Hi folks.

I've been doing some roasts today and I want to share my little experiment to calculate the moisture of the exhaust air while roasting (proportional to the current bean moisture).

I have probed the exhaust of the roaster with this:


(water container with rag- the thermocouple probe is placed inside the wet cloth inside the exhaust. The other thermocouple probe is also inside but not touching the wet rag )

It puts moisture on the lower part and I put a thermocouple probe there (inside where it is wet). Then I place another probe in the top where it is dry.

The principle is this:

If the air is wet, the two temperatures will be closer (since the bottom one is very wet). If the air is dry, the two temperatures will be apart.

So as you can see the gap becomes larger as the roast progresses (beans lose moisture and they become drier). Exhaust fan was kept constant at value 6

Since it is digital (dual thermocouple meter), it can be recorded



orange = dry exhaust probe
black = wet (bottom) exhaust probe

Note: I use a heat gun and BT reads higher than what it should (after dry end).

The software is not ready to record differences (like orange to black) but that is an easy fix. Since this was only a test, I think it came out OK.

Explanation of the collection of the 4 curves. There were two recordings. One was made with the PID (ET and BT) computer, and the other was made with my Omega Dual thermocouple meter using a laptop. Then I loaded the humidity graph as a background (it can be moved up and down) and then I took a pic of the graph with the 4 curves.

Cheers

I have pretty high humidity in my roaster. The excess leaves through my chaff canister vents with the light, mostly recirculating convection the turbo oven provides. I'll be interested in the levels you record. I've always considered too much new dry airflow a negative but with nothing to really prove it.

Rafael, you are a mad scientist I love reading about your experiments.

What on earth are you going to use this data for?

Hi Dieter.

I always was under the impression that reading the moisture level during the roast would be a practical roasting metric. It is just that I don't know how to use it

Next weekend I'll play some more, and I'll try to see if I can find any practicality.

Cheers

Although I understand that knowing the bean moisture content would certainly be interesting, I'm afraid this measurement method has some basic flaws.

A wet bulb reading isn't valid much over 100 C at normal atmospheric pressures. Not only is the relative humidity super small at these temps, your feeding liquid for the wet bulb is likely to be cold enough in comparison to be involved in cooling of the thermometer and/or slight condensation of the incoming steam. I would also like to point out that even under more normal temperatures the wet bulb/dry bulb comparison requires huge air speeds. (up to 60 mph!) I like a lot of airflow, but...

After thinking about this a little while longer, (and feeling bad for criticizing) I wonder if it wouldn't be possible to get at this using a scale.

The Quest weighs in under 25lbs, scales with 5g resolution are not that uncommon in that weight range. (under 15kg) The trick would be to find one at the right resolution/range with a big enough platform, and some type of usable output. Not that it would be easy or cheap, but likely some interesting data!

Hi again.

I did three test roasts today. This time I did not use a heat gun so that the graphs don't get polluted. After graph 1, I moved the top thermocouple a little higher, so maybe the numbers of 1 are different. The temperatures are actual temperatures (two black lines). The airflow was kept at #7. The weight was ~ 1 cup of green beans. At this time I don't have yet a way to graph the difference. No power adjustments were made except were you see a little yellow square. In the third graph, the timer of the roaster ended and the PID turned off for 1 second. So what you see is what I got.

Graph 1



Graph 2



Graph 3



Cheers

PS. Now that I see the BT temps, I think I have the wrong thermocouple type selected on the PID as they look higher than what they should be.