I am attempting to do a simple record of a CW, 250MHz signal at the Rx port of an Ettus B200 using an Anritsu MG369 signal generator. I am using the 10 MHz ref signal provided by the SG (output @ ~9dBm). The frequency/phase of the signal seems to be too unstable from what I should expect. Is this common for the B200? I am expecting a B210 soon and am just trying to get ahead of the game while awaiting the new hardware, but if this is user error I'd like to figure it out in advance if possible as I eventually will need to consider phase closely when test is in its final form. I've seen in Ettus literature that a square wave is best for reference, could a sine reference be the cause of this?
Could it be the PLL misbehaving or an amp somewhere in ref circuitry? RF front end is confirmed to be working okay and can see a clean, stable signal when running CW from Tx through cable to Rx port.
Including flow chart as well as two pics of spectrum to try to illustrate instability.
EDIT: Including better FFT sink snapshots