From what I remember at Counter Culture, the roaster do batch after batch with little time for cooling. So I assume they use a very high drop temp but I am not sure. I wish I was back again to take a better look.

Wouldn't the turnaround temp. indicate how much heat is being sucked from the solid drum. Considering batch size and the drum temp. at drop. Seems more relevant for a solid drum vs. a convection driven roaster.

OK. I've read a magazine article about roasting with a solid drum roaster. Here are the main points.

Ideal roast time = 12-15 mins for a solid drum + convection heat roaster (fan). Over 20 minutes will give a baked flavor. Less than 8 minutes will enhance sour flavors.

Roast profiles are divided by beans densities.

Hard bean (high density) dark roast. High initial heat (90% in QuestM3=9 amps). Ramp very fast till 1.5 mins of 1C. Drop power here to 20% heat till 3 mins after 1C. Then end with 60% power. My opinion: on the Quest m3 roaster I guess you need 150 grams which may translate in about 1 cup because of the high weight of the beans. If you load up the roaster with more weight, you'll need more power at the end.

Medium hard bean. Drop temp is 30 degrees lower than for high beans. 80% (8 Amps) heat till 1 minute before 1C. Cut back power to 45% (4.5 Amps) for a slow 1C till the end. This should give a more linear profile. My opinion: here you can load 200 grams in the Quest M3 roaster.

Exotic beans (very low density). Roast light because they are delicate. Drop temp is 30% to 50% lower than hard beans. Low to moderate heat till 1C. Then sample continually. temperature rate of change no more than 1F per 10 seconds. Heat supply 15 to 40%%. My opinion. Here you can load up to 300 grams in the Quest M3 roaster because it requires the least change of rate.

Drop temperatures.

High bean density ~ 300F
Medium bean density ~270F
Very Low bean density ~ 150F 210F

Regards

That's one of the Boot articles in Roast magazine. They are excellent, but the numbers need to be translatd; they may work for his Probat, but not on the M3 (mine at least), unless you keep changing the airflow and run a very hot drum.

Rafael, I bought the same insulating paper you have, but I'm going to try something different -- a hood scoop for the air intake, so the air is drawn in along the hot skin of the roaster. Not sure it'll do any good, but it'll make an interesting experiment to see if it will speed the roast with lower drum temperatures.

Hi Jim. That's a good idea. Did you tried already putting the roaster inside a box? (maybe the box could pipe the top air back to the bottom)

Next weekend I may put a blanket of insulating material (non reflective) over the top and use an infrared thermometer to see where is the hottest area.

Cheers

Insulating the M3

I repeated Rafael's insulation procedure. Two layers of the paper fits beneath the bars of the stand and the cage. No fasteners are necessary when since the cage and stand secure the paper. There is one gotcha that Rafael didn't mention: let the roaster run at full heat for a half hour, since there's solvents in the paper that have to cook off.

Here's how it looks:



I did two fairly identical roasts, on before and one after insulating, of a no-name DP brazil I'm using for these basic experiments. I apologize for the double exposure in the graph, I overlaid them as two layers in gimp after resizing to equalize their scales.



The insulated roast ran a minute faster, but finished a tad lighter as well: the bean profiles were nigh identical. The drum temperature (blue lines) is 10C (18F) lower on the insulated roast, so insulating slightly increases the heat transfer efficiency of the roaster. The energy efficiency also goes up, and the insulated roast ran around 1.5 amps lower and about 1.5 airflow settings higher.

So insulating provides lots of energy for doing fast roasts with high amounts of beans, but this is probably a poor idea. The speed limit is set by keeping the drum temperature low enough to prevent ashy roasts. On the low grown Brazil, the 260C of the uninsulated roast exceeded that, and the roast tasted ashy. The 250C in the insulated roast was borderline (lots of tobacco, but no ash).

