OSC with PN > -150dBc
I am considering to built Oscillator (crystal) vs Phase Noise about 150dBc/Hz @ 10kHz, can it be possible!?!?!?
David
mmmmmh , it sound hard to built. But someone has made it. The Expensive, very performant, OCXO are perhaps the lowest phase noise oscillators ( even that Rb O).
In my knowledge, the keywords are:
- SC cut rather than AT cut
- excellent thermostatic oven and large time time constant (big mass)
- low noise circuitry
- dynamic range as large as possible
- excellent vibration mitigation
okay let complicate
I considering to built it internal (in to the IC 130nm)
Maybe i am wrong, I am looking for very low jitter (cycle to cycle) low than 100psec
please advice!
Check the low noise crystall oscilator proposed by dr. Ulrich Rhode in the '70. The output is taken from a small resistor (10-22Ohm) connected in series with the crystall., and the amplified.
Since the requirement is @10kHz and fixed frequency maybe you can add a sharp crystal filter at the output to reduce the noise.
I familiar with this technique. this resistor suppose to protect crystal!
In the description of that circuit it said that coupling the output that way ensures low noise, because the signal is filtered by the crystal itself.
Okay,
I tried to find this article but without success, do u have it?!?!
See figures 4-47 and 4-48 below from Ulrich Rohdes book "Digital PLL Frequency Synthesizers, Theory and Design".
He claims 168 dBc/Hz at "a few kHz" carrier offset of a 5 MHz crystal oscillator design.