It's beautiful especially when you think about it that someone made so much effort to just make new sound. It's easy to underestimate it when you think about todays circuitry with microprocessors that just run program with equation that you can program it to do on input waveform. When you don't have such luxury its completely different challenge. Thank you that you are keeping this alive.
Iiiinteresting! I remember once seeing either a VCA or a VCF based on a similar principle of starving the supply rails. But I can't seem to find it now on the googles...
I tried this once on a different opamp than Buchla used, and found that an input voltage greater than the positive supply not only made a protection diode conduct, but triggered a parasitic SCR on the die and latched the output of the opamp. Whoops.
I see the circuit diagram as an opamp with 2 inputs on the -ve terminal with one of them grounded, the other with Vin.. The output will be the sum of these two signals (Vin+Vgnd) at the -ve input.Why do we have a complicated theory? Is there any reason why one should look in your way? Regards
I think you could make a very basic audio compressor out of this, with the deadband cutoff point determining the compression point. . . . This is essentially a core of a noise gate. Op-amp buffers in front and back, possibly translating the signal up to a single-rail reference voltage and you can easily adjust the noise gate volume level by adjusting the core amp's rail voltage.
Hmm, not really and not really. :) What's missing in both cases is a time variation component. This is only an instantaneous voltage distortion effect. What's normally called a compressor attenuates the signal when it gets loud and then returns to unity gain. What you could build with this circuit is more like a soft clipper/shaper. As for noise gate, it would get rid of the noise, but at the same time it would also cut out the middle part of the wanted signal so the tops of the waveforms are moved closer to the zero level. This would create discontinuities in the waveform and make the signal sound distorted and much "noisier" to the human ear.
The deadband wouldn't work as a noise gate since it's effect is instantaneous. It would give you crossover distortion, which might sound weird and cool in its own way... but to do a noise gate you need to get an average of the power of the signal and have the gain respond to that (same with a compressor).
It's beautiful especially when you think about it that someone made so much effort to just make new sound. It's easy to underestimate it when you think about todays circuitry with microprocessors that just run program with equation that you can program it to do on input waveform. When you don't have such luxury its completely different challenge. Thank you that you are keeping this alive.
That op-amp circuit is bananas! Thanks for such a clear and detailed explanation. Analog stuff always breaks my brain.
very interesting Aaron. Looking fwd to next chapter, that Buchla circuit was always so interesting! (259 timbre processor)
Really interesting. Be a really mean exam question though.
I lol'ed at the nintendo shade
Love the lectures. Small mistake at 13:33: the professor meant "continuous" or "piecewise linear" - not "piecewise continuous".
Good catch!
Iiiinteresting! I remember once seeing either a VCA or a VCF based on a similar principle of starving the supply rails. But I can't seem to find it now on the googles...
Finally found it. Published as Four-quadrant analogue multiplier using operational amplifier (Riewruja, Vanchai and Rerkratn, Apinai, 2011).
I tried this once on a different opamp than Buchla used, and found that an input voltage greater than the positive supply not only made a protection diode conduct, but triggered a parasitic SCR on the die and latched the output of the opamp. Whoops.
I see the circuit diagram as an opamp with 2 inputs on the -ve terminal with one of them grounded, the other with Vin.. The output will be the sum of these two signals (Vin+Vgnd) at the -ve input.Why do we have a complicated theory? Is there any reason why one should look in your way? Regards
I think you could make a very basic audio compressor out of this, with the deadband cutoff point determining the compression point.
. . . This is essentially a core of a noise gate. Op-amp buffers in front and back, possibly translating the signal up to a single-rail reference voltage and you can easily adjust the noise gate volume level by adjusting the core amp's rail voltage.
Hmm, not really and not really. :) What's missing in both cases is a time variation component. This is only an instantaneous voltage distortion effect.
What's normally called a compressor attenuates the signal when it gets loud and then returns to unity gain. What you could build with this circuit is more like a soft clipper/shaper.
As for noise gate, it would get rid of the noise, but at the same time it would also cut out the middle part of the wanted signal so the tops of the waveforms are moved closer to the zero level. This would create discontinuities in the waveform and make the signal sound distorted and much "noisier" to the human ear.
The deadband wouldn't work as a noise gate since it's effect is instantaneous. It would give you crossover distortion, which might sound weird and cool in its own way... but to do a noise gate you need to get an average of the power of the signal and have the gain respond to that (same with a compressor).
This looks like it might be abusive to op-amps. Is there a hotline that I can call?
Hah!