Hey Guys/Gals,

I've fallen into some bad roasting habits and am trying to up my game this year and actually develop some good profiles. Like a lot of others I've subscribed to the Rao declining Rate of Rise theory for my 1st six months roasting on my IR-2.5. After a bit of traveling/break from roasting and sampling other Roaster's coffees over the Winter I've realized that using this theory along with Diedrich's suggested airflow settings (50/50 at 270 and thru the drum at 370) has resulted in me 'baking' a lot of my beans.

I'd really like to find a better approach to roasting as I now realize I got extremely lazy and just tried to have a roast with a declining ROR and 20% 'development' time while not really paying attention to anything else.

I'm running my roaster on propane in my garage until I find a permanent home for it and just blow the chaff directly out the cyclone onto my driveway. With no venting I quickly realized that the cyclone's power could really quickly slow down a roast, but for some stupid reason I keep going back to close to the Diedrich suggested settings thinking there must be some reason to follow them.

I played around a lot with the Airflow when I first got the roaster, but got myself into some bad habits of roasting everything the same way. I've attached my 'old style' of roasting as the background in this Artisan shot, but it's hard to see the settings so here's a breakdown:

Guatemalan Atitlan 4.75lbs
Charge: 440ish with gas at the lowest it goes 3WC
1 minute into roast: Gas on full power 11WC
300 degrees: 50/50 airflow
350ish degrees: Gas down to 8WC
370ish degrees: Gas down to 6WC
Airflow switched to 'through the drum' sometime during 1st crack to slow it down even more and generally drop around 400 degrees. I feel I've been going into 1st crack without any real momentum and then sucking all the air out of the drum I stall out my roasts and get that baked/papery/cardboard taste.


I just have the stock Diedrich thermocouple hooked up to Phidgets and it seems my Bean Temps are a bit low. I'd ideally like to get an ET thermocouple...any suggestions for that?

I did some searching on the forum and came across a post from @bohemianroaster who is using a IR3...his thoughts make a lot of sense to me and it seems like he has a similar issue with his thermocouple giving low readings (he estimates it's off by 20 degrees....thread here: Tips for a newbie using a Diedrich IR-7N??) He never uses airflow 'through the drum' unless roasting really dark and "As for airflow, I go CB until 360 or so, depending on the roast level, and then to 50/50 for the remainder. I only go to RD at 420 when I'm roasting dark. Diedrichs roast more with conduction, and adding airflow tends to both strip moisture from the beans, and to cool the roast. Too much airflow can actually stall the roast, giving you baked flavors."

He provided this basic profile that I attempted to follow on my last roast and it had a lot more acidity and overall was a lot better cup...from his previous comment I don't think he meant to switch the Airflow to 50/50 until 360 so I waited until then to switch to 50/50 and never went 'through the drum'.

Drop 415F, gas 8"
Bottom usually around 195F @ 1:45
280F 5:00 to 5:30, gas 10", 50/50
360F 8:30 - 9:00, gas 8"
375 FC starts about 9:30
390 or so . . . back off to 6"
400 FC ends about 11:00
Finish roast at 420 to 424, 12:30 to 13:00


I definitely didn't have the speed to progress through the roast as fast as he does, but the results are really inspiring and have me really wanting to improve my skills and knowledge of how this roaster operates!

What are your thoughts on using this approach with my IR2.5? I was concerned that not going 50/50 and 'through the drum' with airflow at Diedrich's specified levels would result in a ton of chaff, but that doesn't seem to be the case? Still appears to be the same amount of chaff in my cyclone base post-roast. Do you think I'm creating possible problems by using an airflow strategy like this?

I NEED to get a lot better at creating profiles and going into a roast with a good plan! What do you recommend I work on? What should I try next to improve upon this roast?


Thanks for taking the time to read and I'd greatly appreciate any suggestions that can be provided.

I don't have an IR-2.5 (though I've taught classes on them at events) but I do roast on both an IR-1 and an IR-12 and use the same approach with both of these, just being cautious that if I'm trying to develop a roasting plan on the IR-1 for use on the IR-12 that I'm not straying into a rate of change at portions of the roast where those rates aren't attainable on the larger machine for the equivalent portion of the roast. see also:

http://dailycoffeenews.com/2017/03/27/m ... -machines/

Starting with airflow and chaff. Chaff comes off of the coffee fairly early in the roast. If you watch through the window at the front of the drum you'll be able to observe that and the easiest time to get rid of chaff is to do that fairly early in the roast. For reference, on my roasters, I hit the color change from green to yellow at 300F and will generally do an airflow adjustment to RB between 250 and 270 strictly for chaff removal while the coffee is still basically green. That's more effective on the larger roaster than it is on the smaller one, so you might need to adjust that for your own machine, but the sooner you flush that out, the less potential impact that airflow adjustment has on the finished cup.

