Low frequency noise generator for LFO use

[knutolai]

Hi I need a very low frequency noise generator for use as a LFO in a pt2399 chorus pedal I'm designing. Finding any good resources online proved quite difficult. I know I could multiply waves with primenumber frequency relationships, however this seems a little complex and I want to explore other options before this.
Alternatively I could use a Arduino to generate the LFO, but for the same reason as with the wave-multiplier I wish to hear other suggestions before taking steps.
A last possibility could be to create a simple white noise generator as in the image below and then filter away everything above 10Hz-ish, but for some reason I don't think that would work.


Any and all suggestions are very welcome!

[Digital Larry]

Do you want it to give a stepped random waveform like a sample and hold?  Or is it OK if it moves around smoothly?  I think the reverse biased diode (or transistor) noise generator approach is actually well documented and should work, but it might need some tweaking to get the signal into the voltage range you need.

[knutolai]

Do you want it to give a stepped random waveform like a sample and hold?  Or is it OK if it moves around smoothly?

Moving around smoothly is just what I want! Ill have a look around!

it might need some tweaking to get the signal into the voltage range you need.

Don't you mean frequency range?

[Mark Hammer]

Do you want it to give a stepped random waveform like a sample and hold?  Or is it OK if it moves around smoothly?  I think the reverse biased diode (or transistor) noise generator approach is actually well documented and should work, but it might need some tweaking to get the signal into the voltage range you need.

That's just it.  If one simply filters the bejeezus out of a "normal" noise source, and crosses your fingers for randomness, there's a pretty good chance that sample N-1 and N+1 are going to be very close to sample N.  If that's compatible with the desired outcome, great.  If not, then a different strategy is required.

If the intent is to simply have aperiodic modulation, then you'd be surprised what two sources, perhaps one more rampish and one more sawtoothish, can yield.  It's not "noise", but it sure ain't all that predictable.

[PRR]

> simple white noise generator as in the image below and then filter away everything above 10Hz-ish, but for some reason I don't think that would work.

Why not??

Random hiss extends to zero frequency (called "drift"). Invariably it *rises* at some low frequency (typically low audio band).

This is how the early ARP got random low modulations.

I'm not sure about that R2 value, and C5 (and any coupling-cap) might have to be much bigger to pass sub-audio rates.

As Mark says, a truly random source with a heavy high-cut tends to be unexciting.