I’ve finished breadboarding the SSM2044+SSM2164 board for the Shruthi-1. I need a couple of tweaks for the filter’s frequency response (to get accurate cutoff tracking) and that’s it - I hope to send the board for manufacturing in the middle of next week. The funny thing is that I use only half of the SSM2164, so maybe there’s room for another processing to do with the remaining log-VCAs on this chip. I have to think about it…
This board is going to be a “limited edition” thing - I won’t hunt for more SSM2044s, and just let go the tiny stock I have. It’s not as hard to find as the CEM3379, though. The sound is very round and beefy ; but my OTA-based filter doesn’t sound that lame compared to it…
what’s your preference between the three olivier?
If I can get rid of the annoying feedthrough/plop problem, the SSM2044 board is going to be my favorite, because of one extra feature I’ve just added: a 1-pole HPF! (There was an unused section of the SSM2164).
My 4-pole design doesn’t sound as good as the SSM2044 (though for some input signals and ranges of cutoff they have a lot in common), but it’s cheap, the parts are easy to source, and because there’s no “blackbox” chip in it, it serves well my vision of the Shruthi-1 as a gateway/educational project for people who want to get into synth DIY.
The CEM3379 is in a different league than these two since it’s much more aggressive and bold. It really depends on the kind of sounds you want to make and where you want to put them in a mix. For example, I suspect that in a polychain setting, the clearer/rounder sound of the two other filters will give better results.
I have been really impressed by the demos of the OTA filter, lame is not the word that comes to mind.
But then I’ve always wanted a SSM2044 low pass.
The extra high pass is tasty icing on the cake!
I guess the only answer will be to get both
Does anyone have tips on their recommended chip suppliers for a SSM2044?
Australia is a long way from most large synth graveyards, so ebay looks most promising so far.
There are a lot of them on ebay, all from sellers that do not look very reputable or professional.
I got mine from:
This really looked like a scam, but this wasn’t. I’ve receive the chips and they do work.
Perhaps you could buy a large stock to reduce price and sold them with the kit ??
I’m interested by SSM and CEM3379 cards !
Someone in France sell 2 SSM2044 for 15 € each !
At the moment, I don’t want to invest anything more before shipping the first kits. Also, given how shady some of those suppliers can be (cf thread on E-M about the alleged link between one of those and the HK mafia and the kind of dirty wars these suppliers play against each other) I’d rather ramp up slowly in the quantity I buy before sending those guys several hundreds of dollars.
I ended up rolling the dice to China here:
$15 US for 2 SSM2044’s including free international shipping. I decided that kind of price would be hard to get even on a bulk order. Too good to be true? I’ll report back.
Inexplicably they will post anywhere except Italy and Spain…?
Warning, Off Topic:
I have had remarkable luck with things that appear to be scams. Last year I googled my way into a forum and ended up direct transferring many hundreds of unprotected dollars to Bulgaria. Only to have some excellent condition and well-packed Lynx converters arrive promptly at my door for a great price and from a very friendly person.
Risk and reward I guess.
But I digress.
Oliver, please consider this a vote of confidence for your analog troubleshooting skills
Update: I cannot get the filter to work the way I want. Either I do not AC-couple the output (just like any SSM2044-based schematics I’ve seen) and I get a monster DC that makes the output buffer clip, either I AC-couple it and get quite a lot of noise (which isn’t there when the output is not AC-coupled). This seems fishy because none of the schematics I’ve found online seem to work with the chip I have. I’ve bought a chip from a more reputable supplier to double-check that my problems are not due to a bad chip.
Sounds frustrating - thankyou for your tireless efforts.
Mine may well be dodgy too if they ever come, so I surely won’t force you to slog yourself over it
I wonder if people really do manufacture fake chips?
It sounds like a lot of cost and effort, why not make replicas of discontinued chips (to the glad acclaim of synth lovers everywhere)…?
Perhaps it would be more a case of rebadging any old IC…but then you would think it wouldn’t even nearly work.
Either way, electrons sound like they are even more pedantic than lines of code. Your open process is helping me to see the sweat and setbacks that go into making synths. It’s increasing my respect for synth designers many times over.
Fake chips: yes there are. Anything in DIP package can be repainted/relabeled to look like a sought-after chip! Sparkfun even posted about a fake reel of Atmel micros they bought from a chinese supplier.
Faulty chips that did not pass QC? Maybe.
Producing new vintage chips: masks for the old chips are not compatible with modern manufacturing processes. Anyway, setup costs are so high and demand is so low that it wouldn’t be worth it. OnChip still makes CEM-like chips (PA381, PA382, PA397). Minimum order quantities are in the order of several thousands. At least, it’s still easy to get basic building blocks for CV circuits (SSM2164 and LM13700), but maybe one day these far more generic ones will go away.
The 2044’s arrived!
Well, I should say the IC’s arrived, we can be certain of little more at this stage
They don’t look very old. I know that doesn’t mean much, but I won’t get my hopes up early.
Does anyone know a way to find out if these are the real deal or not?
Else I can just be patient. That would probablly be good for me
If the date code is year + week, 8547 doesn’t look silly (of course we can print anything we want on those chips).
I have bought 2 from a trusted source (labelled 8330 and pristine), and 10 more from shady china/HK ebayer (5 labelled 8730 and 5 labelled 0520, all looking grossly pulled from boards). The labelling on the 0520 was in a different font that looked cheap, so I expected them to be crappy logic gates badly relabelled. But they are not, they all filter, they all resonate, they are all going to give me nightmares about DC-offsets at the output.
That’s great and surprising news that they all seem to behave the same. For the cheap chips that is.
Not great for your circuit troubleshooting…
If nothing ends up working you can always rename the DC Offset the ‘integrated analog transient feature’?
I just noticed this in the ‘pipeline’ too:
SSM2164 dual SVF + SSM2164 VCA
We’re in an expanding candy store of filters!
I assume for SVF wer’re talking ‘state variable’ - wow.
Do you expect to be able to have much soft control over each one?
Even I can see there are only so many pins…
However will I choose?
Perhaps I will need to stack all filters in a giant sandwich and use an output to switch between them.
For the SVF, I haven’t decided what will be digital controlled and what will be pads for pots/switches - but keep in mind that there’s only 2 extra CVs besides the 3 usual suspects (cutoff, resonance, and VCA gain).
I only just noticed the SSM2044 with HPF demo on soundcloud!
Sounds VERY good, and I distinctly hear no nasty transients…
Plus the 1 pole HPF has changed to a 2 pole?
Don’t tell me amidst the kit production you’ve managed to make SSM2044 progress as well!
I think I can hear my paypal account stirring…