A thermocouple works by connecting two dissimilar metal wires at the "hot
junction". The resistance/current changes dependent upon the temperature at
the junction. A multimeter can read these changes and deduce the
corresponding temperature based on an internal algorithm. It is excellent
for getting a quick reading (compared to a thermometer)--but when air
roasting, you can't get a true bean temperature because the hot air racing
by is ALWAYS hotter than the bean when roasting (otherwise, the bean roast
would stall). You cannot get a true bean temperature, but towards then end
of the roast when the rate of temp increase slows, you can get pretty close.
When home roasting, it isn't accuracy that is important as precision.
What's the difference? Think of a bullseye target. If you shot 4 arrows,
and they all landed on the outer ring, but north, south, east and west, you
would be "accurate" because the average would be near the exact center. But
you would not be precise. Imagine the four arrows all tightly bunched, on
the east side outer ring. Now you are very precise, but not accurate. You
can work with precision because in time you will learn to interpret the real
bean temperature based on sound, crack, etc.
In my FR, I mount my thermocouple just below the chaff collector in the
middle of the tube. First crack comes at 400-405 and usually ends by 420.
Some slow roasts seem to hit first crack at 395 though. Second crack
usually comes at 450+, but the Peru I roasted last night hit second at 444.
Guess it isn't that high an elevation bean. These correspond pretty well
with Tom's pictorial, indicating that my true bean temperature at EOR is
within a few degrees of what my TC is reading. The results are repeatable
for each bean, so I have great precision and good accuracy at the end, but
early in the roast when my TC reads 350, I might be 20 to 30 degrees off the
bean surface temp.
The key is putting the TC in the same spot everytime. If I lower my TC into
the beans, I get a hotter reading because I am closer to the heating element
and the temperature gradiant between bean and air is higher. But that would
be okay if I did that EVERY time, because then I would know to expect FC at
410 or 415 or higher, not 400 or 405. I prefer to allow maximum heat
transfer from air to bean and then read to corresponding air temperature as
it will be closer to bean temperature.
With your popper, I would lower the TC to 1.5 to 2 inches above the beans
and makes notes of when FC and SC occur. Lowering them into the beans won't
give you a true bean temperature as the air is way hotter and your readout
is going to jump as the tip jumps from bean surface to air to bean to air,
Hope this helps.