[time-nuts] Phase Noise at 5-10Mhz

[Tom Knox]

I was wondering what kind of Phase Noise performance benchmarks Time Nuts have achieved in their labs. I have been rebuilding my Time and Freq test system for the last few years and at times it has been humbling to say the least. I am finally seeing light at the end of the tunnel and seem to have gotten the systemic noise down to where I can really start comparing the individual oscillators I have come across over the years. My approach was to place everything in several Agilent equipment racks and I have even questioned the wisdom of that more then once as I have struggled to set up a state of the art system. My system can be Phase locked, and the various quartz oscillators can be configured in series and parallel so each element of the system can be compare. I know the simplest approach is a Phase Noise test set a Ref and DUT oscillators some batteries loose on a bench with some filters and  devices to break ground loops. So what real word combined uncertainty number have you been able to achieve at 5-10MHz at an offset of 1Hz 10Hz and noise floor. I struggled at several points originally with systemic noise due to ground loops from all the LAN, USB, and coaxial cables interconnecting the system elements and "soaking" between reference signals due to sheilding issues from normal RG/58 cables and the verious Cesium, GPS, and Quartz standards. Slowly the system has improved from a shaky start for 5MHz at around -105db @ 1Hz toward my compromised goal of 5Mhz at -120dB @ 1Hz. I am hoping find ideas on how to surpass -120db @ 1Hz.  I have heard some impressive number from some of the distiguished members and it would be interesting to how those numbers were achieved, and what was used as a reference and measurement system. Thanks and Happy New Years.

Thomas Knox