at least on a laptop...
I learned an important lesson today, and maybe it might be of use or interest to others.
Today I installed the remotely tuned NDB loop outdoors, and began testing. Tuning the loop requires communicating with the loop server SBC via a long network cable. Here I use a CAT5 cable for the long haul, with a pair of USB to CAT5 adapters (and vice versa) at each end, one at the computer and one at the loop. I anticipated some possible noise issues, but nothing like what I encountered. The noise floor was about 20dB higher than what I would like to see, and I had intermittent communication problems which I had not noted when testing in the basement. The communication link would unexpectedly lockup for no apparent reason, and the controller program on the PC would hang and become non-responsive.
I added the obligatory 5 pounds of ferrite to the system without much luck.
The fix was rather simple, and almost embarrassing to say. For this first test, I plugged the USB adapter into the most convenient port, which was the USB port right next to the laptop power adapter socket. It turns out that this USB port is just full of noise from the laptop switching power supply which was being communicated to the loop. It is also likely that the switcher noise was causing the local USB to CAT5 adapter to misbehave.
Moving the adapter to another USB port solved the problem, almost in entirety. There is still some spurious junk being picked up by the loop, but we will work on that in the future.
Bottom line is: choose your USB ports with care, and test them all for best noise performance. While you may not have a USB connected antenna, it is possible that a noisy port may be corrupting a sensitive outboard device such as a USB sound card or SDR.