Very nice job Blackdog!
I just wanted to thank you for having posted all of this, it helped me solving an annoying issue with the dummy load I disegned.
To be honest, I designed quite a lot of those dummy loads, many of them in my working time, using a bunch of components I had lying around, just to test the supplies of the circuit I was debugging.
But with the last dummy load, which was intended to be a bench instrument (!), I had some problems! It was oscillating with some specific supplies I was testing, it seems like the control loops of the supply and the control loop of the dummy load started to "talk each other"! Quite easy to see if we keep in mind the behavior of the output impedance of both supply and load Vs. frequency.
Testing it with a battery or a classic bridge + cap supply gave excellent results, but an LM317 supply didn't want to collaborate with my load!
I solved the thing just beefing up the compensation network (R6-C5 in the first schematic of the topic), giving the load a slower response... that's it!
Also, I suggest anyone who built a dummy load like this to
verify the actual waveform of the current in it, even in DC! The multimeter, in fact, gave a correct reading in DC even if the current was oscillating and jumping like a crazy horse!