Recently I bought a couple of evaluation boards for the ADF4351 Frequency Synthesizer (EVAL-ADF4351EB1Z board) with the aim of using it as a LO for an optical PLL. But it turned out that I could not lock my signal using these boards, so I checked its phase stability. In the attached pictures you can see two signals in an oscilloscope, one is generated by the board and the other one is generated by a very reliable Marconi 2022D Signal Generator; both devices have been locked to a 10 MHz reference signal coming from an atomic clock. You can see that when the oscilloscope is being trigger by the Marconi signal, the other signal is jittering a lot. This phase noise change with the frequency chosen but it is always very high (The pictures have been taken with f=500 MHz and f=1 GHz respectively).
Also, I perform a quadrature measurement, using a frequency of 50 MHz the rms of the phase noise is 105 degrees.
This does not match the specifications of the device and it makes the board useless as a LO. I would like to know whether this has been reported before, or maybe someone could point me out what I am doing wrong.
I am using the drivers and control software provided by analog devices and for this measurements I have use the default settings in the program (only changing the reference signal field to 10 and the frequency field).