So back in my hottub heatpump experimenting days my biggest problem was by far and wide defrost. I was running my 8,000 btu window a/c with the condenser in nice 104* water, and my evaporator in anywhere between 50* and 15*F Temps. I even ran it a few times while it was snowing.
I tried the Clixon, and a couple other methods, but none were reliable with the simplistic methodolgy I was using and the wide temperature range.
Something that I consistently noticed was that airflow dropped when the coil frosted up, so it got me thinking, there's a fool proof way to determine when the coil is no longer moving air and pulling in heat.
My idea, a flap in the exiting air with a flap and switch, when coil frosted over, air flow reduced, flap lowered, defrost initiated. Obviously a flap isn't ideal, but it pushed me to my next idea, a pressure differential switch. No frost = little differential, frost = more suction between coil and blower wheel. So I Googled it, and I'm not the first guy to try this.
Heat Pump Defrost Cycle Information
He had extraordinary success with his design, and it's exactly what I was scheming. This is the direction I plan to take with my defrost control.
Seeing all the threads regarding different defrost methods and problems, I figured this may be of interest to many here.
Happy Hacking,
Adam