- If you don't want to blow chaff around, you could make a screen wire box
to fit over the end of your dryer vent. If it were something like a big
Teaball clam shell, you could open it to vacuum it out every once in a
while. If it were large enough, you wouldn't have to vacuum it every time
you roasted a few grams.
If there were just one exact position for a thermocouple to read correctly,
every other sensor position would be incorrect.
I assume every position gives incorrect temperature sensing. That saves
looking for perfection. What will drive you to insanity is trying to take
readings with a sensor that moves within the hot zone.
Hot gas flow is highly stratified. The gas molecules move past One Heating
Point, one of the heating coils, and several cooling points- the beans that
are being heated.
For roast-to-roast uniformity, the atmospheric pressure (gas molecule
density), the Number and Size of the beans in the roast chamber, and gas
velocity within the roaster would have to be constant, or compensated for,
among others. Variation kills roasting uniformity.
Want to see a demonstration of non-uniformity?
Fill a canning kettle with water and heat it to boiling. If you stick a
thermocouple probe in it, you'll find it's thermal Disneyland while it's
heating up. Even at the boiling point there'll be temperature variation.
Scene 2 is dry, with high velocity gas passing a red hot wire in an instant,
then releasing its heat to a mass of (much cooler) roasting coffee beans.
Anything that slows the gas flow causes it to spend more time at The Wire,
and pick up more heat to dump in the bean mass.
It's the Infernal Dance of the Coffee Bean.
Cheers -RayO, aka Opa!
"The indisputable truth is that no coffee is fresh if it isn't fresh
roasted." - - Martin Diedrich