For airflow adjustments during the roast, I've found two airflow patterns that I really like. One is like what you describe. CB to start, 50/50 at yellow, but I do the RB adjustment much earlier at the transition to brown (330F on my machines). The other is to just start out on RB and leave it there through the whole roast. From a production standpoint that's less efficient because you can't multi-task cooling one batch with roasting the next one, but some coffees are improved by that enough that it's worth making that trade off. Should be less of an issue in a home roasting environment, so it's something to consider. If you try other airflow manipulations, you really don't need to worry about that causing harm to your equipment.

My view on DTR is well known. Rather than rehash all of that I'll just say you'd do better to think in terms of absolute duration within significant temperature ranges as that gives you a more powerful framework to reason about changes to roasting plans and there's no reason to believe that doing that particular division operation makes any sense in the context of roasting chemistry. That's just noise and probably the major culprit in what you're interpreting as baking. Similarly, the shape of the rate is a reasonable starting point, but it's not universal. The solution to being lazy with your roasts is to relax your constraints, try different approaches, and see what you like. There's no substitute for doing the work. See also: http://dailycoffeenews.com/2016/10/10/u ... -analysis/

For air probes, I find that they're less useful if you've got real time rate of change tracking, but if you want it, a good place for that is to take off the chute between the hopper and the drum and run that in on the side so that it's under the door mechanism, then takes a 90 degree turn into the drum far enough in that it's not impacting the fins. Make sure you've cleaned all the metal shavings off well before reattaching that.

Example time, here's a rough profile that I'm using for a coffee from Burundi that I think is similar to what you're trying to achieve:

0:00: 365F, CB (start)
3:41: 250F, RB (chaff removal adjustment)
4:22: 270F, CB (end of chaff removal adjustment)
5:34: 300F, 50/50 (yellow)
6:49: 330F RB (brown)
9:20: 380F (start of first crack)
11:18: 410F (end, about 20 degrees cooler than start of 2nd crack)

Rate of change toward the end of that is still relatively high (a bit under 7F per 30s) to preserve acidity on that coffee. You can try it if you'd like, but you might want to try some basic exercises instead:

1) Progressive roast. Your trier is big enough that you can use that to pull out samples during the roast to cup. Take a roast into a nice rolling 1st crack and pull about a dozen or so samples from there through to a dark roast. Cup the samples. That's a good exercise to try when working with any new coffee to get a sense of where you want to put your roast level for that particular coffee and is the best way to eliminate bias toward or away from any particular roast level. If you're not sure what I mean by this, I demonstrate it in a video:
2) Manipulating time in ranges. What happens if you speed up or slow down different parts of the roast? Rob Hoos' book goes into this in detail and my own data agrees with everything in that, though I think you'll get more out of it if you run the experiments yourself and make your own observations with your own coffee on your own machine. There's a lot of that also in this video series which uses Diedrich roasters:

http://www.youtube.com/playlist?list=PL ... yEoTZAg4oI

Most of these reach conclusions that are more toward medium or dark roasts, but as an interesting aside, one of the top finishers in the recent U.S. Roaster Championship competition used a control technique described in the 4th episode of that series with his Diedrich roaster on two relatively light but well developed coffees.

As you're trying different approaches, think less in terms of "this is better than that" and more in terms of "this change in roast produced these particular changes in these attributes of the finished coffee". Over time you'll find that a change that you might not have liked with one coffee is exactly what you want with a different one.

Thanks for taking the time to write up this great reply Neal, I really appreciate it!

I definitely see the chaff start coming off around 240deg. I went to the Diedrich training a few years ago and Steve was big into going full air right then until you couldn't see anymore through the window....also had a coffee importer there helping train us and he said it wasn't a big deal and that he did it for awhile because Steve said too, but he didn't notice a difference waiting for it to come off later in the roast. I did it for a long time after attending the training when I was using a bigger Diedrich, but since I got the 2.5 at home I stopped as I've been trying to retain as much heat as possible to get to 1st crack at a decent time. What effects have you seen if you don't blow it off in that 250-270 range?

I'm really curious why you switch your airflow to "through the drum" at 330 on some roasts and keep it there for all of some others? (I'm not sure what RB stands for but assuming it means airflow through the drum?) Using a 4.75lb batch and running full gas the entire roast I don't think I'd ever hit 1st crack if I went RB for a whole roast and I'm pretty sure if I did it at 330 I'd struggle to hit 1st crack before 13-14mins. I assume this has to do with my cyclone having no resistance from venting? Ive wondered in the past if there was a good way to mimic venting by reducing the power of the cyclone with a dimmer, but really don't want to add that into the mix. I've also read about and experimented with using middle increments on the airflow lever (only going halfway between 50/50 and RB) so full air isn't so powerful at the end of a roast.

