Last Arabica and Robusta roasts with result of baked / fried coffee :-(


my last 5-6 roasts had the result that i had a hard baked taste. Not drinkable. Can you pls give me some hints what I can do to get coffee without this baked taste.

This was one of the last roasts:

  1. PArchment Robusta: Roast World - Cup, grade, and analyze your coffee roasts in depth
  2. Red Lepar Arabica:
    Roast World - Cup, grade, and analyze your coffee roasts in depth
  • Can you see somethin in the roast curve ?
  • At which timeframe the 1st crack should be ?
  • Do you focus on BT or IBTS when you are roasting ? I have the feeling that BT and IBTS is diverging with 15-30 degrees … and I cleaned the IBTS ebfore the last roast …
  • Do you have a basic roast for me which is running with the most beans (i like 350-500g roasts) that I have a good basis to do finetuning …

Currently I am very disappointed because I am burning much money with this baked Issue …

Need help :slight_smile:
Thanks a lot


THerefore I have

I am no expert… but if it were me… I would drop the drum speed to 6, and roast 500 grams. Maybe drop the PH temp to just over 400. IMHO!

Per advice from Morten Munchow, I generally shoot for 9 mins to FC and 11-12 mins to drop. You’re very close to that, but you are roasting very much lighter than my preference. I generally drop at 215°C to 218°C, which (I think) is a shade darker than the so called Nordic roast but nowhere near dark. SC is in the low 220’s°C.

My “standard” profile is
500g green
PH 250°C
P8 D9 F2 —> 4 mins
P5 D9 F2 —> 200°C
P3 D9 F2 —> End

I roast indoors, so ambient is usually close to 21°C.

Clearly, your mileage may vary!

I’m shooting for more body, less fruity/winey/citrusy.

I think my next batches I’m going to experiment with 10 mins to FC and 14 mins to End.

I always go with IBTS. Bean Temp is affected by so many things (charge size, bean size, ambient etc) that I’m not sure it really means anything and is perhaps only useful for hitting the same place with identical roasts, but IBTS does that just as well.

In the spirit of full disclosure, I have not been crippled by good taste!


Hello Steffen,

First thing I noticed, the charge temp. For 600g Brazilian beans I suggest to be charged at 180 degrees C. For me I focus on BT. The roast is baked because it had took longer than what it should during the milard phase (between yellowing point and first crack). I may suggest the following steps for you :

Preheat 260

Charge 180

Charge P0, D5, F2
After 1 minute P9
Bean Temp 150 F4, P7
Bean Temp 185 F6
Bean Temp 195 P4
Bean Temp 212 End Roast

Good luck

just my two cents, I actually feel like the charge temp is low. Baked usually means the roast has lost too much energy somewhere leading to a baking effect (which seems to be in the development phase in your case looking at your graph). For my 500g I use 473F for a darker roast but since you have 600g Brazil you can adjust it slightly lower maybe 464/455. I use about 20-30 seconds P0 for soaking which you can also adjust to ease into yellowing, then use P7, D8, F2 at the beginning, then it just depends on what marker timings you want and you adjust the rest of the settings later on. your Development phase just seems to crash in ROR and taking around 2:30 is kinda long. Hope this will help.

Thanks a lot for your feedback, I will try and let you know. I let aillio test my logs and everything seems to be ok with IBTS etc. Issue is the roaster in front of the bullet :slight_smile:

It looks like you’re using far too much fan for temperature control. In your second Robusta roast, for example, you hammer the fan super hard around FC (because I think you’re getting nervous about the RoR curves trending upwards). Once you get to F4 and above you’re really adding a lot of cool air to the roaster which is causing your crashes in your RoR during the post-FC phase of your roasts. Even F3 for me, when I’m roasting in a cool environment, will bring in too much cool air and crash a roast.

People generally say it’s those crashes that cause baked flavors.

