Output Variation of HMC733


    I want to generate a chirp signal which sweep from 20GHz to
27GHz. My circuit is just composed of a VCO (HMC733) and a frequency doubler (HMC576)
and both of them are evaluation boards. When I control the tuning voltage, I should
have a discrete chirp signal from 20 GHz to 27 GHz. I do not have a spectrum analyser
to measure the output so I used a power detector (ADL6010) to measure the
output power.

    I found that when I do the sweeping, the output power of individual
frequency point is not constant and the tiny variation is changing slowly with
time. For example, if I record the measurement of 100 sweeping during the band
and just take the 100 samples for one frequency point (25GHz), the variation is
around 0.3 dBm. I know the variation is not huge. For different point, the
variation is different but similar. BTW, my sweeping time is 1 second and the
sweeping point is 1000 from 20 to 27GHz.

    I did the same measurement with the output of HMC733 and I got
the same results.

    The variation is obviously not the output voltage distribution
so I think it comes from phase noise. Could I choose a better circuit to solve
this issue?

    I found ADF5355 can provide better phase noise performance
than HMC733 and it is cheaper. Should I replace HMC733 with ADF5355 to produce
the chirp signal? Will it solve the issue or improve the output constancy with
time? Do I have a better choice? Is there a better solution to generate my chirp

    Thank you very much.