In the past few days my roasts of SM's Sulawesi Toraja Sapan Minanga have tasted totally nasty (brewed) - astringent bitters, thin body, no sweetness, tingly. Just gross. Up till now, this has been like a candy coffee, totally sweet, raisiny, with a thick body. The only things that have changed that I am aware of: the weather has been the warmest and most humid it has been so far this year. The other thing I have noticed is that when I first began using my variac'd P2, I had a set minute by minute profile that I followed by constantly correcting for too much or too little voltage. Over the last week, I have focused more on getting a steady continuous rise in temp/voltage - no back-pedaling. But with this approach, I have noticed that the slower and steadier I go, the less controlled the roast seems to be. My second cracks, which used to consistently begin around 450 now occur in the 430s (1C still is stable at 400). It seems more like the second crack occurs at a given time, not temp. The roasts also seem to be darker looking than normal.

The only other factor that I have noticed is that the roast chamber is becoming quite dark. Is there any chance that this discoloration (oils?) is burning the beans as is heats up? While I am on this topic, how exactly do I go about cleaning this?

Any ideas on which of these factors (if any) are responsible for my now nasty roasts?

The second crack happens based on both time and temperature. A fast roast may get up to 460 before the 2nd starts, while a slow roast may start at 425 to 430.

My guess is that you are dropping the supply temperature at the end of the roast to get the slower ramp. This will kill the flavor. Dropping the ET to lengthen the final ramp may work in a drum (although I doubt it, they can turn off the heat for a while and still get a rising drum temp), but it's a disaster for an air roaster. I've roasted with a P1 since 2003. Trust me, do your heat control using the supply temperature at the bottom of the chamber, making sure it never drops, and use the bean temperature at the top to monitor and end the roast.

On the P1, the supply temperature graph should have the same time markers as the bean profile, but about 30F to 50F higher. 50F when ramping fast, 30F when ramping slow (the bigger the difference between ET and BT, the faster the bean temperature rises). Two tweaks: When changing from a fast ramp to a slow ramp, don't drop the ET, just let the BT close the distance. Start the roast by getting the ET to around 250F ASAP and then take it to 350F by the end of the drying phase, 3 to 4 minutes.

The difference between my ET readings (the probe is at the bottom of the chamber - just going through one of the fins) and my BT readings (probe at the top of where the bean mass rises) is never as large as 30-50F once I am out of the drying phase. The difference is typically between 10-20F max, sometimes just a few degrees difference. Is this problematic? In terms of temp, I do not ever drop the ET below my BT readings, but say for example if my ET is 450 and my BT 430 and my roast target is 435, I will notch the ET down closer to 445 or so to slow the increase. Is this the sort of drop in supply temp to which you are referring? Perhaps then I need to adjust my load (I am using 1/3 cup right now)?

Not a problem, ET temperatures can be all over the place, since even a little closer to the heat source makes for a lot of temperature rise. I measure my ET below the roast chamber, where it's hotter. Your measurement is slightly better, since its the hottest temperature the beans actually see.

In an airroaster, the ET should never drop. Not drop below the BT, but never drop at all, ever, period. If you drop the ET after the first crack, even if it stays above the BT, the bean's surface flavor chemicals will polymerize, and you will get flat, bitter coffee. If you want to slow down the roast after the first crack, you will need to flatten out the ET gradually as the beans get above 360 - 375. This is far more important than having a long gap between the end of the first crack and the end of the roast.

Geez, I dont know if I have read that anywhere before, if I did, it went right over my head.

Is there any way you could, in semi-layman's terms, explain what polymerization is, as it pertains to coffee. I see that chemical polymerization is "a process of reacting monomer molecules together in a chemical reaction to form three-dimensional networks or polymer chains", according to the omniscient wiki, but this is fairly meaningless without an applied example, which we have here. Does it stop the Maillard reactions? (I just hate knowing the reason something happens, but not understanding it - makes me look real bright when I explain to my wife why her coffee sucks and she asks me to explain the scientific term I used, and I just look at her hoping my facial expression is not as vacant as my mind is.)

