HMC363S8G eval board noise at Low Frequency

The "spurs" are indicative that the device is having difficulties. Although the datasheet indicates that it'll work down to DC this is a bit misleading - "close to DC"  would be more accurate. As one of our older parts, unfortunately it was only characterized to the "typical" minimum frequency of 200MHz listed in "Electrical Specification" table on page 1 of the datasheet. Below 200MHz the input and output return loss degrade sharply (IRL is ~ -7.5dB at 200MHz); to over come this the input power must be increased as indicated in the input sensitivity plots. Although the plots appear to go to DC, they really only go to 200MHz. Below 100MHz a square wave is required, the higher the slew rate the better and be sure that the DC blocking caps are sized appropriately for the frequency. We've seen some improvement in the input sensitivity at low frequencies by using ATC 530L or 550L broadband caps. Lastly try increasing the input power to 8-10dBm.

Thanks!

I was hoping that by increasing the blocking caps that I would get some more response down below 10 MHz.  After switching all the AC couplers from 100 pf to 100 nF, I can now see normal operation at 9 MHz now (down from 50 MHz).  I do have to drive my input at +20 dBm (with a sine wave, though).  I'll double check and see if I can get a little lower frequency performance with a square wave input.  

I checked, and if I apply a square wave at +12 dBm, I can get operation down to 5 MHz.

Thanks again!