You should take a look at the Korg Prologue filter. It's a two-pole filter, at first it looks like a state variable, but when you look closely, you see that the feedback paths are arranged differently.
This was fantastic! Thank you! I’m in the process of designing a VCF around a CEM3320 that I have laying around. I liked what Elka did but don’t see the need, yet, for all the configurations (I.e. 4 pole low pass). This video definitely gives me ideas of how I’d prefer to structure the design now. Thanks!
Im not so sure the 2040 had 'reliability issues', more it was a case of 'royality issues' namely having to pay Dave Rossum and eMu a small royalty for each use. When Smith reneged on the deal and switched to Curtis's cheaper version, eMu very nearly went bankrupt but.. it forced them to design the eMu Emulator sampler to stay afloat. The rest is history. Presumably Rossum and Smith resolved their differences eventually. Edit: The Oberheim Xpander used the 4 OTA filter cells in various combinations to achieve 15 different types including all-pass filters used for phasing effects.
@@Lantertronics I believe Rossum had penned the design of the 2040 and had a deal with Sequential, the royalties were mostly put towards developing the ambitious polyphonic eMu Audity (not the later rackmount namesake). This prototype still exists in non-working form in Canada, I gather. There was a gentleman's agreement, but the royalties were obviously a drag on Sequential's earning potential hence the 'betrayal'. Dave Rossum has related the story on more than one occasion and there are some interview videos on UA-cam. Anyway the upshot was the first reasonably affordable sampler (thanks to clever DMA memory scanning techniques I believe) and the Audity was a bit of an uncommercial money pit anyway. Dave Rossum is a self taught, unsung hero of music tech, I believe he not only designed digital filter chips for the highly advanced eMu samplers but he also helped design the digital filters of the Gem S Series workstations as a consultant, but I could be wrong. Anyway not bad for someone who started out as a student of biology and stumbled into modular synthesis by accident.
GTECH was my first choice, but didn’t make it so had to settle for Purdue Electrical engineering…how I wish I had made GT now!!!! This is the stuff I wanted to do EE for!
Nice video! I wonder about the Polyfusion 4-pole lowpass filter's feedback path, as you mentioned it's only dividing down by 6 before going into the OTA, so the OTA should be plenty nonlinear in that operating region? I guess that's probably intentional to put some saturation in the feedback loop and tame the resonance, so that it doesn't go too wild, a bit like the clamping diodes in the Korg MS-20's filter. Would that be the same kind of thing in the Prophet 5 rev 1, where it multiplies by 10 before going into the SSM2020? That seems really extreme though, but maybe the SSM2020 has a wider input range before saturation than a standard OTA?
Another awesome lecture, thank you for putting these together and making them available! I learn so much listening to them. At the end of the video you mentioned using switches to change between different filter configurations. The electric druid website has a link to the "multiple identity filter" which was part of a series of articles in 1979. It uses the CEM3320 and about a dozen switches to do exactly that. I took a crack at building a PCB for it a while back, but made a bunch of mistakes in the design and haven't gone back to it... maybe its time. Anyway, here is the article: www.synthdiy.com/files/2002/MIF.pdf
Awesome lecture! I was wondering if the SSM2044 also could be categorized as an OTA-based 4-pole approach? Sadly I'm a little overwhelmed by the patent (US #4404529). Also the famous Korg NJM2069 Filter IC seems to be a big mystery. The only info I can find is a 'sort of' block diagram in the service manual of the Korg Poly 800...
Thanks for this vid! I thoroughly enjoyed it. I have a couple questions which I hope you could answer. 1. Why does the OBX ota filter stages not have local feedback? 2. Is the type of cap used really that critical? Im curious because when I was prototyping my vcf I used ceramic disc caps because thats what I had on hand. I have read that this is a big no no because of microphonics. To my surprise the cheap little disc caps sound great and I dont notice any undesired behaviour. I may use them in my design.
if the OTA is closed, the knob turned left, isn't the audio just going through the "feedback" loop, opening up the OTA's will subtract their portion of the audio?
