One tricky aspect of this is that fluid beds and other highly convective roaster designs make it hard to get a clean reading of BT by their design, and this grows worse as you turn up the heat and air. The more convective heat you are pushing through the beans, the more your probes may be reading air temp and not bean surface temp.

When I used a HG/BM I regularly measured FC-start in the low 390s, but I strongly suspect that this was at least partly an artifact of the high ETs that are needed to drive a roaster with that design. Switching to a conventional drum roaster, FC-start dropped down into a range between 370 and 385F. ETs are also much lower in this setup, which draws more of its heat input from conduction & radiation, so I suspect this is closer to a "true" value.

The reason I bring this up is that, to go through ramp faster in a convective setup, you'll probably bump up ET a fair amount, and thus automatically increase your measured FC-start temp due to measurement error alone, which makes it harder to say whether the increase is a real effect of the profile change or not. Conversely, dropping it will also drop your ET and thus your measurement error.

I haven't tried to measure this specific thing in isolation from everything else, so all this is just speculation. I can try to do a side-by-side comparison next time I fire up the roaster, and if so I'll post my results.

[creative nickname] wrote:One tricky aspect of this is that fluid beds and other highly convective roaster designs make it hard to get a clean reading of BT by their design, and this grows worse as you turn up the heat and air. The more convective heat you are pushing through the beans, the more your probes may be reading air temp and not bean surface temp.

I'd have to show a picture, but my BT probe only protrudes into the roasting chamber a 1/2" or so. The beans only loft from a column the center and the beans around the probe just roll around. Not much ET air hits the beans around the probe.

That said, I can actually change the BT reading to whatever I want. If I stick it in more, I get more ET influenced readings. At a certain point pulling it out, I don't get any and the readout stabilizes. In other words, I can get a 1C at 400 or 370 depending on where I put the probe. The main thing for me is to be consistent. I calibrate it so I start to get that bread smell around 300*.

If anything I think fluid beds give a more accurate reading of BT, since the probe is immersed in a bean mass that is barely moving. And even in a drum, there is still a lot more convective heat than conductive. At least from what I've read, beans are roasted by hot air from the drum as much or not more than contact with the drum itself. So the probe would be exposed to hot air more with beans rolling quickly around a drum, then in a fluid bed scenario where its sitting in a relatively stable bean mass.

Almico wrote:I'd have to show a picture, but my BT probe only protrudes in the roasting chamber a 1/2" or so. The beans only loft from a column the center and the beans around the probe just roll around. Not much ET air hits the beans around the probe.

That's what I would have said about my old setup as well. I still noticed a significant drop when moving to a drum roaster, with its lower ET for a given profile. Anyhow I'm not trying to make any strong claims either way, just wanted to raise a potential reason why it might be risky to draw strong inferences from measured differences in BT if you are varying ET significantly, as the former data stream is often biased towards the latter value.

I just put a Joper BSR5 profiler on my Xmas list. I better be very, very good.

My ETs tend to run 50-75* above BTs throughout the profile.



I had been getting into the habit of shorter 10-12 minute roasts, but I'm remembering why the chocolate bomb coffees taste so good from 13-14.

Interestingly, the Artisan evaluation algorythm still rates this as OK/Acidic in the development phase!

@Alminco thanks for kicking off this thread.. I think you have the closest setup to mine, a home built air roaster, and am interested in comparing notes. I have had the BT reads high criticism in the past and my answer is to just switch off fan and heat anywhere (usually at the end) and wait for BT to stabilise - there can be no air effect then and I find BT drops a bit less than 1 degree C which I put down to the inside of the bean catching up with the outside. Try it yourself with something you don't mind baking. My 1C is almost always way above the drum guys and you also - typically 207C with very little variation (less than +-1degC). I have done quite a bit of experimentation with RC design and air flow and always if I am getting a low 1C it is an indication of a bean flow problem - some beans are staying in the ET airflow too long and getting overheated so getting up to 1C before the bean mass is up there. I always thought there must be a higher amplitude microprofile in other roasters as there does not seem to be much variation in 1C between different beans (apart from decaf and MM lower and Robusta higher). Interestingly I was getting exactly the same 1C with my introductory popper rig. All measurement with TC4 & typeKs.

That's an excellent suggestion. I'll throw 2 pounds of coffee in the roaster get it up to about 1C and shut the airflow off and see what the temperature does. Of course it will cool overtime. But if the ET airflow is making a significant contribution to the reading, it should drop off immediately.