1/f noise

Was doing my morning reading of various emails and magazines I came across this Analog Devices chopper op-amp. The claim to fame is :"This architecture enables temperature drift suppression to 20 nanovolts per °C (typ) and eliminates 1/f noise. "

http://www.analog.com/en/amplifiers-...Amp/press.html

Anyone familiar with why there would be no 1/f noise in a chopper op amp? What is the mechanism?

Why I ask is, can we apply the same architecture, somehow, to microwave oscillators and eliminate their 1/f noise? Before you laugh, we already have class E microwave amplifiers, which are basically "chopper" or "digital sampling" amplifiers. Why not an oscillator too?

Rich

The chopping action nulls out 1/f noise but the chopping action produce wide band noise in return.

What is the mechanism? A high Q resonator can "filter out" far out noise, whereas 1/f noise, according to leasson's equation, determines the whole show.

http://metrology.hut.fi/courses/s108...o4/chopamp.pdf

I don't see why it wouldn't work. In fig 3.47, if you are using a parallel low frequency amplifier A2 to null out the 1/f noise in the microwave amplifier A1, you should end up with a no-1/f noise oscillator.

Previous discussions on this board about a DC feedback loop with high (90 dB) gain will do the same 1/f noise nulling in ultra-low phase noise XTAL oscillators. Maybe here is a way to do the same thing, but with a simple chopper.