low noise JFET, how about this one

[jackinnj]

Saw this on Wenzel Associates website -- an ultra low noise FET amplifier for measurement of phase noise -- very low noise at very low frequencies --

An externally hosted image should be here but it was not working when we last tested it.

 

[tschrama]

Thanks jackinnj!

 

[DRCG5]

Thanks for the suggestions!

Indeed it seems like a RF device, but what do you think: should it be low noise down to the 1KHz region? 10Hz even?

I was dreaming about building a < 0.5nV/sqrt(HZ) amplifier.. should I stick to the 2SK170? Or BTJ, given that the Rsource is < 1KOhm...

I have a article decribing a 0.33nV/sqrtHZ JFET amplifier using 9 sony 2sk152 transistor in paralellel ... actually .. i wanna beat this thing


thanks!
Thijs

Hi -

Microwave transistors typically have a 1/f corner in the hundreds of kilohertz at least and could be as high as MHz.
Transistors which are diffused and not ion-implanted are the quietest. Ion implantation shatters the crystal creating traps which are leaky and noisy. This has to be annealed out and is not very well done at lower frequencies.
JFets, while avaiable at 1nV/Rt-Hz at say 100Hz, have very bad 1/f noise.
A good chopper op amp like the LTC1250 chopper op-amp has a noise of 100nV at 10Hz, which is not the greatest, however the LTC1250 noise decreases at lower frequencies dropping to below 25nV at 0.001Hz.
The lowest noise op-amp in the world (for 1/f) is the MAXIM MAX4238 chopper op-amp. The MAX4238 has a minimum closed-loop gain of 1000V/V.
These op amps have an open-loop gain of a million and are dead linear at gains of 1000 V/V.
JFets are trouble.