Nyquist Aliaser build

[PBE6]

I just built the Nyquist Aliaser developed by JC Maillet (genius dude!!) and it's quite fun once you get it going - I was playing the theme to Super Mario Brothers all day yesterday Zelda next!! Here's the project file:

http://www.lynx.bc.ca/~jc/nyquistAliaser.html

I do have to admit it was extremely difficult for me to get it biased properly. Has anyone built this and had similar problems? The 3 trim pots were very finicky. Each has its own narrow operating range to hunt down, which is complicated by the fact that they seem to interact slightly. A lot of tinkering was required to minimize the clock noise, get the control pot working useably over its whole range, and just plain getting it to make sound! Well worth it in the end though.

I did make a few additions to the circuit. I added a bypass 2.2 uF bypass capacitor on the input JFET amplifier to increase the signal to "whine" ratio (which also added some dirt), as well as a 0.1 uF capacitor across the control pot. I'm actually a little unclear as to why this works, but it had the effect of reducing the whine without losing too much range on the pot (larger caps reduced the whine even more, but killed the control too much). Anyone know why this would happen?

Anyways, thanks JC!!


BTW, has anyone tried building JC's improved JFET tube-emulation circuit? Basic schematic is on his Viva Analog website, going to be my next project!

[Mark Hammer]

JC's Aliaser is essentially a sample & hold unit, with sampling taking place in the audio range.

Stock, the gated JFET allows for the .001uf cap to hold the sampled instantaneous voltage of the audio signal, and the JFET that follows it buffers that momentarily stored voltage so that it doesn't leak out.  The .001uf cap is a much smaller value than one might normally see in a S&H, but since the interval it needs to hold onto the voltage for is very short, the circuit permits that cap to charge up fully and quickly and doesn't need it to hold onto charge very long.

But one of the things that allows a person to tinker with is the stability of the sampled signal.  I've recently detailed a "droop" mod for the FSH-1 S&H Filter in another thread, that drains off the stored/sampled voltage, and I think it can be played with here, too.  Try out a fixed resistor in parallel with the .001uf storage cap.  I can't imagine one wold hear very much, or very much of interest, at the highest sampling rates.  But once you start to slow the sampling down a bit, I would imagine parallel resistances in the 33k-470k range might provide some interesting changes to the tone.

By draining the sampled voltage off, instead of accurately holding onto that momentary sample, one further "destroys" the fidelity of the signal, by having each sample drain off before the next one is grabbed, instead of producing a perfect "stairstep" image of the input signal.

[PBE6]

Cool thanks Mark! I will give that a try.

[Processaurus]

I'll have to try drooping mine, neat idea Mark!

I built JC's design into a wah pedal, there was an interesting and fortunate side effect to the oscillator, that when it is at its fastest, it acts like it latches up and sends out a DC signal, so when the rocker pedal was all the way back, the effect sounded like it was bypassed, because the JFET doing the sample and holding was stuck on.

Mine actually mysteriously stopped working a while ago, and I never got around to debugging it, maybe I'll dig it out and fiddle with the trim pots.

[PBE6]

Weird, I had a similar but opposite experience. On guitar, decreasing the sample rate all the way acted as intended, but when I switched to bass suddenly the end of the pot acted like a bypass! Very odd. I also had a situation where it stopped working, and for the life if me I couldn't figure it out. Luckily it fixed itself on the way back to the scope!

This circuit is easily the most finicky I've ever dealt with, fun once it's working though

[moosapotamus]

Yeah, this is a really fun circuit. I added an envelope follower and an external CV input when I built one back in '06.
http://moosapotamus.net/ideas/jc-maillets-nyquist-aliaser-project/

~ Charlie