Any suggestions for a digital ready random noise circuit?

[blueroomelectronics]

Looking for simple circuits that generate randomly variaing width pulses or amplitude waveforms. As simple as possible would be nice, commonly available / inexpensive parts preferred. It's for the next JPUG issue (A PIC newsletter)
**broken link removed**

 

[on1aag]

Hi Bill,

The only thing that pops up in my mind is a MM5837 from
National Semiconductor, but that chip is a real dinosaur.
A quick look at the datasheet might give you a clue on
building your own from standard logic chips.

MM5837 datasheet pdf datenblatt - National Semiconductor - Digital Noise Source ::: ALLDATASHEET :::

on1aag.

 

[blueroomelectronics]

Thanks for the link, making random noise is an art it seems. I recall something using transistor junctions...

That digital noise chip is similar to the software one in BASIC

   Seed = Seed * MagicA + MagicB
   result = LongWord(Seed >> 16) Mod RndMax

Wait till audioguru sees the old 741, well I do want noise

 

[nickelflippr]

I tried XOR'ing a light to voltage chip with the software random generator. Seems about the same as just the software. Kept reading about using a zener diode as a random generator, but didn't actually see how you implement that.

Since this was for a RGB led, a roll of die scenario might have been better. The random generator gives lots of crappy colors and inumerable shades of white. The best case seemed to be just using the random number, and discounting a wide swath of middle values.

 

[blueroomelectronics]

Well the article is for the Junebug magazine, one project shows a software based pseudo random number generator rolling two dice. Another is a freerunning modulo 6 timer waiting for a user to press a button (truly random unless you're a robot). I'd like to add a true hardware version, but this may be too complex for practical reasons (IE the ones I've seen require about 10V to get going)

 

[audioguru]

Hi Blue Bill,
It is good that you are powering the lousy old 741 opamp from 18V since it probably will not work from 6V when one 9V battery is old.

The transistor noise generator might not work from only 6V to 9V but will work fine if you feed them from 12v to 18V from both 9V batteries.

The LM386 little power amp will probably oscillate at 300kHz without the RC Zobel network at its output to ground as shown on its datasheet.

 

[nickelflippr]

Yea, when using a timer as a seed, which in turn is accessed in software at a more or less predictable time or sequence seemed to be my problem. The light values seemed to take on a repeating color pattern, even tho that pattern may be developed over tens of cycles. No means scientific, but more like a perception thing. Keeping a log of the values, or reading them out on hyperterminal might have proven otherwise?

The human intervention aspect has to be the ticket I would think.

 

[blueroomelectronics]

the lousy old 741

Have you made that a macro key

I guess I'm hoping for the improbable with only a standard USB 5V supply, if it's more than a couple of transistors and an IC or two then I'll pass on it.
One solution might be a pink or white noise program for your PC and feed from the line out to an input pin on the PIC. Just a thought.

 

[Pommie]

Have you considered the 5V supply itself. If you divide it in half and feed it into the ADC then the bottom bit should be pretty random. Read it eight times and you have a random byte.

Mike.

 

[blueroomelectronics]

Interestingly enough the Swordfish random number module suggests leaving an A/D pin floating to seed the generator.

 

[Mikebits]

How about two unstable oscillators driving the D and clk pins of a FF.

 

[crutschow]

Kept reading about using a zener diode as a random generator, but didn't actually see how you implement that.

You can generate random noise by amplifiying the noise current from a zener diode. Attached is an old ap note which describes running the zener current into the base of a transitor. It shows a 10V zener and a 26V supply but it should scale for lower voltage zeners and supply voltage.

You can apply the noise to a comparator input if you want a digital signal.

(Sorry about the picture being upside down. You'll have to rotate it in the Acrobat viewer.)
View attachment Noise Gen.pdf

 

[audioguru]

The emitter-base of a silicon transistor has avalanche breakdown at from 5V to 10V like a noisy zener diode. That is what Blue Bill has in his circuit then the nisey transistor is amplified by the second transistor.

 

[k7elp60]

Bill,
Here is a schematic of a circuit I built several years ago that lights 1 of 8 leds. It seems to be in random fashon. With pull up resistors instead of the LED's and the single resistor it could be a random 8 bit word.
Ned

 

[ericgibbs]

hi,
Maxim have a white noise gen, using zeners.

https://www.electro-tech-online.com/custompdfs/2008/06/AN3469.pdf

 

[crutschow]

The emitter-base of a silicon transistor has avalanche breakdown at from 5V to 10V like a noisy zener diode. That is what Blue Bill has in his circuit then the nisey transistor is amplified by the second transistor.

Yes, it is similar to the zener circuit I posted.

But a significant difference is that the zener cathode is decoupled by a resistor and capacitor to eliminate negative feedback from the collector, which would otherwise significantly reduce the noise gain. That's probably why that circuit can generate 15V noise with only one stage.

 

[Roff]

It's difficult to get true average 50% duty cycle from an analog noise source, due to harmonic distortion, etc. Think about running the resulting digital signal through a fast (relative to your noise bandwidth) toggle flip-flop. The resulting signal should average 50% 1's and 50% zeroes, with the only error being the skew between propagation delays of clock-to-output on positive vs negative output transitions.

 

[Ubergeek63]

Linear feedback shift register - Wikipedia, the free encyclopedia



you could easily find many more variations