If you like the flavor of leisurely drum roasts, insulation is unnecessary. But if you want faster, brighter roasts, it's a useful tweak.

An air scoop, to blow in heated air rather than cool air, may not be so good an idea, since the cool air is needed to change temperatures quickly. I haven't tried it yet, since it would require a thin molded part mounted to the control box, and that's beyond my fabricating skills. At this point I don't think I'll go for it; since I now have the heat transfer efficiency I was looking for.

Wuau. You've been busy at full throttle

When I did my insulation I attributed the smell to the glue of the stainless steel tape I used. I think it took ~ 10 roasts to be gone. The stainless tape will make it easy to clean the surface of the insulation. I used only one layer of the fiber insulation so that everything fits under the horizontal arms. But I guess two layers work better than one.

I bought the materials to do the exhaust pipe. the four pieces are only $15 and are very easy to put together (flange, 4x3 reducer, flex bow, 3' long pipe. HomeDepot). With it, I am planning to roast inside and exhaust the smoke by the window (it is starting to get cold to have the window halfway open). It will shoot the the smoke farther out and will keep the roaster at room temperature. I'll do some pics on the weekend.

Cheers

Hi Jim,

A big fan of your roasting posts, and others on this forum. I'm a professional roaster, but I frequently turn to the home roasting community for thoughts and theories that are founded in science and not roaster-specific empirical data.

Anyhow...this is backing up a bit on this thread, sorry about that, but you posted your "desired" drum roaster profile, and I couldn't help but notice that it has a declining MET. Up to this point, you've appeared to be fervently against a dropping environment temperature. However, this profile seems to contradict your previous position. Is this the case, or am I missing something?

Cheers,
Phil

Phil & Sebastian Coffee Roasters

I may be eating crow on that one.

I've just been running a simulation that can do both air and drum heat transfer. From a math standpoint, the biggest difference is that in the beans in an air roaster will rise from 40 to 50 degrees per minute for every 100 degrees the environmental temperature is above the bean temperature. In a probat style the figure is between 20 and 30 degrees per minute per hundred degrees difference.

So what does this mean? Schenker in his roasting dissertation used an environmental temperature ramp running at around 325F for 4 minutes, rising to 450F in 4 minutes, and staying there to the end of the roast. This was for an air roaster or industrial high convection roaster which has similar properties. At 50 degrees per minute per hundred difference this ramp gives a very familiar profile: the slow dry period, fast ramp to the first and slowed down roast finish. I can confirm that this is very close to how my P1 works.



So how do we get the same profile in a drum that runs at 25 degrees per minute per hundred difference? You can see it on the graph, you need to drop the ET during the drying phase and roast finish, and raise it rapidly during the ramp to the first. For someone experienced in drum roasting (unlike me), I think, this simulated ET curve may look familiar.

I think the idea that dropping ETs are bad is based on fast heat transfer roasters (Carl Staub must use a lot of airflow, since he's the one who said this first). The reason he said it is bad is that the roast tastes flat due to flavenoid polymerization. It makes more sense to think this happens not by environmental temperature but based on what's going on inside the bean. I guess stalling the roast even slightly creates this problem. That would happen whenever the ET drop s in an air roaster, but not in a drum roaster.

This makes sense to me, but it's still needs to be confirmed.

another_jim wrote:I may be eating crow on that one.

Maybe not. I'm suspecting that on our baby drums, ET management is what's actually different. Just because your ET probe reads a certain temp doesn't always mean your beans are actually being subjected to that exact temp! A couple weeks ago, I posted my latest profile, and one reason behind that profile is ET management on our small drums. I now believe:

coffee.me from Max's Final Hottop Drum Roaster Profile... for 2009! wrote:you have an ET that's pretty tough to figure out AND heating elements that are hard to control: if you go high heat, you could get your ET too hot, if you go low, you lower your ET -- both result in less-than-great roasts.

So what's the solution? Either recirculate that hot air or stop that fan completely.