(price included the espresso sampler)
The I-Roast is my first roaster. Given that I had never roasted before, I wanted a roaster that would lead me by the hand to a good roast. The temperature control and computer-controlled temperature ramps of the I-Roast seemed perfect for me. As far as I could tell, the I-Roast was the only machine offering such capabilities, so I took the plunge and ordered it from SweetMarias. They included their "espresso sampler" of green beans.
My first roast was not good! I aborted the roast too early, and the beans came-out under-roasted. It was clearly operator error.
My second roast was much better. The beans became nicely darkened and oily, and evenly roasted. Stinky smoke filled my garage, making me very happy that I didn't try roasting in the kitchen under the fumehood.
My next roasting experience was miserable. It was a cold day here in Northern California, perhaps 50 degrees in the garage. The beans were simply not getting hot enough. I watched helplessly as the timer counted-down to the end, while the temperature refused to rise. The machine is supposed to regulate temperature via air-flow. For higher temperatures it reduces the flow, and for lower temperatures it increases the flow. It seemed to me that the machine could have made a more valiant effort to reduce the airflow to achieve the proper temperature. Perhaps it's simply under-powered? For my next roast, my garage was still cold, but this time I tried blocking the exhaust to slow-down the airflow to achieve higher temperatures. It worked! I got the temperatures I wanted, and therefore the roast I wanted. I was very glad to machine had a built-in thermometer, which I monitored all the way through the roast. Smoke again filled my garage. In case you've never roasted before, let me tell you: the smoke is stinky and icky and it lingers. I did not want my garage to start smelling like this stuff permamently. So I cut a hole in the garage wall and installed a dryer vent, and, using the adapter they provided, attached a venting duct to the machine. Fantastic! Most of the smoke went outside, and the restrictive airflow of the duct gave me the hotter temperatures I needed for a good roast. Unfortunately, near the end of the roast the machine created a lot of smoke, and it leaked everywhere. Don't count on a stink-proof roast even with the vent duct attached.
I roasted another batch, again using the vent attachment. Disaster! The weight of the vent duct easily twisted-off the chaff lid, and chaff FLEW everywhere. The beans cooled way down with the sudden increase in airflow, and chaff continued to fly. I tried to put the lid back on, but it was flesh-searingly hot. I finally did get it back on, but not without losing my fingertips to the heat, and not without losing the roast to the catastrophic cool-down. So I am very careful now to arrange the duct so that it is least likely to twist-off the lid. The lid should be re-designed to click into place more firmly.
When summer comes and temperatures climb into the 90's in the garage, I hope that I don't over-roast the beans, especially with the vent duct attached causing higher temperatures. I do not have a high confidence in the automated temperature control of the machine. Perhaps I can simply program-in lower temperatures. Again, I am very glad the machine has a built-in thermometer, which I can monitor to be sure the roast is not stalling, or running too hot. The temperature sensor, by the way, monitors the temperature of the air as it enters the pot, not the temperature of the air WITHIN the pot. The temperature reading is typically 30-50 degrees LOWER than the temperature set-point, which I find weird since... well, isn't it a feedback control system? Why the huge steady-state error? Perhaps the designers know that the temperature of the air within the pot is typically hotter than the temperature measured at the sensor (but that doesn't really make sense either...), so they apply a fudge-offset. Who knows.
Bottom line: this is a sturdily-built machine with a very useful built-in thermometer, and with some attempt at temperature control. In the best circumstances, the temperature actually does ramp the way the operator wants it to. However, the machine is very sensitive to ambient conditions (including whether a vent-duct is attached), which easily overwhelm the control algorithm. If there is one thing that could use improving, it's the temperature control algorithm. Perhaps it's a more difficult engineering problem than I realize.
Make no mistake: I am very happy with the output of the machine when the roast temperatures come-out right. The beans get roasted very evenly, and since I can see the beans so well during the roast I can stop the roast at just the right point to achieve just the right roast. My biggest issue is with the under-performing temperature control system, resulting in the machine's sensitivity to changes in ambient conditions.