Old Part Day: Voltage Controlled Filters

For thirty years, the classic synths of the late 70s and early 80s could not be reproduced. Part of the reason for this is market forces — the synth heads of the 80s didn’t want last year’s gear. The other part for the impossibility to build new versions of these synths was the lack of parts. Synths such as the Prophet 5, Fairlight CMI, and Korg Mono/Poly relied on voltage controlled filter ICs — the SSM2044 — that you can’t buy new anymore. If you can source a used one, be prepared to pay $30. New old stock costs about $100.

Now, these chips are being remade. A new hardware revision for this voltage controlled filter has been taped out by the original hardware designer, and these chips are being produced in huge quantities. Instead of $100 for a new old stock chip, this chip will cost about $1.60 in 1000 unit quantities.

The list of synths and music boxes sporting an SSM2044 reads like a Who’s Who of classic electronic music machines. E-Mu Drumulators, Korg polyphonic synths, Crumars, and even a Doepfer module use this chip in the filter section. The new chip — the SSI2144 — supposedly provides the same classic tone but adds a few improvements such as improved pin layouts, an SSOP package, and more consistent operation from device to device.

This news follows the somewhat recent trend of chip fabs digging into classic analog designs of the 70s, realizing the chips are being sold for big bucks on eBay, and releasing it makes sense to spin up a new production line. Last year, the Curtis CEM3340 voltage controlled oscillator was rereleased, giving the Oberheim OB, Roland SH and Jupiter, and the Memory Moog a new lease on life. These chips aren’t only meant to repair broken, vintage equipment; there are a few builders out there who are making new devices with these rereleased classic synths.


57 thoughts on “Old Part Day: Voltage Controlled Filters

      1. The “is” before ‘market place’ should be “was” ?

        ..I don’t email on this device – perhaps all posts should contain a link so we can easily contact someone with suggestions.

      2. You need to chill. I know you deal with this a lot…

        But it’s hardly tactful to flame someone for correcting your fuckup, even if it’s minor and nobody else really cared. That said IDK if this particular commenter has been problematic previously. BTW, to *find* your email one has to go to About, then your twitter, then your github; or possibly another path I didn’t find. Perhaps if you want people to email about suggestions/corrections, an address for such things should be displayed; doesn’t seem right to congest the tip line with that either.

        Great article though, contains information we care about in a good format. Great to see someone making old synth chips again. SID when? :P

      3. Brian, I can understand your frustration as an author, but frankly, no indication of an email address that you could send info to on the website. Nor on the “about” site. For me to send you a message via something unintuitive like twitter, I’d need you to follow you. So really, this is the only direct way of reaching out to the author. And: it’s the *comments* section. Which is, as far as I can tell, for *comments*. Like, “Hey, I think you misused a word”.

        Also, for christ’s sake, this is the internet. People aren’t always polite about criticism. It’s usually the very worst reaction to act offended. Also, don’t know if you’ve got history with Bru, but come on, “Do you mean realising it?” isn’t even proper criticism, much less “shitting up the comments.

        Don’t troll yourself. You’re supposed to be an internet professional, I think. Don’t think that includes reacting sour to short corrections.

        1. I don’t think we’ve got history; I’ve been reading hackaday for years but only very recently started commenting.

          To an extent I understand his frustrations. If he’s excited about this and wants to talk about it with someone, having the first response be “Hey, you fucked it up” can be very disheartening.

      4. I clicked on Brians name then the About on his google and there was his addy. Two clicks, no sign ups, no nuthin’. Not that hard for supposedly techy people to figure out I’d have thought.

        1. Clicking on his name brings up a list of his articles. Clicking on “about” at the top of that page brings up an “about hackaday page”
          I’m not sure what you mean by “About on his google”
          I’m glad you found his address, but I would have needed to leave a comment if I had wanted to suggest a correction.

        2. Do you mean you clicked on his name in the comments? His comment wasn’t there until AFTER the typo correction was posted. If you click on his author name at the top of the article, you’re not given his email address at all.

