, OSC1 routing to filter (set high) or direct to output (set low).

uhm why would I want to not use the SID filter at all and still need the feedback loop? I thought the whole intent of doing feedback at all was to increase the VCF's resonance, no?

It is just a suggestion, you don't have to do it. I did get the impression from the bigmesh article that this was also a way to make complex, gritty sounding waveforms, with or without the use of the SID filter. Think of the feedback loop in an FM algorithm synthesiser. I thought that this sounded at least possible. If it serves no benefit, then by all means leave it. I don't want to tie you up with minutae, when you are still working on the VCA board...

Link to comment
Share on other sites

  • 10 months later...

I guess it is time to wake this thread up. I never saw the pros to use the SSM2164 in a MB-SID configuration. There are allready envelopes and stuff. Please explain, what makes this module usful?

please allow me to quote myself :tongue:

While not as obviously exciting as a VCF, a VCA is a very useful addition - e.g. it serves as a noise gate for 6581 lovers like me, adds the possibility for true velocity sensitivity. Another great effect in conjunction with a VCF like my SSM2044 module: if you set the VCF to full oscillation with max keytracking, you can use it as a fourth (sine wave) oscillator. Sound demos to follow in the next days :)

It seems that last promise was a lie :whistle:, but I'll do some audio demos next week.

Personally I think that VCAs are quite useful for a MBSID. But of course, as a good VCA does only change volume but not alter the sound, the benefit is less spectacular and obvious than when adding nice VCFs. So if you can't imagine what to do with VCAs and maybe you're not too much into synth/sound programming anyway, you probably wouldn't need these extra VCAs.

But if you already have good VCFs and you think it's fun to expand your MBSID even more, and you like to program/alter sounds and explore the possibilities, adding this module is definitely a good thing :thumbsup:.

S

Link to comment
Share on other sites

After a quick search I have not found a schematic for this board, correct me if I'm wrong :)

yes that's right, so far there's no documentation for this board. It was a spontaneous decision to offer pcbs for this module, too, when I set up that multi-bulk order, so I didn't prepare any docs/parts list etc yet. I'll change that in the next 1-2 weeks.

Would this PCB also be suitable for putting into a modular system?

If yes, then I presume it is ok running on +/-15V?

again, quoting myself :)

The logarithmic response is kind of unusual for a synth VCA and not very suitable for modular synths (because most envelope generators already output exponential CVs - some reading: http://www.synthmuse...ne/linexpo.html).

Just like the SSM2044 module, this one wasn't designed with a modular system in mind. It's meant to be used together with a (linear) DAC/AOUT, doesn't have multiple inputs/outputs per channel, no connectors for pots (e.g. for adjustable overall volume or depth control). Instead, it's designed to be a small, cheap, easy to build addition to Midibox/AOUT projects like MBSID or MBFM (maybe you could also use these pcbs to build a simple "vca automation" box in conjunction with a MB64 core, as this thing has a really good s/n ratio etc). I'm not saying it's not possible to use it in a modular synth, but there are other SSM2164 based VCA modules out there that are designed for that usage and save you the hassle of adding external stuff to expand it to full versatility. - That being said, 15 V should be fine :ahappy:

S

Edited by seppoman
Link to comment
Share on other sites

This VCA is a good workaround to the SID ADSR bug... you just set the oscillators to A=0 D=0 S=15 R=0 and control volume through the SID engine's envelope.

Well actually you are not limited to just Env->Vol modulation... you can do whatever you like in the mod matrix and modulate volume. There's even a parameter to switch whether volume modulation will modulate the SID's internal 4-bit volume control or an external analog VCA.

V2A (Volume to Analog): Volumes are forwarded to external CV outputs as well (assignments have to be done in the setup_*.asm file). Note that the effective volume value behind the modulation path is taken - accordingly you are able to realize an external VCA to overcome the ADSR bug.
Link to comment
Share on other sites

  • 3 weeks later...

Hi Seppoman,

it is the second time that I take part for a buy and I am not so experienced.

Where do I have to compile the form for SSM2164 PCB bulk order?

I have not a Paypal account but often I buy through Paypal with my card, is it the same?

Thank You for any answer

Regards

Link to comment
Share on other sites

  • 2 weeks later...

so finally I managed to do a little demo using both the ssm2044 filter and the 2164 vca (with mbsid and 6581 sids).

0.00 - a sound with very high resonance, cutoff is modulated by key velocity. The filter is at the beginning of self oscillation.

0.15 - same sequence but using the vca to mute the oscillation during pauses.

0.28 - another high resonance sound with cutoff frequency tracking the played pitch, i.e. using the filter as a sine oscillator.

0.35 - same sequence using the vca.

0.46 - again a high resonance sound with cutoff as a high octave.

0.54 - same sequence using the vca, playing around with gate length and vca offset ("influence").

I guess everyone can imagine how key velocity controlling note volume sounds so I didn't record an example.

Please disregard the zipper noise present in the second and third example - I did some experiments with capacitor values last week and only remembered that after having recorded the demo and listening to the mp3 with headphones... Problem here is that the AOUT update rate on MBSID is quite low (every 2 ms), so very fast envelopes make the "ramp" of a fast attack consist of only one or a few value changes, i.e. there's huge jumps in the CV signal and the 2164 reacts with some noise. This is compensated by low pass filtering the CV a bit, in this example it's obviously not enough filtering. There's only a fine line between having these noises and filtering too much, i.e. killing the snappiness/punch of the VCA, but I'll specify a good compromise value in the final parts list.

S

2164demo.mp3

Link to comment
Share on other sites

  • 2 months later...

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
 Share

×
×
  • Create New...