I roast 1LB/454g at a time and have had the best results by just keeping the fan at 2 the entire time while modulating with power level changes. (I’ve also had good results where I would increase from F2 to F3 at yellowing but I’m trying to keep things simpler).

Here’s a natural Ethiopian I just roasted that came out well taste wise.

I roasted this one in an ambient temperature of about 14C.

1 Like

Hi @hercdeisel I tried your Roast World - Cup, grade, and analyze your coffee roasts in depth
natural ethiopian profile you mentioned in your last message.
Today i tried it. It is completely sour (and a taste of baked).
Do you think it could be the problem that I adjusted the power to often because I wanted to reproduce your roast profile.

I appreciate any feedback and hints.

The Santos I tried is sour as mentioned and seems to be more a very light roast.


Overall it looks like you’re putting a lot less power/heat into your roast than I did. I’m usually not getting below p5 or p4 until after FC. Sometimes lately I’ve been mixing in a fan move from F2 to F3 as the first temperature decreasing move after FC while keeping the power level at 5. You’re down to P4, P3, P2 range 5min into your roast. I didn’t get under p5 until the 9min mark on my roast. I also started a bit hotter at 215 charge instead of 205.

On my machine, at least, roasting in the 400g range, F2 seems to be the peak heat transmitting fan speed. Roasting at F3 slows the roast down. That’s how I’ve landed on F2 and F3 being the right fan ranges for me.

I’ve just roasted a few naturals from El Salvador and with those I charged at 205 but immediately hit P8 & F2 at charge. Here’s an example. I think starting the bullet strong on power/heat rather than doing any kind of soak at the start is best. The pre-heating routine it’s keeps the power element pretty low level so there’s a natural ramp up time/soak to be the electric element coming up to temp.

My general plan for washed 1lb/454g roasts is to hit P9/F2 at the start, with a 215-220 charge. For naturals, I charge at 205 and hit p8/f2 at the start. Sometimes I’ll build a move to p9 at the BT 0 ROR point. All in all I’m generally trying to simplify and reduce the number of adjustments I’m making during a roast.

1 Like

Hi Gray (@gfh65wi )

thanks for your remarks. Today I tried your standard profile.
I have not tasted it:

But could it be that your US 110V machine has not so much energy like the 220V Bullet in Europe. You mentioned you are focussing on FC at 9min. With your standard profile i had the first crack at 6:41 min.

What do you think, why is this the case ? i think the Fc is too early ?

Thanks a lot.


Hi @alghamdi.majid1b1s
thanks for your feedback: i will try it.
what do you think how much time should be between yellowing point and first crack approx. to prevent this baked coffee issue ?


Hello Steffen,

Phase between YP and FC called Milard / Caramelization. In general I try to make it between 3:00 and 3:30 minutes. I prefer clarity, specially when I roast beans with fruity and floral aromas. The longer this phase takes, will allow more chemical reactions, it will increase the complexity. While shorter milard phase, allows less chemical reactions to occur before hitting the first crack, leads to a very clear notes.

1 Like

Hello @alghamdi.majid1b1s

thanks for your response …sounds good. I will try out next time and analyze my roasts re time in milard phase.

And at which time do you prefer 1st Crack ?
And how long are you roasting until which temp ?

thanks Steffen

180C charge for 600g → is this also the Bean temp you are focussing ?

Are you also focusing on bean temp when you are roasting smaller batches like 350g ? or are you then using IBTS ? Do you have a standard profile as basis for 350g maybe like you presented above ?

thanks a lot. Steffen

Hello Steffen,

FC mostly range between 192-202 based on the beans themselves. I keep myself close to the roaster once I reach 190 C, to make sure I hear the FC.

There 2 ways to decide how to end your roasts D-Temp and DTR.

