Reducing noise in hex inverter distortions?

[psychedelicfish]

Recently I've been playing around with my CD4049 on the breadboard, with Runoffgroove's Double D as a base. I've got some pretty decent sounding distortions out of it, but anything with a reasonable amount of distortion is far too noisy. My questions to you are:
What causes circuits using this chip to be noisy?
How can I reduce this noise?

I have googled this topic and have come up with some information (at the "other" place), but I still think more can be done... Perhaps using smaller input and feedback resistors would help?
Thanks,
Edward

[PRR]

CMOS are just hissy.

Different suppliers/batches may be different hiss.

Use low-noise gain before the CMOS, high CMOS supply voltage (so the clipped signal is as far as possible above the somewhat-fixed hiss), then knock the signal down to size after (reducing both signal and hiss).

[Mark Hammer]

Paul's right,  Some are hissier than others.  Given how cheap they are, it's worth having a handful on hand, using a socket for your build, and doing taste tests with the chips you have on hand to identify the least noisy unit.

Personally, I find that I prefer the sound of a 4049 when I'm not asking too much gain out of it.  A simple op-amp or JFET gain stage ahead of the invertors lets one get the desirable qualities out of the 4049 (or 4069) without  adding too much of their undesirable ones (hiss).

Finally, think over how much bandwidth you want and need from the circuit.  That's ne of the other thngs I like about using a simple op-amp stage to push the input of the first invertor.  You can limit the upper part of the bandwidt hitting the chip, and then remove even more bandwidth from each successive stage so that you end up with an output that is fizz-free, not hissy at all, yet still has as much bite as your need.

[PRR]

...with Runoffgroove's Double D as a base.

Sorry, missed this.

That's a good plan. Q1 is the "lo-hiss gain ahead" that I suggested; the volume pots are the "knock-down after" which I suggested. (Which means they should probably not be on "10".)

As Mark wisely says, control of un-necessary high-end response is important, and those hunnert-pFd caps are part of that. Smaller resistors *may* help, but 1Meg is only 14uV audio hiss, and I think the CMOS tends to be more.

[Paul Marossy]

CMOS are just hissy.

Different suppliers/batches may be different hiss.

Yeah, and PT2399s are CMOS devices too. No wonder they are generally noisey...

[pinkjimiphoton]

if it's on the breadboard, it'll be noisier than when boxed, too.

why not use one or two unused inverter stages to make a simple tone control? i found just a really simple cap to ground from a pot can help a lot.
takes out the fizzy and crackly, and makes it much smoother...

look for my "boogie snake" here on the boards for an unusual way to do it that makes little sense, but man, it sure sounds good.
the guy who bought it has it on his live board and loves it.

[duck_arse]

does/would a 4007 wired right make less noise? you'd need a pair for a double'd obviously ....