How to achieve stable decrease of ROR with only 3 power settings while reaching the right drop temps?

Hello dear roasters. A while ago, I asked some question, where I got pointed at the CoffeeMind guide on how to roast coffee. I tried to follow the rule, where the guy says we should be able to achieve these, with only 3 power settings:

  1. Get the FC at 9th minute
  2. Get the development last 3 minutes.
  3. Drop at temperature between FC and SC to achieve agtron 75 (if I remember correctly).

Now I am roasting small batches (350g) before I feel more confident, but I struggle a bit with getting a good ROR curve with only 3 power settings.

This is my last roast, which is not at all bad (any feedback is welcome, esp. your thoughts on how dark you think it is).

Now as you can see, at the very end, I failed to keep the 3 power settings only rule, because the ROR started to go up. At the same time, I am failing to reach temperatures for the Agtron 75 (assuming it would be around IBTS 215C) in the “right” time.

If I think about it, those rules (in my case) seem to go a bit against each other. As you see, I got my FC at 7:58, while I was aiming for 9:00. At the same time I have a hard time to reach IBTS 215C in time (3 min) after FC without a growing ROR. It all just doesn’t add up.

Any ideas? Thanks a lot!

Where did you hear about the 3 power settings only rule?

I know that is common for large cast iron roasting configurations, but the dynamics of single layer stainless 1Kg small thermal mass induction electric may be too sensitive for that.

Start with Andrew COE’s roast, and if that is too light, @bab has a good looking roast on his profile if you want to take it a little deeper.

The three power settings (power drop slightly before FC, and power drop at some point during FC) come from Morten Munchow’s introductory CoffeeMind material for the Bullet.

His approach challenges somewhat the “conventional wisdom”, if you want to call it that, espoused by many, that a smoothly declining ROR is the key to a good roast.

@zakotar.t63F, I’d recommend first focusing on roasting with your senses and judging mainly what you taste in the cup. If you like what you’re roasting, it doesn’t matter what your curve looks like. You can find award winning roasts whose profiles look incredibly janky if judged only on ROR prettiness.

1 Like

Thanks a lot for this information. My aim was to nail the “starting point” so that I can calibrate my “taste” as well and take it from there. Otherwise, if I just experiment randomly, I fear I’d fail to associate taste to curves and be lost in the chaos. I will however be a bit more daring going forward.

I’d be happy for any references to useful material on how curves (or else) affect the taste. Thanks

If you want to follow the approach of Morten Münchow then perhaps you should consider watching this other video of his where he argues that ROR is isn’t a good reference point for optimizing flavour in a roast:


zatokar.t63F Bear in mind that the 3 power settings thing is just an exercise to get you acquainted with how your roaster works. Munchow never said that it would produce good coffee or a ‘good’ RoR curve or that you should actually roast that way. Being restricted to a minimal number of power changes makes you focus on placing those changes where they work the best and seeing their effect on getting to the goal. Getting 1C at 8:00 minutes instead of 9:00 minutes means you made the first change a bit late. Adjust for that. Then you can work on where to put the second change to get smoothly into and out of 1C under enough control to reach your end goal. Then you will know what you can accomplish with minimal changes. The RoR curve has nothing to do with this exercise. Also, the Munchow exercise keeps drum and fan constant so you focus only on power settings and know what you can accomplish with just that. Then move on. The RoR thing is something else. As nicolas mentioned, Munchow has his own opinion on that as well as the Development Time Ratio. Cheers.


Thanks so much. Good thinking! I kind of expected I could meet all the “goals” at once, including a good RoR.

Perhaps someday, and then you can teach all of us!