Halla - *&%$#)@ - lujah! It works!

 I've been struggling with managing the slope of the Bean Temp curve (data from the thermocouple) after 1C. The slope of the line would just tank after 1C got going. Nothing I did seemed to adequately affect that plunge and that threatened to stall the roast.

I had looked at a bunch of roast data (most of the 350 roasts I've done plus roasts from other people) and I saw varying degrees of the same problem (at least I see it as a problem). I had tried several things involving power changes and varying fan speed but they just didn't seem to have the controlled effect I was missing.

Then I recalled that I had been working with slow drum speeds at the start of the roast, gradually increasing drum speed from D4 to D8 just before 1C. When I thought about what slow vs fast drum speed meant to the beans, I realized that at slower drum speed the beans are staying in contact with the drum for a longer period. Yes, they drop away sooner because of less centrifugal force but the total time against the drum is longer which increases heat flow into the beans. At higher rotational speeds the beans are thrown against the drum a little longer but the total time against the drum is still less and that also means less heat transfer.

So if the slope of Bean Temp is flattening too much and raising power or dropping fan speed is inadequate to significantly control that slope, why not try lowering drum speed? That ought to increase the time the beans spend held against the heat source (the drum) and increase the heat moving from the drum into the beans.

Works like a charm! There are still big swings in temp as 1C continues, but when I visualize bounds around those extremes I can see they contain a trend that is not collapsing but is actually walking toward my target of getting to just under 10F°/min by the time I dump the beans.

Excellent.