[evlatests] correction Re: L302 phase noise.

[Vivek Dhawan]

I forgot about the eternal factor of 2. If the spectra are SSB,
(I believe they are), then the phase noise is really 1.414 times
higher,

i.e. 1 Hz to 1MHz
old: 3.5 deg rms,
new: 6.4 deg rms.

Still not a significant loss: < 10% at 50 GHz.

(I hope the spectrum analyzer is not being smart and adding 3dB
before it displays dBc/Hz vs. offset?)

Vivek Dhawan wrote:
> Given 2 spectra one each for the old and new designs,
> measured over 10Hz to 1MHz. The new one is worse by over
> 20dB in the range 1KHz to 100KHz, and worse by about 5dB
> at 10Hz (-45 dBc old, -40 dBc new, 10Hz offset from carrier)
> 
> The integrated phase noise from 10Hz to 1MHz for each is:
> (at 12.8 GHz carrier)
> 
> Old: 0.9 deg rms.
> New: 3.5 deg rms.
> 
> The crucial range from 1Hz to 10Hz was not on these plots, but
> Rob Long reports (from another measurement) that the phase noise
> was certainly 30dB or more down from carrier. Assuming a
> conservative -30dBc for each, and extrapolating in reasonable
> fashion from 10Hz to 1Hz, the integrated phase noise 1Hz to 1MHz
> is:
> 
> Old: 2.5 deg rms.
> New: 4.5 deg rms.
> 
> No significant loss here, assuming the same spectra are valid
> at all times/antennas.
> 
> Vivek.
> 
> _______________________________________________
> evlatests mailing list
> evlatests at listmgr.cv.nrao.edu
> http://listmgr.cv.nrao.edu/mailman/listinfo/evlatests
>