Diode noise generator

SparkyEngineering said:

Hey guys,

I'm working on designing a noise generator, and I was hoping I could get some info on the characteristics of diode noise--specifically noise in a zener diode operating in reverse breakdown. Here are my basic questions:

1.) What are the characteristics of diode shot noise?
is it fairly flat? what does it look like when plotted over frequency? does it drop off at some point?

2.) What is the best configuration to produce broadband noise?
multiple diodes? Does biasing/current in the diode affect noise?

3.) I'm trying to get a noise generator from the low khz to somewhere in the 5Ghz range. Is this even possible? Tips on how to do it?

4.) Any other suggestions from broadband white noise production would be appreciated

~I'm one semester away from a B.S. in electrical engineering, so you don't have to oversimplify stuff. Any responses appreciated. :lol:

Thanks in advance!
~Sparky

Click to expand...

jiripolivka said:

1. Zener diodes biased in reverse (at Zener voltage like 6 to...32 V) do generate a good noise level but due to a high junction capacitance, the bandwidth extends to several MHz only.

2. A common P/N junction can generate noise in avalanche mode. I have used many RF and microwave noise generators "for free". Take a RF SMD transistor, any type, and connect its e-b diode through a ~1 kOhm resistor to a DC power supply. A typical reverse breakdown occurs at 6 to 8 V (I used a 9V battery for the power source)_ and use a variable resistor to adjust 5 to 10 mA current. I connected such e-b diode in an half-wave dipole, to make C-band and Ku-band "noise radiators". At 5...10 mA the noise bandwidth was quite flat, over 9 to 12 GHz, and the measured ENR was ~30 dB. Compared to expensive ($80) noise diodes, this is good solution.

3. If you need a higher ENR, a RF or a microwave amplifier can be added. Such "amplified" noise sources are offered e.g. by NoiseCom and other companies. I tested also a 3x3 matrix of my "noise dipoles", to achieve the >5x noise intensity on-axis, no amplifiers needed.

If you are interested, google Polivka : Microwave noise fields about my experiments.

Click to expand...

SparkyEngineering said:

Thank you so much for this great information! I was definitely not expecting somebody so knowledgable to reply to my post. I'll definitely browse through some of your research when I get some spare time.

I believe i'll do exactly what you said and play with a diode in avalanche routed through some filters and amps. This will definitely give me a few hours of quality one-on-one time with a spectrum analyzer haha.

Couldn't help but notice your location is Santa Barbara. I'm actually applying for acceptance into UC Santa Barbara to get my masters in RF/Microwave telecom

Click to expand...