      5. If you wish for someone to contact you for corrections in a specific manner, just put a link a the bottom of the article for submitting corrections. It could be a form, a link to a “wiki”-style editor or just an email address link but without that indicator, people are going to “shitting up the comments”. It also counts as a human hack because it’s modifying their behavior to do what you want based on an understanding of their psychology (inner workings).

          1. ^^^^^^^ THIS!!!

            It’s fucking 2017 already!!! We have the technology… Maybe the comments software IDE uses the same programming that the comments section does…. forever enshrined in digital stone.

      6. I would like a function that lets me select to hide ALL spelling or grammar related comments in Web-Forums. In 99% it’s clear what the author want’s to say and the discussion of language related things just distracts from the technical content.
        From that point of view a possibility to silently inform the author about this – without bothering the other readers – would be nice. Just to keep meaningless clutter out of the comments.

      7. Actually you shat up the comments, Benchoff. You have really got to stop lashing out this way. It is becoming more of a contention point in the comments and is hardly professional. I can only guess you are under some tremendous stress, like your dad having a stroke or something. Just correct/delete the comment and move on bro.

      8. Firstly, responses like that will drive people away from HAD. It shows a lack of professionalism.

        Secondly, people shouldn’t have to jump through hoops or traverse a maze of links to find your address to “tactfully email” you (or any of the other authors).

    1. Technical content is actually far more important than an English lesson. I appreciate the new info this article shares. I need to make one technical correction however. No Sequential Circuits Prophet 5 synthesizer ever used a SSM2044 chip. The REV 1 & 2 Prophet 5s used the earlier SSM2040 filter chip and the REV 3 used only CEM chips.

    1. Oh don’t, I remember the classifieds in the mid/late 80s in Music Technology etc with Seq Pro-ones and sh101s going for £50 or so… Moogs a little more but you could still find rogues/prodigies for well under £100. Sure some big chain were clearing out 303s for around £30 new… :o

    2. Yup bought an SH101 for fifty bucks in ’98 :) Sold it for rent in 03 at a slight profit but man, I really regret it at this point. I recently bought an A01k which thankfully included s/h and has CV/gate for all of my diy gear :) Cool to hear they will make some new chips to help keep the old gals making noise :)

  1. This is great but I hope they can diversify. The reason old chips sell for a higher price is that demand exceeds supply. I can imagine that the market for new (re-make) chips will saturate very quickly.

    The schematic of the chips is weird! There’s an op-amp in there with triceps or something. It’s like a freak show for circuit symbols.

    1. People are working on an FPGA-Implementation of SID. Considering how well the Atmel-based SwinSID (ab overclocked ATmega) was already working, this could solve the problem.

      1. An FPGA implementation would have a digital filter and would not fully emulate the characteristic distortion of the shitty MOS transistors; it might still be musically nice but that’s not a SID. It’s not about digital versus analog per se, it’s that the characteristics of a real physical SID filter are not easy to emulate, whether or not it ‘sounds better’.

        AFAIK sidplayfp is the closest software emulation yet; you could probably run a ported synthesizer portion on a modern ARM microcontroller (i.e. Cortex-M4F) without too much difficulty.

        1. For clarification, ‘without too much difficulty’ is relative. Realtime DSP that was designed for a PC is probably going to require at least minor optimisation.

  2. @[Benchoff]

    As an avid reader here who very much enjoys your articles, I often see an article that has no comments and I would like to make a comment but I also know that the leading comments can often dictate the tone of following comments so I leave it be for a while so others can comment first.

    Obviously not all people are going to do this so perhaps there should be an option for HAD or editors to re-arrange the order of early posts.

    Obviously that is not your responsibility so this post is more toward the HAD team rather than directly to you.

    This facility could even be available to uses if you think they could handle such a responsibility. And going by the fact that these pages are more or less un-moderated, I think that the users here would be able to do so in a responsible manner.

    PS: I think this article qualifies to share the tag Analog.