@@Lantertronics Hey professor, well, a "general" OTA / VCA ic based low pass filter "feedback" circuit. to me it seams the audio is bypassing the filter through "feedback" when the OTA filter cells / stages are CLOSED, the frequency knob turned left. nothing is fed into the capacitors. i think opening the filter cells / stage (turning the knob to the right) causes subtraction from the original audio going along the "feedback" path. allowing the capacitors to charge. thereby chewing off a portion of the audio. Afrotechmods - Fun with ultracapacitors!! ua-cam.com/video/EoWMF3VkI6U/v-deo.html Afrotechmods, has a good explination.
@@tired.mp3 well, the "general" low pass filter pretty much everyone builds ( some 200+ producers are found on modulargrid alone ), like Ray Wilson, Mutable Instruments, Rene Schmidt, ...
@@tired.mp3 well, it seams the audio should go through the feedback, if you close the filter cell / stage nothing passes through yet the filter sounds. i'm not sure how much the 100K / 68K ( as Roland used in the Jupiter 8) / 20K ( as seen in professor Lanterman example) resistors found in the feedback loop affects the gain of the filter. since there aren't really voltage dividers.
This is the best example of circuit theory I have ever seen, thank you.
Thank you for explaining the 4-pole LP filter theoretically for both OTAs and transistors.
You're an amazing teacher Aaron, thanks for existing
As always: Thanks for sharing your lectures and insights!
You should take a look at the Korg Prologue filter. It's a two-pole filter, at first it looks like a state variable, but when you look closely, you see that the feedback paths are arranged differently.
This was fantastic! Thank you! I’m in the process of designing a VCF around a CEM3320 that I have laying around. I liked what Elka did but don’t see the need, yet, for all the configurations (I.e. 4 pole low pass). This video definitely gives me ideas of how I’d prefer to structure the design now. Thanks!
Im not so sure the 2040 had 'reliability issues', more it was a case of 'royality issues' namely having to pay Dave Rossum and eMu a small royalty for each use. When Smith reneged on the deal and switched to Curtis's cheaper version, eMu very nearly went bankrupt but.. it forced them to design the eMu Emulator sampler to stay afloat. The rest is history. Presumably Rossum and Smith resolved their differences eventually.
Edit: The Oberheim Xpander used the 4 OTA filter cells in various combinations to achieve 15 different types including all-pass filters used for phasing effects.
Interesting... I thought the fallout was over a patent related to keyboard scanning. I didn't realized the SSM chip designs were part of that as well.
@@Lantertronics I believe Rossum had penned the design of the 2040 and had a deal with Sequential, the royalties were mostly put towards developing the ambitious polyphonic eMu Audity (not the later rackmount namesake). This prototype still exists in non-working form in Canada, I gather. There was a gentleman's agreement, but the royalties were obviously a drag on Sequential's earning potential hence the 'betrayal'. Dave Rossum has related the story on more than one occasion and there are some interview videos on UA-cam. Anyway the upshot was the first reasonably affordable sampler (thanks to clever DMA memory scanning techniques I believe) and the Audity was a bit of an uncommercial money pit anyway.
Dave Rossum is a self taught, unsung hero of music tech, I believe he not only designed digital filter chips for the highly advanced eMu samplers but he also helped design the digital filters of the Gem S Series workstations as a consultant, but I could be wrong. Anyway not bad for someone who started out as a student of biology and stumbled into modular synthesis by accident.
GTECH was my first choice, but didn’t make it so had to settle for Purdue Electrical engineering…how I wish I had made GT now!!!! This is the stuff I wanted to do EE for!
Great video..
Thanks!
Nice video!
