AnsweredAssumed Answered

ADF4351 Output power and phase noise

Question asked by mtwieg on Sep 11, 2015
Latest reply on Sep 21, 2015 by mtwieg

Hi all, I just got my EVAL-CN0285-EB1Z (wideband TX modulator with ADF4351 and ADL5375) and I'm just looking at the outputs of the ADF4351. I've disconnected the LO inputs of the ADL5375 from the RF outputs of the ADF4351 by removing C70 and C71, and am measuring RF power on one of the LOI connectors. I'm using default settings, with the RF frequency set to 250MHz, using the 25MHz reference, so the VCO is at 4GHz with DIV=16. I've also replaced the RFout choke inductors with 390nH to operate at the lower RF frequency. My output power is set to +5dBm. Below is what I'm seeing on the output (marker table on the bottom shows the carrier power at -2.46 dBm, and phase noise at 100kHz and 1MHz offsets).


So I expect about -1 dBm of power since I'm only measuring one of the differential outputs, but I'm only getting about -2.5 dBm. But much more importantly, my phase noise is much higher than expected (from the ADF4351 datasheet and ADLsimPLL). At 100kHz I'm getting -105.5 dBc/rt(Hz) where I expect less than -120 dBm/rt(Hz). At 1MHz I get -120 dBc/rt(Hz), where I expect less than -150 dBc/rt(Hz).


The only obvious cause I can see is that I'm not doing a differential measurement, but I would only expect that to make a 3dB improvement in phase noise, not nearly enough to explain this poor performance. Also I notice that my phase noise plateaus until about 400kHz, but the loop filter on the eval board is supposed to be 70kHz.


For what it's worth, I had looked at the phase noise and power before modding the inductors and capacitors, and saw similar performance (at 2GHz, anyways), so I don't think I broke anything.


Any ideas? Do I have a defective part somewhere?