    1. Actually… that has me thinking about something similar reddit did to control toxic behavior

      Scores on reddit threads aren’t updated in real time; there is a bit of lag between an up/downvote action and the actual displayed scored to prevent people from gaming the point system

      it’s not like these comments are time critical…. maybe a 1 hour embargo on unregistered comments where people can write their FIRSTS but they won’t actually be posted to the article until the timeout, then they’re all posted simultaneously?

  3. Before DSP took over the role of controllable audio filters, there was the Switched Capacitor Active Filter (SCAF) which was controlled by a clock signal.

    Of course, the clock was often controlled by a voltage controlled oscillator.

    For me, the go to part was the MF-10 which is still available from several manufacturers.

    This part was the basis of the MagicNotch automatic audio notch filter where the frequency was controlled by a voltage output from a phase locked loop to remove constant frequency interference.

    I suppose arguing over whether to use DSP or one of these old chips is like arguing about vinyl vs CD or tube amps vs solid state.

  4. It’s interesting that Behringer is remaking the original Curtis voice chips as well. Getting quite a bit of flack over it on MuffWiggler and such, but I personally am excited to have other options. The Deep Mind 12 sounds fantastic and this is coming from an owner of a few classic polys (Juno 106, AX80, JX3P).

    I can’t wait to check out their Model D clone. Crazy excited that it’s in Eurorack spec and for $400 MSRP, you’d have to be crazy or an elitist to not pick one up.

  5. Can people cool it with the annoying comments? Even if he had a typo – who cares?! And if you want to suggest fixing it do it in a civilized manner. If getting his email is too complicated and you get mad because you have to look for too long – you can stop and not correct him. Can’t we simply enjoy the content? Is that too difficult to do?

  6. I thought this was going to be about voltage controlled filters in general, not a specific IC.

    People made synthesizers before there were VCF ICs, or other electronic music ICs. The Moog filter is famous, didn’t use ICs and the patent s out there. PAIA got by with some primitive circuits. Lots of filters could be made voltage variable once you had some element to vary. The CA3080 OTA made it easier.

    Don Lancaster wrote quite a bit about electronic music in Popular Electronics, at one point I expected him to have an “Electronic Music Cookbook”. There was Electronotes, a newsletter with lots of theory and circuits, they put together a photocopied Handbook, you supply the binder.

    By the time the fancy dedicated ICs like this VCF arrived, things had shifted off to niche publications. So we read about the special electronic music ICs in Popular Electronics or maybe Byte, but most of the circuits were elsewhere.

    Of course, Byte had some articles about using computers for electronic music, including one that was sort of a hybrid, using a bit of hardware (regular TTL) to do some of the work, since 8bit CPUs running at 1MHz or so were a bit too limited.


  7. I have mixed feelings about re-released components and equipment. There seem to be 2 major markets for them: one of them being the “the design works perfectly and I don’t want to make the effort/expense to change it to accommodate a newer design, even if it is better” (if it ain’t broke, don’t fix it!) market and the other being the “nothing could possibly be better than the original design” (stuck in the past) market. There are merits to each side, for sure. But, is it worth the expense to relive your (or a long-deceased person’s) glory days?

    Now, excuse me while I look for some more 8″ floppies…

    1. And 3v3 single clock instruction cycle Z80 and 6502 that run at higher speeds.

      It’s getting to the point now that you use the original 5Volt CPU (cos no choice) some level translators and everything else is 3v3.

  8. Well darn, I was hoping my handful of Curtis ICs would fund my retirement. Or a boat. Just a small one.

    Oh well, I guess I’ll just have to build something with them.

    PS – Hackaday needs proofreaders. If you’re not interested in doing that, at least stop whining when commenters point it out. At least make it easier than chasing several links to find the back-channel to report typos.

    1. Vintage chip $100.
      A remake of the chip stirs up a buzz but forum gurus claim that even though it sounds really good it doesn’t nail that organic warmth of the original.
      Vintage chip is now $220

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.