Comparing John Ackermann's data from 2007 with my data.
2 thoughts on “combined_pn_plot_with_notes”
For measuring PN of the distribution amp something with a lower PN floor than an 8607 is desirable,
Even the OCXO in some Thunderbolts is a lot quieter than an 8607.
Whilst tere is some cancellation of the source PN it isnt perfect and a lower noise source is useful.
The thunderbolt output signal SSB PN is severely degraded by the choice of output buffer amplifier.
Much longer measurement duration is also advisable particularly for low offset frequencies.
I typically use at least 1000 sec sometimes several hours.
Particularly when using correlation techniques together with an inteferometer to reduce the instrument PN floor below -200dBc/Hz.
Hi Bruce,
I can test a few OCXOs we have and try to find the best one.
Yes I should measure for a longer duration. I have one noise-floor test that ran for 12 hours and it starts out at around -180dBc/Hz at the right edge (100kHz) - so the floor is up to 10dB better and also less noisy in the 12h data.
For measuring PN of the distribution amp something with a lower PN floor than an 8607 is desirable,
Even the OCXO in some Thunderbolts is a lot quieter than an 8607.
Whilst tere is some cancellation of the source PN it isnt perfect and a lower noise source is useful.
The thunderbolt output signal SSB PN is severely degraded by the choice of output buffer amplifier.
Much longer measurement duration is also advisable particularly for low offset frequencies.
I typically use at least 1000 sec sometimes several hours.
Particularly when using correlation techniques together with an inteferometer to reduce the instrument PN floor below -200dBc/Hz.
Hi Bruce,
I can test a few OCXOs we have and try to find the best one.
Yes I should measure for a longer duration. I have one noise-floor test that ran for 12 hours and it starts out at around -180dBc/Hz at the right edge (100kHz) - so the floor is up to 10dB better and also less noisy in the 12h data.
Anders