Thank you very much for this quite essential piece of info!!

Just roasted a batch with absolutely no voltage reduction, only increase. The result after grinding out of the roaster w/ 30 minutes rest: a hot cup of candy and a very happy Noah!! Thanks Jim.

see, this is why I love this forum! I can learn so much from you guys!

Tim

another_jim wrote:In an airroaster, the ET should never drop. Not drop below the BT, but never drop at all, ever, period. If you drop the ET after the first crack, even if it stays above the BT, the bean's surface flavor chemicals will polymerize, and you will get flat, bitter coffee.

I believe you, but can you explain why this is so? Is a certain amount of energy transfer rate needed to keep the polymerization from happening? Even if that is so, why wouldn't any amount over that be OK?

Best,
David

I once thought it was true of all roasters; but I've had terrific heat gun roasts, where the ET obviously drops, so this is not true. I also doubt that there's some sort of magic going on here. In other words, I doubt that if the ET drops a tiny bit, the beans will suddenly metastasize into something horrid. My recommendation is as a policy for everyday air roasting rather than a piece of physics.

That being said, I keep trying to control my roaster on a bean temperature PID, trying to tune it so that the ET never drops. But there are always some roasts where the ET goes flat and slightly wobbly at the end of the roast. Moreover, this is more likely to happen on lighter roasts, where the effect is more dramatic. Cupping these roasts side by side with roasts where the final ET slightly rises is always no contest; the cup quality of the slightly rising ET spanks the wobbly one. So despite a "full-auto" bean PID control being much easier to use, I keep returning to running my roasts on an ET based controller.

Why is this so? I'm as puzzled as you are. If you follow a single bean in an air roaster, it circulates around the roast chamber, bottom to top and back again. In this circulation, it gets exposed to the maximum temperature at the bottom of the chamber, then travels up into air roughly the same temperature as the bean's surface, then back down. It dos this maybe one to three times a second (I've never timed the circulation rates on air roasters). So as the bean circulates, it is repeatedly exposed to falling and rising temperatures. A bean in a drum roaster experiences a similar temperature fall and rise as it moves from the lower surface of the drum towards the center and back again.

So as long as all these temperatures stay above that of the bean, it shouldn't matter that the hottest part gets cooler; the bean is always moving through hotter and colder in any case. But in an air roaster (or at least my P1) the cupping evidence is clear, at least to me: it makes a big and consistent difference.

noah wrote:Just roasted a batch with absolutely no voltage reduction, only increase......

One point worth note here is the relationship of voltage to MET temp in a popper. (I'm using a poppery 1 here) When raising the variac to a higher voltage, there is a fairly fast response in the MET temp probe as the heater element puts out more power quickly, then the MET rise slows down but it does not stop rising for several minutes. I suspect this is mainly an effect of the thermal mass in the heater housing and chamber taking on heat, so your poppery II may not show this effect nearly as much.

What this means for me is that voltage levels on the Variac must gradually drop in order to maintain a constant MET (ie; during drying or a light roast finish). A typical profile for me would have voltage gradually droping for the first 2 minutes of drying (to maintain 325F) then ramping to first crack requires progressive higher voltage. When first crack gets underway I generally need to drop voltage gradually a few volts to maintain a constant to slightly rising MET during finish. ....So while MET never drops through the entire roast, voltage to the heater does drop. (I hesitate to suggest this because you need a reliable MET reading to do it) The closer MET is to BT, the more critical that it doesn't drop. My profiles have the closest BT-MET only during final finish.

A couple other things I have found with MET readings on the Popper; there is variation depending on where you measure around the bottom of the chamber. Some are cooler and some areas warmer. The nichrome element is not entirely even in distributing the heat. This could account for alot of the differences noted in BT-MET margins.