What batch sizes and gas settings are you using in your example profile? Running 85% of capacity batch sizes I've kind of stuck close to the Diedrich recommended charge temp of 420 and sometimes up to 450...should I be trying to get down to 365 or is that just for batches that are maybe 50% capacity or less?

I'm roasting tomorrow and will try out your recommended progressive roast exercise...that was actually one of the 1st YouTube videos I ever watched on roasting a few years ago so I'm excited to rewatch it now that I've got a few hundred roasts under my belt! I also just read that Roaster Manifesto book by Hoos' last week and while a bit over my head I think I get the basic concepts of experimenting with different range times.

I'd love to hear if you have any other suggestions for exercises I can do or other profiles I could experiment with. I just feel like I've gotten myself stuck and wasted a lot of time focusing on declining ROR & 20% DT. Do you have any general good starting points for DT's and ROR's for before and into 1st crack....you say 7F per 30s is pretty high, so what's more standard? I feel like I've been stalling out a lot of roasts and heading into 1st crack really lacking energy to get a rolling crack. I've got some other coffees I want to bring the best out of and lots to learn. For instance, what's a good starting profile for a natural Ethiopian vs a washed Ethiopian? What are some good resources for learning more about profiling different beans?


Thanks again for all the help!

On the chaff issue, chaff has a bitter taste to it if it doesn't get removed. If you can have that removed by the roaster itself instead of having to rely on more manual methods later, that's easier, but how much this matters is really entirely up to you.

The airflow adjustments are done to balance heat transfer efficiency with bean development. The through the roasting bin (RB) setting reduces pressure inside the drum leading to a lower heat transfer efficiency, but encourages better physical development of the coffee. My installation seems to be better than average when talking to other Diedrich owners (our installer was really paranoid about not screwing up his first roaster installation) and it seems like I have an easier time pushing my profiles around where I want them and running larger batch sizes. To a certain extent everybody needs to sort out the capabilities of their specific installation, though the way to do that is by trying these different control manipulations and seeing what happens. You might be pleasantly surprised by what the machine is really capable of doing.

If I were doing the example profile on my IR-1 that would be with a 2 pound batch out of 2.2 pound maximum rated capacity (so about 91%) and for the IR-12 I'd need to be a bit more conservative with that. I'm usually roasting that closer to 30-40% capacity for production. If I needed to go higher on the IR-12, I'd push the starting temperature up higher, but my work on profile translation analysis suggests that we're talking about the portion of the roast that's least significant to the finished cup. It looks like you're using Artisan in your screen shot, which is fine, but as far as I know they haven't added any features for incorporating this sort of analysis into a live display, which would make it harder for you to take advantage of the extra flexibility that affords in the early roast (I could be entirely wrong about that and would be glad to learn that this feature from Typica made it into yet another program). Gas adjustments are as needed to stay on track and will vary depending on my batch size and how much roasting I've already done that day.

Hi. I'm not using a diedrich but 4kg gas drum roaster with infinite control on drum speed airflow and gas

I think a key thing to look for in your airflow is the colour of the beans at about 142 c. When they show the colour change. brighter yellows mean a drier bean at the end and it doesn't get that moisture back Conserve the oils and flavour up to this point. Then up the airflow to first crack (I use about half capacity) and 3/4 after first crack to end to get smoke out and control and cool the roast from first to second.

Here's what I do

Drop in beans low air setting and low gas setting on a preheated roaster about 180 c at air port
At 1.5 minutes gas to full air to 1/3 of your max
At 5.30 air to 1/2 of your max gas stays on full
At first crack gas to idle air to 3/4

It doesn't stall as you have got the heat in early by turning on the gas and getting the heat into the bean without blowing it and drying it over the surface with to much air. This allows you to turn off the gas to idle at first crack while air is turned to high cooling it down and glide your way to roast end as there is a lot of momentum and heat in the bean from not using to much air up to about 5.5 minutes. If you use to much air early in the roast it might stall using this technique because the inside of the bean is dried out and can't hold much heat anymore. You should get a nicely pleasant lively espresso from it as the oils and acids haven't been destroyed with to much air to early. Obviously you do need a quantity of air in the beginning to move chaff but a lot of roasting machines don't have enough air control for this technique. And many machines are never the same so what I have said may not work or needs to be tried in terms of General principals

Hope this is useful in general terms
Always interesting to hear what other roasters are up to.