Great references. There seem to be some reservations towards the polarity of the input signal and the control signal when using a 4066, apparently when switching a bipolar signal the control signal needs to be bipolar as well. OK, but for uni-polar signals this should work. And a 74HC4316 seems to remedy that. Something to look into some more, me thinks.
Curious to know whether this would work I experimented a bit following the suggestions in the reference. I used a 555 as a pulse to control the switch on a 4066 while feeding it the signal from a synth. The 4066 was powered by 8 Volts to GND, so was the 555. This resulted in a chopped signal, but there was clearly something clipping the negative part of the input signal. When I connected the negative supply input of the 4066 to -6, the chopping was much better. A small signal during the āoff periodā, and there were clicks, which is to be expected when chopping a signal. Raising the negative supply to -8V stopped the chopping. The overall supply voltage of the 4066 can not be higher than 20V according to the datasheet, so if you want to use this to chop some audio, you need to keep an eye on the supply and control voltages.
Interesting, I hadnāt really thought about chopping an audio signal but thatās got some good possibilities. As you mentioned earlier there probably are better switches for that purpose.
I called it chopping because that is in essence what the 4040 would be doing, wouldnāt it? B.t.w. the chopping may result in audible clicks as it is not synchronous to the signal.
Are you planning what I described here, only with the wave shaper?
If the wave shaper is a wavetable, you should be able to feed this into it for the same results since the pulse will tell the samples if the waveform to sound at different frequencies. It might sound like a janky aliazed mess, but it could sound fine.
The so called wave shaper is a phase locked loop feeding a ripple counter. The /8 output of the ripple counter is used to drive the phase comparator of the PLL with the result that its outputs are x8, x4, x2, x1, /2, /4, /8 (thatās all the outputs that are used). In the original Barton circuit those go into a mixer with manual controls. In what Iām looking into, those outputs would control analog switches whose inputs are control voltages. (No reason the inputs could not be audio signals, with probably crazy and interesting results, but I had slow CVs in mind.) Then the switch outputs are the ripple counter outputs scaled by the CVs. Those go through a mixer to the output. So itās just a voltage controlled version of the wave shaper.
Not sure what youāre getting at, but wavetables donāt enter into it.
The fast counting would however chop them up in very short pieces at the speed of the input waveās pitch (and multiples of that) that would be added together in the mixer. You want that summed signal to be a CV again I think. What purposes would you use the resulting fast jumping signal for?
The moduleās input is an audio signal. The outputs of the ripple counter are audio frequency square waves with amplitude equal to VCC and frequencies equal to the input *8, *4, *2, *1, /2, /4, /8. The switches just convert them to the same audio frequency square waves with amplitude equal to CV. They just act as VCAs. Theyāre mixed and the output is an audio signal.
Again, itās just the wave shaper but with voltage controlled mixing of the different octaves.
I just didnt know what the waveshaper was, thats all. I just have experience with an astable 555 and using the 4040.
Something like this?
Lets give it a name: Barton Wave Folder with a Twist
Note, this is an untested draft !
V1 is an audio signal, V2 is some CV, V0 is the output voltage.
Iāll give this a try tomorrow.
No, because a single CV on all octaves just alters the overall amplitude. Might as well just put the Barton circuit output into a single VCA. Rather, the idea is to have (up to) 7 control voltages, so that each octave can be modulated independently. Put attenuators on the CVs instead of on the mixer inputs for manual control (with CVs normalized to +5V or whatever).
That is not entirely so. Each octave has its own potentiometer thus its own level. Putting them all through a VCA would change them all in the same way.
āso that each octave can be modulated independentlyā ?
Should that not be the other way around ? The octave (i.e. each counter output) modulates the CV?
I think it is time you sketch a draft because we seem to be missing each others point.
Did you ever try this draft you had drawn up? Iām curious to know if itās worked for you as I would be interested in trying it.
No, I decided against pursuing the idea because I believe the switching, given that it is not necessarily synchronized to the signal being switched, will introduce all kind of harmonics. This could be prevented if the switching took part at the zero crossing of the signals the 4066 is switching. So if you want to give that a try Iād be interested in hearing the results.
Finally got my voltage controlled version done:
defiantly would like to hear more about this , I have one of Bartonās 4046ās and like it . CV would be really nice add to these !
ok that was quick ā¦ I just saw this .
More evidence @analogoutput is a timelord.
He does go by the name of The Doctorā¦