How do you troubleshoot poor roasts?


Hello all,

I’m a newbie to roasting, and have been trying to roast beans from Sweet Maria’s. I’ve obtained a good handful already, and tend to start my roasts from the Ikawa profiles, and then move points around until the beans visually and audibly have the desired outcome (with regards to first crack). I sometimes take notes on when the various stages happen (Pale, Yellowing, etc) also, though I have a very limited understanding of how they impact the final product.

The next day when I finally taste the coffee, I may get something I like, or I may get something that’s awful. In either case, I honestly don’t know which stage of the roast to start tweaking to change the taste of the coffee into something desirable.

I completely understand there’a trial-and-error aspect to things, but I’m also interested to know if there are rules of thumb people have developed from their experience that we might be able to learn from?


What have I done wrong?

Hello @Kvangels,
This is THE question, actually. And it very much depends on the level of understanding you have (and plan to have ;))
My approach is to test different profiles I have at hand (all extended so that I can roast to the point I want - usually some 30s after end of 1st crack - and asses from tasting which profile shape I like. I usually do few different lengths too. I dont manipulate the shape of the curve as I dont feel like I have enough knowledge yet to do so, but I plan to get there.
If you want to understand and manipulate the curves read what others that are doing that write about it (deven for example) and read through the stuff in fruit for though thread, as there is a lot to help with understanding.
But the experience still comes from practice… :wink:


How do you stop the roast? By pressing the start button? Or twisting the coffee loader knob? Or something else?


I would say each bean is different (and represents different flavors and cup styles) but keeping great notes of the roast and personal inputs, is important to speed up seeing the similarities and what beans like what. Adjusting a roast is easy if you know what the bean needs. I’m still trying to understand this roaster and how a bean handles heat over time.
Breaking up the segments to represent the different stages is something I do. But I don’t know how to adjust certain segments. For example, when roasting the last of my Columbian El Hato, one of the roasts I got such a sweet candied fruit, right towards the end of the 4th segment on the nose and I kept trying to adjust and move the profile but I couldn’t find it again. Other times I roasted a bean and tasted it after and it was fruited and juicy and then 20minutes later the fruit was gone. Maybe I didn’t cool the beans fast enough or slow enough there are just a lot a things to take into consideration when roasting. That is to say; roasting to beat the best roasters out there.
At the point of learning to make your own profiles the worst thing you can do is add multiple varieties of beans. Buy a lot of one bean, get varieties sure but get a lot of quantity, with proper storage beans can last years and age. That’s to say about not being worried about buying too much of one good bean. I wish I bought 10lbs more of the Ethiopian Sidama from SW because after 3.5lbs I found the profile and enjoyed blueberries for 2 months without getting tired of it. Roast to get the flavors on the notes, even the attempt of profiling with a goal is giving a purpose to the roast. With a Sumatra I found a flavor not listed on the bag but it was great. With the El Hato I just roasted the last 1-2lbs and was coming up on some great flavors but could not get it in the cup. So…5lbs gone with not really any great results.
The more info you read the better your understanding of how to roast will be. There is plenty I’ve already written about on this forum, there is a library or resources etc etc etc. But the question you are asking is the question we are all trying to figure out. So go out, take notes and share something if you find anything interesting. GL!

@mshextra yes we press the button, which starts the cool down cycle. There is no way to eject the beans without a cooldown cycle. I would suggest keeping the cool down 1-2 minutes and cool the bean all the way down on a table or different set up because I feel the roaster doesn’t cool down the bean enough and it is a bit weird to be cooling down beans in the same chamber as you heat them.


Yes the start button stops the roast a makes it go to cooling phase.



So I’ve been thinking about this question for the past couple of days.

Missing the mark on a roast is disappointing. Usually I start by working my way backwards. Starting with the coffee in the cup: what color is it? How does it smell…is it baked, or is there no distinct smell?

Answer: if it’s too red, roast longer but not necessarily higher temperature. If it smells like baking spices (cinnamon, cloves, allspice, etc.) Then there was most likely a stalling from the start of yellow to the 1C. With a cup that doesn’t have a clear aroma, increase the speed around 1C depending where the speed can give clearer flavors in the cup but beware going too high a temp or too fast can result in ashy/carbon-y flavors.
Sometimes those negative flavors (baked or ashy/roasty/carbony) will dissipate or rest out. You will notice that during the 1-10days after that the flavor of the coffee changes.

Is the coffee dry(ing) on the tongue?
too long and too hot at the end of the roast

How to get a juicy cup?
An even steady ramp with a slower speed than the previous stage, but a faster ramp than the next stage, I think, helps create this.

How to get fruited/floral?
This the goal for me, what I believe to the the actual flavor of coffee. Also I believe this stems the “origin” flavor ie the amalgamation of indistinct flavors that come from that region of bean. Like how I taste in all French wines a certain note of cheese that, when I drink them, I know the wine is from France.
Getting the greens to a bright green and extending that green stage without turning them yellow until you want them too be. Turning them yellow cuts the maturation/formation of fruit. How accurately you can stay in the green zone and how long you stay there determines florals --> candy --> fruits.This a guess But to affirm my guess try getting in the green area and smelling the crazy aroma that comes out. How to keep the aroma going and carry the momentum through the yellow and crack/dev is still unknown.

Chocolate I feel is the creation between a med-long extension of green while also extending the point of dark brown to 1C, This phase I like to call expanse beacuse the beans get a little bit bigger but do not puff like popcorn and get smooth I think the smoothness is from a high enough external temperature searing the outside of the bean .

Getting strong undefined acidic flavors, I feel is a waste. Those acidic flavors can be developed into real, discernible, flavors by allowing the beans to spend more time during the green to yellow aka drying.

The speed of the roast is important, too fast and nothing develops, too slow and things die off. Too hot and things get ashy and too low a temp and not enough momentum/speed/rate of rise, things don’t crack.

Hope this makes sense.

I also forgot to mention, read and read. I’ve read a lot of stuff but at the time they don’t make much sense, but after I’ve spent some time on the roaster and re-read things, they make a lot more sense as I get more familiar when roasting.

Also don’t forget to check out Scott Rao and other’s blogs…Barista Hustle et al. If you want to become a better flamingo you must think like flamingo, stand like a flamingo and maybe, one day, you will fly like a flamingo.


Here are some examples modifications I’ve made to Andino profiles that have been successful for various beans.
My method has developed with time and is bound to change, but it goes kinda like this:
The bean tastes underdeveloped with the Andino profile, but pretty good and not too far off. I usually raise the first point the most, raise the second less, and the third the least. I might try raising pt1 20, pt2 15, and pt3 10. I also may try to move pt4 back a bit. I roast it, wait a few days, brew it, and then see.
I adjust again, and end up somewhere good usually in part due to luck.



I see you’ve somewhat flattened the fan profile on these. What were you trying to accomplish? Thanks.


I needed to just to get the beans to rotate, other than that stayed with the Andino.
Note: when I raise the temp, I usually need to shorten the roast. Surprisingly this can work. I believe it is because the “momentum” of the chemical chain reaction is a more potent factor than time. Even though I will get through each roast phase faster, it seems to accomplish more than spending a longer time in a phase with mild change. It is good to have a strongly pronounced phase, I believe, and the shorter roasts like the ones above are what has done it for me.
Also, I’ve lowered and/or shortened the end points on some of these since posting. Btw, Is there an automatic way to share new profiles between different Apple devices? Would be nice.