D-Temp : is the end roast temperature (Drop) - First Crack temperature, lets say your roast achived the FC on 200 and your dropped the beans in cooling tray on 210, Your D-Temp = 210-200 = 10 Degrees. How to use it to decide when you end your roast!? Add the 10 degrees on the FC Temp. If the FC occurred on 195 and you want to end your roast with 12 degrees D-Temp, then you should end your roast at 207 Degrees. Based on the cupping results, can adjust the D-Temp either you increase or decrease it.

The second method is the most common, DTR (Development Time Ratio). Most of the recommendations says DTR advised to be between (15% and 20%) of the total roast time.

Maybe in the beginning you use one of them, but later when you advance in roasting you will use both.

For example, I try hit 10 D-Temp degrees in 15% DTR. If I hit desired D-Temp in less than 15% DTR, thats indicates my development time was faster than it should and I should slow it when I adjust the profile.

About the IBTS and 350 or less batches, unfortunately I didn’t roast less than 500g, but if I had to roast 350g or less, I will consider the IBTS. But for 500g and above I use the Bean Temp.

Hey Steffen,

Can’t say for sure, but I would imagine that, except for P9 perhaps, the actual power input should be the same for both 120V and 220V machines for a given power setting.

There are a lot of things that affect FC times such as ambient temp (my roaster is in the dining room so the ambient temp is almost always near 24°C); bean variety; and bean moisture content, all other things being equal.

Another factor that seems to be rarely considered is exactly how you define First Crack. Is it that first loud pop, or the second or third, or is it when the sound is almost continuous? For my purposes, I say First Crack happens when I hear several pops in quick succession. This seems to generally happen in the 198°C to 203°C (IBTS) range for most varieties under the conditions I roast.

6:41 does seem quick, but Morten Munchow claims the flavor of coffee is 80% final color; 15% development time (FC to end); and 5% time to FC, so your quick time to FC probably doesn’t really make much difference. (His work is based on considerable blind taste testing with both rained and untrained testers. To arrive at the above percentages, he did hundreds of three cup tastings where two of the cups were the and the third was different by some particular parameter and noted how many of the tasters could pick out the different cup.)

I’m a coffee Philistine that drinks mine with a sprinkle of sugar and a lot of cream, so take my views with as many grains of salt as you feel necessary!


jpndemir has deleted his post, but it did raise an interesting matter that might benefit from a bit of elaboration.

In addition to the statement that 80% of the coffee flavor that is attributable to the roast is final color and 20% is timing, Morten Munchow said that, of the timing part, 5% is time to yellow; 15% is time from yellow to FC; and 80% is time from FC to end. (I am relying on Mr Demir for the exact percentages, here, but I remember something quite similar from one of Munchow’s webinars I listened to a while back.)

These are, of course, percentages of the 20% that Morten attributes to timing. Thus, as percentages of the contribution to the final flavor profile, time to yellowing is 1%; time from yellowing to FC is 3%; and time from FC to end (“development time”) is 16%.

Because it is such a small percentage of the final flavor, I shoot for 9 mins to FC, per Morten’s suggestion for a starting point, but don’t sweat it very much if I miss it. I can’t imagine my palate is sufficiently discerning that I can actually notice.

I pay more attention to the timing between FC and end, because here is where the balance between fruit/acidity and body is apparently struck.

But mostly I pay attention to final color. This is something I know I can actually taste. Because I don’t have a color analyzer (and because I don’t have a need for absolute repeatability), I use final IBTS temp as a proxy for color, at least within the same bean variety.

Morten largely poohpoohs concerns about RoR except as a tool to hit roast milestones (FC, SC, end) at he time desired. He says his tasting panels have largely been unable to detect RoR differences.



I deleted my post as I noticed that I misremembered the Munchow teachings,

He did indeed say that 5% is start to FC, 15% is FC to end of roast and 80% is the color of the roast.

To oversimplify: As long as the drop color is the same all the time you don’t need to worry TOO much about how you got there.

This is me as well.

I wonder if Morten’s ever done an experiment where he out and out crashes the RoR into zero/negative territory, and checks if that sort of profile can be detected double blind.