So if your not finding an exact linear correlation what does this mean? Hint: might the beans have something to do with why this happening? What is this telling you that will be useful in monitoring a roast?

Side note, MET and ET can be a bit confusing. I use MET in 2 ways. First, it's the highest measurement within the main heat transfer envelope (that Jim mentioned). Second it's a maximum temperature within that envelope that I determine safe for the specific beans being roasted during the various points in the roast. So really the first is my ET and the second my ET safety limits or MET. This is more of a factor when batch size approaches limits with the desired speed of the roast and heat that must be applied to maintain that desired speed.

The correlations between different roasting temperatures are not direct but time shifted, with some temperatures showing more heat storage and time lags. On the Quest, measuring the air outside the drum is the most direct measurement with the least storage. The bean temperature is the slowest with the most storage. The drum temperature (after the first roast) will probably respond more quickly than the beans, but slower than the air, and stay higher than both. If so, it's not adding a lot of info.

On the other hand, it's a lot less of a bell and whistle than a rate of rise meter.

There's a distinction between the minimum number of control required to do the best possible roast, and the controls a given person likes to use.

Even an experienced hobbyist may not want to be a minimalist in their roasting controls but have a few bells and whistles on hand. That's a good thing; but one needs to distinguish the cool stuff from the necessary stuff. Some sort of source temperature indicator is necessary, while having both the drum and the air probably isn't. But necessary or not, it still may be nice to have it while roasting.

IMO, roasters should be sold with the necessary stuff, and we should get to add the cool stuff. I'm just a somewhat grouchy minimalist.

Some ET and drum temperature correlations. Mini 500, Yirg DP G3 from SweetMaria's. 300g. Drum temp measured by a Fluke IR thermometer.

Batch 1. First batch of the day. Charge temperature 20c higher.
Time/Temp/Drum/Heat/Vent
0:00/160c/246c/0/1
1:30/98c/209c/110/1 Turning point
6:50/150c/230c/160/3.5 Drying finishes
/188c/247c/50/3.5 Heat down in anticipation of first crack
10:50/191c/50/3.5 First crack
13:00/202c/243c Drop

Batch 2. Charge temp 140c, 20c lower than the first batch.
0:00/140c/180c/0/1
1:30/98.7c/173c/110/1 Turning point. Notice ET and drum temp are different compared to the first batch
7:42/150c/234c/160/3.5 Drying finishes. Notice similar temp compared to Batch 1.
/188c/250c/50/3.5 Heat down in anticipation of first crack
11:19/191c/50/3.5 First crack
13:19/203c/229c Drop. Notice the ET is similar to batch 1, but the drum temp difference

Batch 3. Charge temp 140c, but vent was minimally open to decrease convection.
0:00/140c/194c/0/1
1:30/88c/184c/110/1 Turning point. ET and drum temp compared to Batch 2.
7:42/150c/220c/160/3.5 Vent open for one minute, then minimally open to decrease convection
188c/246c/50/1 Heat down in anticipation of first crack
13:10/193c/50/3.5 First crack. Note it takes longer to reach first crack when convection is decreased. The temperature is also 2c higher. Vent open to decrease smoke particles.
15:25/206.9c/229c Drop

The three batches differed in time from start of first crack to drop by about 10-15 seconds (2:10, 2:00, and 2:15). The final roast color is similar, at least I cannot distinguish, but the bean volumes/expansions are significantly different.

I generally find it easier to concentrate on just one temperature, ET or BT. The drum temperature is too sporadic, as it differs by convection.

Hi

Nice info. It can get very tricky to measure metals with IR (unless painted). It will read the surroundings. Then the distance of the meter-drum and the size of the drum would need to be withing the conical range of the meter.

Cheers

I have an IR thermometer as well, but I can't imagine it providing highly useful data. Like Arpi said, if it were painted a nice non relfective dull or matte black, the drum would likely be easier to monitor and transfer heat energy more efficiently. But I'm not going to drill a hole into the side of my outer casing just to get a drum temp.

It's by far the hardest thing to control (for me) but of all three temps that I can monitor, only one can I directly affect, everything else is just off the downstream effect. I hate dialing in my MET to control the roast, but if I can't control that, then chasing BT or ET will only have me running in circles since they are controlled only via alterations of MET.

I wish it were easier for me. I bought my Quest back in September, I think... and it has a pre-drilled hole for ET monitoring. I got sick of having digital temps for my BT and MET and analogue for my ET, so I popped for one of Eric's matching thermocouples.

The little nuggets of gold that Jim leaves in these threads help me get better results. I veered away from KISS, and thought having more data would lead to better results, but I found that I can get mired down in data, rather than paying attention to the physical sensory characteristics of the roast. My greatest challenge is not blowing past a safe MET, and properly controlling it, without stalling the roast, yet pulling the maximum subtleties out of the particular beans.

I watched the folks at Four Barrel Coffee today, here in San Francisco. They have a 30lb Probat. Sure enough, those beans were more of a hornets nest than wet cement tumbling and sloshing around the bottom of the drum. I'd love it if there was an easy add on to speed up the drum on the Quest. Apart from that, the only solution is to go super low on the bean mass. I haven't tried less than 125 yet, but perhaps I should give 100g a try. At least if I mess up, I've ruined less $ on greens that way