Noise generator using device self-noise

[Rob Strand]

I did try Zener for a noise source but it was pretty feeble.

FWIW,  the other day I did a half-hearted poke around through my stuff on the subject.  Unfortunately my notes are on subject got archived when the house was re-carpeted.  A common problem is tuning the current (that includes reversed VBE) so the noise is strong but also also nice - sometimes you get really weird waveforms.   

The RF guys quite often use 6.2V zeners and they tune the current because there's a small window of currents where the noise is strong.  This straddles the Zener/Avalanche region and is also near zero TC. One paper mentioned 6.2V zeners have a low noise TC.  Unfortunately I can't remember if the noise was any good for audio.

I may be wrong, but "resistor noise" should add in a rms way

It's correct.  You just need to add the squares of the noise.
Each resistor,      Vn1^2 = 4 k T B R1,  Vn2^2 = 4 k T B R2
Series resistors, total noise voltage  Vn^2 = Vn1^2 + Vn2^2 =  4 k T B R1 +4 k T B R2 = 4 k T B (R1+R2)
so the noise is the same as using the total series resistance.

[PRR]

OT: Tom makes digital noise sources and sells them honestly cheaper than your breadboard labor is worth.

Digi-noise has a bad name from the old short-lines, but today's PICs allow lengths which will not repeat for "142 million years".

https://electricdruid.net/noise2-white-pink-noise-source/
https://electricdruid.net/white-noise-source/
...etc...

[Rob Strand]

https://electricdruid.net/noise2-white-pink-noise-source/
https://electricdruid.net/white-noise-source/
...etc...

Nice write-ups too - thanks Tom!

Following through some of the web-sites,  there's this interesting paper about subtleties of noise, [click on the blue dot in the calendar]
https://web.archive.org/web/20110501000000*/http://users.informatik.uni-halle.de/~thielema/Research/noise.pdf