Understanding roast profile

[emilesilvis]

Hi all. I very recently started roasting on the Nano 7 (I've roasted about 10 batches so far). Loving it so far.

I noticed something interesting the last time I roasted. I had the following sequence of roasts, all on the Raost v4 profile:

Logs saved on NANO7 2021-11-02 22-06-36.png (27.4 KiB) Viewed 4954 times

What confuses me is that the time for the roast at 1.2 (5:52) is more than the time for the roast at 1.3 (5:24). I would have expected 1.2 to be shorter than 1.3?

Did the sequence of roasting influence it? Or does it have to do with end temperatures? Can someone help me understand a bit better?

I also attach the logs:

log0012.klog is 1.2:

log0012.klog

(62.64 KiB) Downloaded 272 times

log0010.klog is 1.3:

log0010.klog

(50.75 KiB) Downloaded 261 times

[nsolling]

Hey.

I can’t pull up the graph as I am on my iPad.

Is it the same bean you are roasting and is it the same load? I.e is one of them 120 and the other 100g?

But you are correct: if these parameters are the same the lighter roast should be quicker.

[fnq]

Hi emilesilvis
Taste them first, if they both taste good then it is time to happily drink coffee!!!

If you think the shorter roast (but supposed to be higher roast level) is not as good as the other then

a) possibly juggle your roast batch size etc as suggested by nsolling

or if I was to analyze your log ( and i am a novice here so likely to be wrong)

b) your errant roast rises above the profile line from about the same point as zone boost one ( 4.35 ) and the PID is disabled so the roast just gets too exothermic and reaches set temp too quickly
you could increase the roast batch size as above or
( more of a guess)
put a negative boost parameter in boost zone one ( -1 in the zone 1 boost profile setting). BUT this may impact how the other roast follows its profile line.

(NOTE AGAIN the author of the profile put the zone boost in for particular effect.... so changing this may have other unexpected effects)

Because everyone's power supply/ bean selection / temp/ humidity etc is different sometimes profiles need a tweak to accommodate these differences.

If i wanted to try this roast again i would first just up the roast batch size a tad and see if that tames the heat. Cheers Darryl

[emilesilvis]

Hey.

I can’t pull up the graph as I am on my iPad.

Is it the same bean you are roasting and is it the same load? I.e is one of them 120 and the other 100g?

But you are correct: if these parameters are the same the lighter roast should be quicker.

Exactly the same green bean and exactly the same load (100g).

[emilesilvis]

Hi emilesilvis
Taste them first, if they both taste good then it is time to happily drink coffee!!!

If you think the shorter roast (but supposed to be higher roast level) is not as good as the other then

a) possibly juggle your roast batch size etc as suggested by nsolling

or if I was to analyze your log ( and i am a novice here so likely to be wrong)

b) your errant roast rises above the profile line from about the same point as zone boost one ( 4.35 ) and the PID is disabled so the roast just gets too exothermic and reaches set temp too quickly
you could increase the roast batch size as above or
( more of a guess)
put a negative boost parameter in boost zone one ( -1 in the zone 1 boost profile setting). BUT this may impact how the other roast follows its profile line.

(NOTE AGAIN the author of the profile put the zone boost in for particular effect.... so changing this may have other unexpected effects)

Because everyone's power supply/ bean selection / temp/ humidity etc is different sometimes profiles need a tweak to accommodate these differences.

If i wanted to try this roast again i would first just up the roast batch size a tad and see if that tames the heat. Cheers Darryl

They both taste good, but I still expect the lower roast level to be shorter than the higher roast level, especially since the green beans and loads were exactly the same.

[nsolling]

Hey,

I had a look at the logs you sent.
The curves looks a bit strange for 1.3 that it is not following the power curve very well.
Is this something you are able to replicate or was it a one off?

Did you turn off power of the machine when doing back-to-back roasting? (you should leave it on)

I have once had a similar weird experience where a bean got stuck on the temperature sensor - So maybe something similar happened.

[Wayne]

Hi emilesilvis

Thank you for this.

I am looking at the logs and they are actually tracking fine. I would make an adjustment, up the preheat power by minimum 100 to give you more momentum through the 3 minute mark.

It looks like the coffee in log 10 is responding to the fan profile just before 5 minutes and pitching higher. The recommended and temperature is reaches quicker than in log 12. Thats why it stopped 30 seconds before. It's hitting the set threshold line. Threshold line is sitting at 0.333.



Log 12 roast finishes at 216.8 degrees at level 1.2 at 5:54
Log 10 roast finished at 219.1degrees at level 1.3 at 5:24

The levels are based on degree temperature at the recommended end point and takes the threshold line into account. If the threshold is reached, the level will adjust itself to the closest level. The level 1 (215) and level 2 (222) temperature distance is quite large. This means that the temperature scale during adjustment is varying. Bring this difference down and the levels change will be more manageable.



If I look at the logs, log 12 reached the threshold line before reaching the actual recommended end level. Log 10 also hits the threshold but in a shorter time to log 12. However, due to the profiles level setting between level 1 and 2, the 2.3 degree increase pushed the level in the next recommended level.

Increasing the threshold degree will help align these or experimenting with the "roast end by time ratio" will align this.

I hope this helps.

Regards

[emilesilvis]

Hey,

I had a look at the logs you sent.
The curves looks a bit strange for 1.3 that it is not following the power curve very well.
Is this something you are able to replicate or was it a one off?

Did you turn off power of the machine when doing back-to-back roasting? (you should leave it on)

I have once had a similar weird experience where a bean got stuck on the temperature sensor - So maybe something similar happened.

I haven't tried to reproduce it. Might be once-off fluke indeed.

I didn't turn the machine off.

[emilesilvis]

Hi emilesilvis

Thank you for this.

I am looking at the logs and they are actually tracking fine. I would make an adjustment, up the preheat power by minimum 100 to give you more momentum through the 3 minute mark.

It looks like the coffee in log 10 is responding to the fan profile just before 5 minutes and pitching higher. The recommended and temperature is reaches quicker than in log 12. Thats why it stopped 30 seconds before. It's hitting the set threshold line. Threshold line is sitting at 0.333.



Log 12 roast finishes at 216.8 degrees at level 1.2 at 5:54
Log 10 roast finished at 219.1degrees at level 1.3 at 5:24

The levels are based on degree temperature at the recommended end point and takes the threshold line into account. If the threshold is reached, the level will adjust itself to the closest level. The level 1 (215) and level 2 (222) temperature distance is quite large. This means that the temperature scale during adjustment is varying. Bring this difference down and the levels change will be more manageable.



If I look at the logs, log 12 reached the threshold line before reaching the actual recommended end level. Log 10 also hits the threshold but in a shorter time to log 12. However, due to the profiles level setting between level 1 and 2, the 2.3 degree increase pushed the level in the next recommended level.

Increasing the threshold degree will help align these or experimenting with the "roast end by time ratio" will align this.

I hope this helps.

Regards

Thanks for your detailed reply Wayne! I had to read this four or five times before getting the gist of it, but I get it (I'm also confident this will become more straightforward as I get into actually using Kaffelogic Studio to design profiles — I know there's a few good resources available for that, so looking forward to dig into that).