Uneven color and uncracked beans

[utkuyzc]

Hey everyone, this is Utku from Turkey. New to the community and like to learn on the way.

So I just started roasting and trying to get a better understanding by tweaking some numbers.

I Bought Kenya AA fresh beans and roasted them with 1500-2000 RTD profile. It didn't follow the line I played with it; here are the results.

I attached all the logs but a summary;

In

v1.0.klog

(91.09 KiB) Downloaded 977 times

, it didn't follow the curve and developed at 31.1%

In

v1.1.klog

(82.55 KiB) Downloaded 1196 times

, I pulled the last dot of the profile curve, so it developed at 24.9%i but didn't follow the curve, and the color was uneven

In

v1.2.klog

(81.64 KiB) Downloaded 1304 times

, I increased preheat power to 890 (recommended one)

In

v1.3.klog

(82.13 KiB) Downloaded 1160 times

, I dropped the zone 2 boost to 1 and zone 3 to 0, it didn't follow the curve, but this might be the best result I got,

In

v1.4.klog

(83.31 KiB) Downloaded 1069 times

, I dropped the zone 2 boost to -1.5 and followed the curve perfectly, but developed at 18.4%, and the color was uneven,

In

v1.5.klog

(83.13 KiB) Downloaded 914 times

, pulled the last dot of the profile a bit further so it can develop over 20, increased the first two dots of the fan curve from 1470 to 1480, a little bit drifted away from the curve, developed at 22.2% and not only color was a problem, but also there were uncracked beans.

One more detail, just before the v1.5, I went into fan mode and pressed some buttons(!), and immediately I decided not to play with it; I turned everything back to its original values, but maybe I broke something.

So just when I was expecting the best outcome, it was the worst. Can anyone give me some useful tips?

[SeanBrews]

I have seen this pop up, and after troubleshooting and testing I have come to the conclusion that this is caused by the coffee itself. You can see that the roaster is reducing the power on your power line to compensate for it and try to bring the coffee temp back onto the bean curve. Something to try would be to create a boost zone at the start where it begins to stray and change the boost value to -10, -20 or -30 (this will act as a %, leaving it on 0 will hold the power at the value of when it enters the zone. Changing it to -10 will reduce it by 10% of the power and -30 will basically act as if you are cutting the gas on a production roaster.) You can try this to see if you have any better results.

The algorithm stopped itself from dropping the power too far in order to prevent any form of stalling; it thought it would perform better if it had left it to rise.

[utkuyzc]

Thanks for the reply, mate. I'll try the same profile with the fresher and quality beans I'm sure of.

[SeanBrews]

If you are going to be using different beans, including different freshness then I would suggest roasting the profile as is to see then changing it using the 'gas off' simulation. Change one variable at a time. Hope you come right!

[fnq]

A little late here but will add my comment just in case.

I have found Power reduction through the development phase is going to result in a muted roast.

Sometimes even flat lining the power curve ( often due to power supply in the house itself) can spoil the roast

[utkuyzc]

Hey fnq,

Thanks for the reply; I'm also keen on hearing your suggestion. How would you correct it? The power in the house is pretty stable, and I don't use an extension cord, just in case. As MrBrews suggested, it's due to the beans themselves because they were old and probably not going to have the same issue with fresher green beans. But as a rookie, I'm open to hearing as much information as possible.

[fnq]

Hi utku yzc

A couple of things

(My opinions and offerings.... so not gospel)

The default MASL profiles are generally aimed for a more medium roast level as default.

I have found the further away from the recommended level ( ie 3.2 to 2.4) , there are more chances of your roast going off profile.

For medium light or less - I would suggest trying the RAost profiles ( user contributed profiles subsection) and maybe look at one of the versions that reads true for you. ( Maybe aiming for an end temp between 222-224 depending on your take on those curves)

Note Benjamin mostly uses 80 gram load sizes. Good luck

[utkuyzc]

Thank you so much! I'll definitely give them a try and i think i gotta complete the training sessions first

[utkuyzc]

Okay, a little update on here,

I couldn't give a chance to MrBrew's suggestion as I'm still a rookie trying to get the hang of the KL. But he guessed it right because new, fresher beans worked out great.

I also tried other profiles, Raost, Hoo's, Firestarter, and NordicLight, they worked out great, and the results were even and looked good; still waiting to taste them, but I believe I'll be fine.