Post Go back to editing

PFD Spurs ADF5355

We have our own PCB design - a fixed frequency application. Target 5GHz.

PN measurements have turned out to be in line with simulations, so that s all good.

Icp .9mA, FRAC1=FRAC2=0 for testing, PFD between 15 and 200 MHz.

We are consistently getting strong PFD spurs at Fc +/- n*Fpfd order of magnitude -45dBc (n=1..4) only typically. Several boards have been tested.

- bleed influences somewhat, a few dB, but this looks dramatically below what I would expect from datasheet (around -80dBc).

Our application use a very clean reference Crystek 100M  locked to GPS/10M using an ADF4107. Nothing wrong with that. Using single ended ref on the ADF5355. Using correct init.

I have been reading up on the datasheet and forum, but running out of ideas.

Can you at least confirm that this is not normal behaviour? 

Thanks for any pointers.


Spectrum plot below for PFD = 50MHz

update plot
[edited by: xtof huygens at 4:47 PM (GMT -4) on 26 Sep 2020]
Parents Reply Children
  • Thank you. I am quite convinced we have good decoupling everywhere, and we have 3 different ADM7150 for the supplies (2 3.3V AVDD DVDD and 1 V5P). We have modified a board with extra decoupling but the result is the same, The coil on the on board oscillator is removed and the board is fed single ended using a either the Crystek or an HP8662a low PN generator.

    The spurs go down as we decrease PFD frequency, and Fc+ +/-2Fpfd is always highest. 

    • ADF5355schem