[hfsig] Sidebands on Raspberry Pi WSPR carrier

[Ron Ott]

Last night I completed some revealing tests with the Raspberry Pi (=RPi) TX with TAPR 20m WSPR kit.  The bottom line is that grounds are the cause of the observed spurious sidebands.
* I see no sidebands when the TX is connected to a 50-ohm load and my HP scope with a floating ground input.  I'm receiving on an SDR with an outside 20m vertical antenna, picking up stray RF.  In this setup the RPi TX is fully isolated from any grounds in the shack.
* When I move the 50-ohm load to the input of an HP frequency counter with a grounded-shield (1Mohm) input, I see several sidebands on the carrier that are multiples of 60Hz. In this setup the shield side of the RF output coax is connected to the mains ground through the frequency counter.  I decoded four (4) WSPR signals simultaneously - the carrier and three of the sidebands.  
* I played with ground leads to the RPi USB connector shells and to the ground side of the RF output terminal strip, but there was no change in the sidebands.  Since the RPi is a small computer, the grounds are not that easy to discern and it may take some experimenting to discover what works. 
* Potential solutions: Perhaps running the antenna lead directly to the transmitter without a ground connection in the shack might help - an experiment for tomorrow.  An un/un transformer might be used to isolate the RPi grounds from the shack ground and avoid the ground loops.  I'll look into making an un-un myself or buying a commercially built one from Balun Designs.

Ron, W6XY
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.tapr.org/pipermail/hfsig_lists.tapr.org/attachments/20160905/5044f5f4/attachment.html>