Sub Audio Noise Source and Multiplier for Random Modulation

I was referring to driving a DAC from a say white noise wave file, not using commercial p[layers. Very simple actually, done it several times.

Reply to
<698839253X6D445TD
Loading thread data ...

PS the effect you describe was always fun with analog tapes, you would hear the RF bias when playing it very slowly...

Reply to
<698839253X6D445TD

I was thinking that the zener noise source won't work for you because it's mostly ~>1us random spikes.. no voltage density at 30 Hz. But you could send the spikes into a flip flop to make random 'square' waves, and then divide that down to some lower frequency. And then filter it if you don't want the sharp edges.

(20V zeners at ~10uA to start.. so you'd need at least a 24V supply.)

George H. (If you mean Tim Wescott he hasn't been around for a while.)

Reply to
George Herold

That is an interesting concept. Instead of a zener, would a BC547 transistor reverse biased at 12VDC provide noise at 30Hz and below? See circuit below. I would remove the high pass filter to earth on Q2.

formatting link

However, for the sake of simplicity I very much like your idea of a divider instead of a filter. Would a CD4060 work?

What would the output look like at the highest possible divisor which is

16384? A train of very narrow pulses at random intervals? These could be fed to a flip flop as you suggest, and then rounded off.

Would it pay to put a simple Sallen-Key low pass filter before the CD4060? If I roll off everything above 500KHz, that divided by 16384 is about 30Hz.

Yes, I did mean Tim Westcott. Thanks.

Robert Martin

Reply to
Robert Martin

Any of those zener-type noise sources is going to have horribly uncontrolled levels of 1/f noise, which will change with time.

I don't know what you're trying to accomplish with such a poorly calibrated noise generator, and I'm beginning to think that perhaps you don't either.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

Am 02.11.18 um 14:39 schrieb George Herold:

OMG. No voltage noise density at 30 Hz?

<
formatting link
>

You can see in the BZX84 series how fast noise grows with Z-voltage. Sorry, I had no interest in large voltages since I had not much interest in lots of noise.

OK, apart of my Agilent 346B noise source.

The LM399 may serve as an example for 7V. And there is a lot of 1/f, and the corner may be quite high for some types.

The lowest trace is the self noise of the preamp. The best test objects may look worse than they are at lowest frequencies because of a coupling cap that was to small. That has been healed in the mean time

0 dB is 1nV/rt Hz, 20 dB is 10 nV/rt Hz and so on.

cheers, Gerhard

Reply to
Gerhard Hoffmann

Yes, I am actually looking for chaotic noise.

My purpose was stated at the outset, and in response to your most recent post I tried to ask a few simple questions.

I am sorry if I missed your point.

Robert Martin

Reply to
Robert Martin

n

as

en

s
e

I never looked for 1/f noise in zeners. Mostly I get big (hundreds of mV) pulses out of them. If instead of narrow pulses I turned it in into random 'square' waves, that would lower the frequency a lot.

Right, I'd have to measure the noise density. (Robert is still learning the difference between Gaussian and white noise. One wonders if just sweeping the frequency and recording results would work better.... noise gives no more information than a bode plot.)

George H.

Reply to
George Herold

To be more precise, at this point Robert would settle for any practical advice regarding the use of random sub audio noise as chaotic modulation.

He does not care if it is based upon amplitude, phase or frequency.

I do know what a sweep is, how to solder and cook my own food, but that was not the intent of my original question.

Robert Martin

Reply to
Robert Martin

t's

ter

535945536/

Besides not being Gaussian the the zeners are not symmetrical. I did one noise source where I added two of opposite polarity, they never quite balance, but much better than a single. The one goal was to have the noise flat to 100 kHz (limit of SRS spectrum analyzer) I think it rolled off about 1 MHz, mostly opamp slew rate limit.

You know random spikes (from a HV zener run near the knee) should (if perfect) give Gaussian white noise over a long enough time frame. Measuring white is easy. I've never tried to measure Gaussian (the amplitude distribution). Can I just use my 'scope trigger? (with some switchable gain ahead of it.)

George H. (shot noise from a photodiode is a great source)

s.

Reply to
George Herold

No one has mentioned the use of a micro to generate the noise, and to provide the appropriate filtering. Either using an on-board DAC or PWM output.

I'm sure the footprint would be less than any 30Hz analogue filters.

--
Mike Perkins 
Video Solutions Ltd 
 Click to see the full signature
Reply to
Mike Perkins

Nah, you can do passive RC filtering with about four 0402s. 30 Hz is a

5-ms time constant, e.g. 0.022 uF and 220k. Given the vagueness of the spec, RCs are probably fine.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

So, you are saying I can filter the white noise with a two stage, passive RC filter? There might not be much left after that.

What should the output level of noise be in order to feed into an op amp?

So far, I have observed nothing in the sub audio spectrum. Can't even get to first base. what type of noise circuit is my best bet to produce an actual output at those low frequencies?

Robert Martin

Reply to
Robert Martin

Well, your noise source is producing 2.5V RMS. If you run it at 3 kHz, then in a 30-Hz bandwidth you'll get about a quarter of a volt RMS. How much do you need? You'll probably need an op amp to buffer it, but op amps come in SC70, so that's no biggie either.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

They are usually AC coupled so you could choose either to take out the decoupling entirely or increase the capacitors to let your sub audio noise signal out. Slowly varying noise signals are easy in software.

ISTR this trick of removing the output decoupling is popular amongst the DIY laser light show brigade for driving the deflection mirror servos.

10x larger output capacitors would probably be good enough.

Daqarta might actually be able to mock up what you propose to do and then you can decide whether your really do want what you have asked for.

formatting link

--
Regards, 
Martin Brown
Reply to
Martin Brown

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.