I wonder about the Polyfusion 4-pole lowpass filter's feedback path, as you mentioned it's only dividing down by 6 before going into the OTA, so the OTA should be plenty nonlinear in that operating region? I guess that's probably intentional to put some saturation in the feedback loop and tame the resonance, so that it doesn't go too wild, a bit like the clamping diodes in the Korg MS-20's filter.
Would that be the same kind of thing in the Prophet 5 rev 1, where it multiplies by 10 before going into the SSM2020? That seems really extreme though, but maybe the SSM2020 has a wider input range before saturation than a standard OTA?
Good question. I'd need to dig into the SSM2020 Datasheet.
Another awesome lecture, thank you for putting these together and making them available! I learn so much listening to them. At the end of the video you mentioned using switches to change between different filter configurations. The electric druid website has a link to the "multiple identity filter" which was part of a series of articles in 1979. It uses the CEM3320 and about a dozen switches to do exactly that. I took a crack at building a PCB for it a while back, but made a bunch of mistakes in the design and haven't gone back to it... maybe its time. Anyway, here is the article: www.synthdiy.com/files/2002/MIF.pdf
I think the Oberheim Xpander did something along those lines.
Awesome lecture! I was wondering if the SSM2044 also could be categorized as an OTA-based 4-pole approach? Sadly I'm a little overwhelmed by the patent (US #4404529). Also the famous Korg NJM2069 Filter IC seems to be a big mystery. The only info I can find is a 'sort of' block diagram in the service manual of the Korg Poly 800...
Yeah, the SSM2044 basically has OTAs with differential outputs. It's then sort of kind of a cross between a SSM2040 style filter and a Moog ladder.
Alas I don't know anything about the NJM2069 -- as you say there isn't a lot of info out there.
I still have a hard time understanding how the "local" feedback works in each of the 1p cells.
Have you seen this lecture? In it I focus on these one-pole cores: ua-cam.com/video/pTHHzFsa4Ss/v-deo.html
See around the 13:30 mark
I need to learn math to understand this. Any course on the web?
I recommend the classes on basic circuits on Coursea taught by my colleague Bonnie Ferri.
@@Lantertronics Thank you, I will check it out.
Thanks for this vid! I thoroughly enjoyed it.
I have a couple questions which I hope you could answer.
1. Why does the OBX ota filter stages not have local feedback?
2. Is the type of cap used really that critical? Im curious because when I was prototyping my vcf I used ceramic disc caps because thats what I had on hand. I have read that this is a big no no because of microphonics. To my surprise the cheap little disc caps sound great and I dont notice any undesired behaviour. I may use them in my design.
if the OTA is closed, the knob turned left, isn't the audio just going through the "feedback" loop, opening up the OTA's will subtract their portion of the audio?
Hi there! Sorry I don't think I'm following your question... which part of the video are you referring to?
@@Lantertronics
Hey professor,
well, a "general" OTA / VCA ic based low pass filter "feedback" circuit.
to me it seams the audio is bypassing the filter through "feedback" when the OTA filter cells / stages are CLOSED, the frequency knob turned left. nothing is fed into the capacitors.
i think opening the filter cells / stage (turning the knob to the right) causes subtraction from the original audio going along the "feedback" path. allowing the capacitors to charge. thereby chewing off a portion of the audio.
Afrotechmods - Fun with ultracapacitors!!
ua-cam.com/video/EoWMF3VkI6U/v-deo.html
Afrotechmods, has a good explination.
@@tired.mp3 well, the "general" low pass filter pretty much everyone builds ( some 200+ producers are found on modulargrid alone ), like Ray Wilson, Mutable Instruments, Rene Schmidt, ...
@@tired.mp3 well, it seams the audio should go through the feedback, if you close the filter cell / stage nothing passes through yet the filter sounds.
i'm not sure how much the 100K / 68K ( as Roland used in the Jupiter 8) / 20K ( as seen in professor Lanterman example) resistors found in the feedback loop affects the gain of the filter. since there aren't really voltage dividers.