Comment roast logs - 5 different beans using Steve's 9min profile

[mr. bean]

Hi All (Looking at Chris in particular!),

I have watched a few of the recorded zoom workshops, learning to make more sense of the roast logs to detect issues. I was wondering if you could take a look at these and comment on anything unusual or unwanted...
The profile is Steve's 9min, which demands a pretty high wattage, although I don't think that ended up being a problem. I roasted (in order of below): Kenya Lena, Ethiopian Limu, Peru Kontika, Costa Rica La Lapa, Columbian Excelso (forgive me if I've remembered the regional names wrong - can't be bothered looking them up!).
My aim was to finish them all around 23% DTR (only the costa rica took longer than expected and finished at 18%,with setting 4.2).
ROR is amplified by 10, and smoothed by 20.

Kenyan

Screen Shot 2020-06-08 at 9.58.42 AM.png (104.93 KiB) Viewed 5058 times

log0066.klog

(60.54 KiB) Downloaded 273 times

Ethiopian

Screen Shot 2020-06-08 at 9.59.42 AM.png (104.2 KiB) Viewed 5058 times

log0067.klog

(62.38 KiB) Downloaded 282 times

Peru

Screen Shot 2020-06-08 at 10.00.37 AM.png (108.06 KiB) Viewed 5058 times

log0068.klog

(63.62 KiB) Downloaded 271 times

Costa Rica

Screen Shot 2020-06-08 at 10.01.32 AM.png (113.56 KiB) Viewed 5058 times

log0069.klog

(63.39 KiB) Downloaded 257 times

Columbian

Screen Shot 2020-06-08 at 10.10.16 AM.png (113.74 KiB) Viewed 5058 times

log0070.klog

(64.98 KiB) Downloaded 273 times

To my mind, there appears to be a fair amount of wobble in the red line, and worse for some beans than others... or is it not an issue?

log0066.klog

(60.54 KiB) Downloaded 273 times

[Steve]

Just my 2 cents.

Ive played around fairly extensively with sticking 1.5mm and 3mm sheathed K type probes into the roast chamber situated on the side opposite the hot air inlet = where the black creosote build up is most. Logging these in Artisan with no smoothing was somewhat interesting yet uneventful / not much difference in profile shape just lower recorded temperatures.

My conclusion was that there is a large amount of noise and over sensitivity generated by the bare probe. This has its positives and negatives, sometimes the sensitivity causes the control of the system to react in volatile / jerky ways which are not optimal. Good news is these can usually be easily smoothed out with very careful / VERY small tweaks in zone boost, profile shape or fan profile.

There are a few over reactions in those logs that could be tweaked, but would you be able to taste a difference is the question? That being said ive found dealing with big ROR swings early in the roast prove to be most useful in that it automatically smooths out problems later in the roast which were only caused by the PID control getting stuck in an over reaction / correction cycle.

I continue to find that most important for sweet well developed roasts is that the black power profile remains relatively smooth, no big crashes, flicks or prolonged periods of flattening / decline.

Ive been working on tweaked versions which use similar fan profile / power profile but use less power watts so as to be more usable on lower Voltage, its a bit tricky though.

[mr. bean]

Thanks Steve. If I was going to try something to help smooth the bigger swings in ROR mid way through, what would you recommend starting with?

Btw, I tried to message you a while ago to ask what of your profiles you recommend now? I noticed that you've got some that you now say are not so good, and to use this other one instead etc... but the comments are littered through different posts.

Would you mind posting a new topic with your currently recommended roasts, and if possible comment on what kind of bean / roast degree you recommend them for?

[kaffelogic]

Thank you for posting a comprehensive set of logs.

My quick reply is:

1. You are correct that your power is sufficient for this profile
2. The ragged look to the mean temperature lines is normal. As Steve points out the probe is picking up variations almost between individual beans as they make contact with the probe. Higher quality, better graded beans tend to have less ragged temperature lines.
3. I would recommend some positive boosting around first crack to reduce the crash/flick that is still appearing with some of the bean varieties, provided that this gives some improvement in the cup.

Wayne and I will be discussing your profiles tomorrow in the Virtual Cafe. You are welcome to join us here: https://kaffelogic.com/index.php/home/virtual-cafe. I will record the session and post the link to the video here: https://kaffelogic.com/community/viewto ... p=941#p941.

Cheers
Chris

[mr. bean]

Thanks Chris - that'd be very interesting!

I'll aim to join in.

Cheers, matt (aka mr. bean : )

[kaffelogic]

I continue to find that most important for sweet well developed roasts is that the black power profile remains relatively smooth, no big crashes, flicks or prolonged periods of flattening / decline.

Agreed. In other words, jagged temperature curves are not necessarily a problem, but big swings in the power curve are. The roast will naturally create swings in temperature, but if these appear strongly in the power curve then the roaster is being spooked into over-reacting. Your work in uncovering this principle is much appreciated.

Power profiling is the logical cure for this. I'm happy to provide more assistance with using this feature if desired.

Ive been working on tweaked versions which use similar fan profile / power profile but use less power watts so as to be more usable on lower Voltage, its a bit tricky though.

We are all interested in the outcome of this work. However, your profile does demand conditions that the roaster can really only deliver if it can provide 1400W. So long as the profile sets roast required power to 1400 or 1450 the roaster will give a warning to any user that tries to start the roast with too little power available. This might just be the best option otherwise you are having to mute what you are trying to achieve with this profile.

[Steve]

I continue to find that most important for sweet well developed roasts is that the black power profile remains relatively smooth, no big crashes, flicks or prolonged periods of flattening / decline.

Agreed. In other words, jagged temperature curves are not necessarily a problem, but big swings in the power curve are. The roast will naturally create swings in temperature, but if these appear strongly in the power curve then the roaster is being spooked into over-reacting. Your work in uncovering this principle is much appreciated.

Power profiling is the logical cure for this. I'm happy to provide more assistance with using this feature if desired.

Ive been working on tweaked versions which use similar fan profile / power profile but use less power watts so as to be more usable on lower Voltage, its a bit tricky though.

We are all interested in the outcome of this work. However, your profile does demand conditions that the roaster can really only deliver if it can provide 1400W. So long as the profile sets roast required power to 1400 or 1450 the roaster will give a warning to any user that tries to start the roast with too little power available. This might just be the best option otherwise you are having to mute what you are trying to achieve with this profile.

Thanks Chris, you have created something truly unique and it has helped me continue better my understanding of how coffee roasts in general.

Yeah the power profiling is the next logical step, i played around with it a few times and stepped back from it so as not to have another variable to try and cup against? A bit of fatigue (winter).

I think for slighly longer espresso roasts shallowing out the gradient of the power profile has some interesting flavour developments to a certain point.

[Steve]

Thanks Steve. If I was going to try something to help smooth the bigger swings in ROR mid way through, what would you recommend starting with?

Btw, I tried to message you a while ago to ask what of your profiles you recommend now? I noticed that you've got some that you now say are not so good, and to use this other one instead etc... but the comments are littered through different posts.

Would you mind posting a new topic with your currently recommended roasts, and if possible comment on what kind of bean / roast degree you recommend them for?

No worries Matt replied to your msg earlier today.
Probably a good idea to have an updated profile thread instead of having things scattered all over. I will try to collate my tasting notes with current profiles this week.

Cheers