Rob French (KC4UPR)
Well, I guess this answers this... I have so far had no luck killing the fan electrical noise. It's in the low-hundred Hz, with some harmonics (though at least on my digimode waterfall, the only spikes I see are < 500 Hz, and I don't see anything visible in the spectrum display of my TX output). Very noticeable (audibly) when I turn my volume down, mostly unnoticed when I turn the volume to normal listening levels.
I use a 1000uF electrolytic. No resistor in series, though I guess I could try 10 ohms or so... I had a larger resistor in at one point, and it slowed the fan down too much. I also have a 1N914 diode in series... not sure what I was thinking, but I haven't removed it. Large (megaohm?) resistor across the electrolytic terminals.
I have found the fan indispensable for running digimodes. I have my uBITX v5 setup for a solid 8-9 watts on most bands (except 10M). FT8/FT4/etc, the finals stay cool to the touch, even with many many transmissions in a row... 15 second transmissions just aren't long enough, and with the fan running continuously, any excess heat is quickly shed during the RX periods. However, with long transmissions (3-4 min) using Olivia, the finals definitely get uncomfortably hot to the touch, even with the fan running. Now, if the other guys transmits for just as long, the fan does manage to cool the finals back down pretty quickly during RX.
Is this fan EMI primarily conducted, or radiated? I kind of assumed it was conducted, and that maybe some DC ground isolation would fix it, but I don't really have a good way to do that (though I know there are specialized ICs for this). Would probably be way simpler to forego the fan and switch over to the aluminum block!
Thanks for humoring me,