Slow Light Roasts

[JonFairhurst]

It’s been challenging for me to do slow light roasts on the KL. Is it just me, but every bean seems to roast very quickly.

Let’s start with the elevation Rest profiles. Even with high elevation beans, I find that I need to use low elevation profiles, and they can still be on the fast side (FC @ ~5:00.) I’ve also tried Raost and Adaptive Raost as starting points.

Additionally, I will sometimes get a shutdown due to excessive ROR on the same bean and same profile on different roasts. Looking at the logs, the temp wasn’t excessive. It was just playing catch-up.

On other roasts, I sometimes get a shutdown with the heating too slow error. I hate that! I’m shooting for a slower roast! Keep going!

I’ve been using 1/4 batches (113g.)

BTW, with the faster roasts, I get non-uniform results where the overall roast is medium, but some beans are oily after a few days.

I’m coming from a Behmor, which tends to lack power and isn’t able to do very dark roasts, unless the batch size is small.

Any guidance for doing slower, lighter roasts is appreciated. Also, are there settings that I can apply to make the machine more tolerant of low or high ROR?

Finally, are there calibrations I need to do? There was a link to calibration instructions in one of the threads, but it’s a dead link. I’ve looked at the manual and couldn’t find it. My machine is a 120V model and was shipped with the new chaff collector. Where do I look for the fan calibration? FWIW I’m at 2,000 ft elevation, if that makes any difference.

The machine has a lot of potential, but I’m a bit frustrated by the software right now. Too many failed roasts.

[Geronimo]

Hi

Out of interest what is your fan calibration factor. For example mine is 0.96

Cheers
G

[rosemarycole6]

Here is the link to the fan calibration/re-calibration instructions https://kaffelogic.com/community/viewto ... tion#p2899.
Additionally, which beans are you trying to slow roast and what were their processing methods?

[JonFairhurst]

Thanks! I’ll check the calibration shortly. Maybe I need more fan speed at my elevation (600m+).

I’ve roasted a variety of beans, and I believe that most were washed. Rather than a bean issue, it seems difficult to get FC past about 5:30, and uneven roasts are chronic. I have a log and can list them later…

[JonFairhurst]

My calibration is 0.97.

A recent, difficult bean was Kenya Nyeri Gatugi AB. Washed. 1900m. I started with the 1200-1500 Rest profile. FC at 5:03. I then did 0-1200 Rest. I got it to 6:20 for FC, but preheat was low, and the log was very unstable. I stopped manually at 18% development, which was only L0.1.

My worst was Tanzania Sepukila Coop AB. Washed. Under 1500m. I had started with Raost Adaptive, but scaled slower. I kept adjusting the preheat, but I got four runaway temperatures in a row. The beans seemed to be moving well, but now I’m thinking that I needed to scale up the fan speed more.

I had another frustrating Kenya washed pea berry 1800m as well, with multiple runaway temps. That was also based on Adaptive Raost.

More cowbell. And more fan!

[Geronimo]

Hi

Out of interest can you try one of your higher altitude beans like the Kenya with the Ninja Turtle profile. I’m using Kenya AA beans and FC is around 7’45”. Ethiopia Yirgacheffe around a similar time. Noting that I’m recording FC when Ive got a few cracks happening.

Cheers
G

[rosemarycole6]

JonFairhurst,
when you say "run away temperatures" do you mean at the end of the roast or the start of the roast? Many of the roasts do runaway at the start but the roaster drags the temp back down to the curve fairly quickly. I've not had that problem at the end of a roast with the KL profiles only with ShapeShifter - where the temp at the end accelerated way to quick. You can also adjust the zones if you're finding these are boosting temps too much.

The Roast Level I find isn't that reliable or consistent across profiles or even roasts of the same bean & profile combo sometimes. So I stand over the roaster and gauge the best time to eject based on a combo of colour and DTR and depending on the processing method. I use a spreadsheet to log all the details and then assess the colour, DTR and then taste profile against each roast to work out repeatability.

[JonFairhurst]

@Geronimo - Thanks for the suggested profile. I also wait for a series of cracks before I log FC. I don’t trust the odd outlier.

@rosemarycole - Yes, I get the runaway temps at the end, typically right after FC. FI get the rare “too slow” warning during the drying phase. I also don’t tend to use the Level to end the roast. I generally set it to L5.9 and end the roast manually at a given percentage after FC. That said, I plan to start ending at a goal temperature, rather than time or percentage, because marking FC can be more art than science.

[JonFairhurst]

Success!!!

I increased my fan calibration from 0.97 to 0.99 and tried the Ninja Turtle profile with a washed El Salvador, 1,000 to 1,300 m. The four 113g roasts were consistent with nice, stable tracking. Here are rough averages…

Dry: 172C, 4:45, 57%
FC: 199C, 6:35, 22%
Drop: 216C, 8:20, 21%

The first roast started a little hot. I targeted a 20% dev time. It ended up a bit light and smelled sharp and spicy. It stopped at L1.0.

The next roast started a bit slower and finished a bit hotter. FC was a bit later, so my 20% development was also later. Turned out a bit darker at L1.4. The smell was dull by comparison.

I did the final two roasts at L1.2 to split the difference. Sure enough, it gave off a more muted spice smell, and seemed sweeter. The Goldilocks roast.

I couldn’t be happier with the results. Consistency, great tracking, predictable. No hint of tipping. I can’t wait to try this with a denser bean.