Coffee

My Coffee Stuff

Coffee Links

Equipment

Beans

Zines

Other

Nov 6, 2003: Since I wrote the following article, I have been experimenting with a stovetop popcorn popper and a heat gun for roasting. As much as I admire the level of control I can achieve with the variac-ed FR+, it requires too much time to roast a useful amount of coffee. As there is already plenty of literature out there on the popper, I will probably write up some stuff on heat gun roasting next.

Some Roasting Info

Some say there are two approaches to roasting, a technical approach and an artistic approach. I've found that I lean more toward the artistic side of the spectrum. What I mean by that is I try to achieve a good roast by being in tune with my equipment and the beans. (Be the ball, Danny)

Conversely, the technical approach involves things like carefully documenting various quantities like mass of beans to be roasted, temperature at various stages, roaster or voltage settings and timing. I'm not saying there's anything wrong with this approach. I just find that trying to keep track of the data ruins my enjoyment of roasting. It also makes it harder for me to achieve a good roast. I tried keeping track of times, temperatures and voltage settings for about three roasting sessions. After letting the roaster slip into the cooling cycle prematurely because I was busy looking at the thermometer, I realized this approach wasn't for me.

In short, my roasting philosophy is to first concentrate on roasting the beans as evenly as possible. I don't mean all the beans have to be the same, I mean the internal temperature of the beans should be uniform. All my fiddling around is aimed at that goal. Profiling for me is deciding how fast the different parts of the roast should be within the parameters of having evenly roasted beans.

I have a general procedure I use that I tweak a bit for different beans. I use a Fresh Roast Plus with a KRM 5 amp variac (overfused) and a Marshall thermometer. The only other "mod" I use is to prop one side of the roaster up with a paint stirring stick to change the air circulation. My basic procedure is ...

• I've found that smaller batches are more controllable and surprisingly roast a bit slower. Therefore instead of using two rounded scoops of beans, I use two slightly less than full scoops. The exact amount I use varies by which bean I'm using and I tend to adjust it a bit from batch to batch.
• I usually start the roast at around 115 volts. There are two reasons for using a relatively high temperature to start. The fan speed varies with the heat when you have one variac controlling the whole thing. You need good air circulation to get the beans moving. The other reason is to drive the water out of the beans to make them lighter. This also helps to keep them moving.
• The next bit is hard to describe because the exact order of everything varies by how the roast is going. As the beans get moving, I start moving the variac down, eventually reaching 105 volts. Also, I give the beans a couple of 30 second rests by stopping the roaster. The first rest starts anywhere from 45 to 90 seconds into the roast. The second rest usually happens a minute or so later. The idea is to get the beans as even as possible before the first crack happens. The rest plus the lower heat seems to help.
• Once the beans look even, if the first crack hasn't started, Ill kick the voltage back up to 115 for 10 seconds, then 110 for 10 seconds then back to 105. If I think things are proceeding okay, I'll try backing the voltage down to 100 or even 95.
• Depending on what I'm going for, I'll either finish the roast at this low voltage, or more likely kick it up to drive the second crack.
• I've added a refinement here. I have been regularly bringing the heat up to drive the second crack. Once that is going, I stop the roaster for 45 seconds or so and let the beans roast each other. This seems give more flavor. After 45 seconds I do my cooling routine.
• When I'm ready to stop the roast I'll manually move the knob to cool and crank the voltage up to 130. I'll keep the cool going until my thermometer reads less than 175. Then I cool the beans the rest of the way in a mesh colander.
• The thermometer is there as a guide. I don't kid myself that I'm measuring bean temperature but it does give me a sense of what's going on in the chamber. I take more of my cues from the sound, smell and appearance of the beans. When I'm going for a Full City roast, the visual cue is not so much the color but the sheen. If the beans look satiny the roast is just right. If the beans look glossy, it definitely time to stop the roast.
• Smell is difficult to describe. You really have to learn the various smells. Prior to the first crack, the smell is often described as grassy. It doesn't smell anything like grass to me (either kind) but I can't think of a better descriptive term. From just before to just after the first crack, the smell is desribed as bready. Again, it doesn't really remind me of bread but I don't have a better word. After that you can start picking up caramelly notes to the smell. In theory you can know when to stop the roast by the smell. Mainly I use the smell as a sanity check on my eyes and ears. The color of the beans can be deceptive. If you are going for a lighter roast and don't want to hit the second crack, then you can't really use sound as a good cue.