In the end I found the problem. The Firewire PCI Express card retaining screw to the computer cardcage was about 1 turn from tight. The Firewire cable had multiple ferrite clamp on beads over it too. This is the Crux of a lot of problems people experience with the Flex3K/5K. Speculation appears to suggest that the Fireware cable at the computer side needs a robust RF path to the computer chassis. How to to that is another question.
Jim
Wd5JKO
Hi Jim,
Nice job tracking down the loose retaining screw. Having run full legal limit from 160-10 meters with a Flex 5000 and now with an ANAN-100 I can say that I had a few adventures tracking down the causes of RF ingress. Once I bonded the SDR, the computer, and the external audio chain to a common point on the ground bar at the back of the bench most of it went away. I also found that using a 1:1 600 ohm transformer at the line input of the SDR helped to keep RF out.
One tool that helped me more than anything else was a field strength meter. I powered up the rig and amplifier, dumped around 500 watts onto an empty frequency and then held the field strength meter near every wire and cable that entered and exited the station. I found the last RF path into the station using that method. This one was quite a pain because on 160 meters my "touch control" Samsung monitor was powering down randomly while I transmitted. Of course this is more than inconvenient when you need the screen to see the operating software! At any rate, it turned that RF was flowing back to the operating desk along the rotor cable which terminated in the Rotor control box that sits neatly right under the "touch control" power button of the monitor. At the window threshold where the cable enters the station I looped the rotor cable several times and snapped a few big mix-31 ferrites over it. That cured it and I have not had a trace of RF feedback issues for several years now with the SDR stuff.
At any rate, I hope you can enjoy your Flex without any further RF ingress issues!
73,
Rob W1AEX