Thanks. I'll just ask for that profile then. Still, I guess it could be interesting to compare how close they are if someone's game.

drgary wrote:I think a key part of dialing in your use of any roasting machine is tasting your results and doing that over different beans and roast levels within your range of preference.

Absolutely, tasting is key that's how I've found that sometimes/often extended flat RoR can lead to exactly the kind of bad results you describe. But for some beans there might be one 'pretty' batch and one wonky, with the latter tasting a lot better. Sorry if this is a bit off topic.

Hi,
I am very interested in the CB profile you have devised. I am trying with a very similar hot air roaster, and I too am interested in the two-step roasting, and I also share your stance of countering the theories. We are getting adjusted to a pretty good point. Have there been any updates to the CB profile since post #176?

And one more question. Misama phase is unique, but some people think that hydrolysis can cause astringency if a lot of water is left behind, but you dare to caramelize it on the outside with a lot of water left behind, coating it to prevent water from escaping and making it retain more water. Going from the perspective of hydrolysis, it would be better to enter the Maillard phase with some of the water skipped, but I am interested in the rationale behind daring to take the exact opposite approach.
Does it mean that more water is needed to keep the Maillard phase longer?

Best .
Tomo.

Tomo2023 wrote:Have there been any updates to the CB profile since post #176?

And one more question. Misama phase is unique, but some people think that hydrolysis can cause astringency if a lot of water is left behind, but you dare to caramelize it on the outside with a lot of water left behind, coating it to prevent water from escaping and making it retain more water. Going from the perspective of hydrolysis, it would be better to enter the Maillard phase with some of the water skipped, but I am interested in the rationale behind daring to take the exact opposite approach.
Does it mean that more water is needed to keep the Maillard phase longer?

Best .
Tomo.

My latest version was posted here: Humidity sensor on exhaust - other gas sensors ?

Link to profile: https://share.ikawa.support/profile_hom ... oAIgAoAA==

The numbers (very specific to late-model IKAWA roasters):
30sec = 220C inlet
5:30 = 210C inlet
6:00 = 265C inlet
9:00 = 255C inlet

I bumped the development phase duration to about 2:30 post first-crack, and gave it a full five minutes in the browning/Mailliard/caramelization phase. At nine minutes, some might consider that too 'bakey', but I'm not a fan of 'bright' third-wave coffee; much prefer '2.5 wave' coffee-coffee. Beans come out with a subtle sheen, just shy of Full-City.

Regarding "Misama":
I think running the browning phase for five minutes is enough time to reach well into the core of the bean, and the moisture data I'm collecting now seems to corroborate that. Going shorter can increase astringency, but much of that can be compensated for in the development phase.

I've read here and there that the presence of water can be essential for certain tasty chemical reactions to occur, so I thought, why not for coffee too? That's, like, the full extent of my rationale. Small-batch air roasters have a propensity for driving out moisture really fast, hence getting to caramelization temperatures as quickly as possible, such that some amount of water is still present for a minute or so.

The stepped profile is much easier to replicate on other air roasters like the FreshRoast. Just two heat settings. Duplicating a linearly-declining RoR profile on a FreshRoast is much harder. That said, lately I find myself toggling between this stepped profile and a linear RoR profile, and happy with both.

GDM528 wrote:
I bumped the development phase duration to about 2:30 post first-crack, and gave it a full five minutes in the browning/Mailliard/caramelization phase. At nine minutes, some might consider that too 'bakey', but I'm not a fan of 'bright' third-wave coffee; much prefer '2.5 wave' coffee-coffee. Beans come out with a subtle sheen, just shy of Full-City.

Regarding "Misama":
I think running the browning phase for five minutes is enough time to reach well into the core of the bean, and the moisture data I'm collecting now seems to corroborate that. Going shorter can increase astringency, but much of that can be compensated for in the development phase.

I've read here and there that the presence of water can be essential for certain tasty chemical reactions to occur, so I thought, why not for coffee too? That's, like, the full extent of my rationale. Small-batch air roasters have a propensity for driving out moisture really fast, hence getting to caramelization temperatures as quickly as possible, such that some amount of water is still present for a minute or so.

The stepped profile is much easier to replicate on other air roasters like the FreshRoast. Just two heat settings. Duplicating a linearly-declining RoR profile on a FreshRoast is much harder. That said, lately I find myself toggling between this stepped profile and a linear RoR profile, and happy with both.

Thank you for your detailed reply. I like a few light roast coffees as well, but most of them are rarely tasty.

On the other hand, I love roasts with lots of Maillard reaction, sweet and chocolatey. 2.5wave is a good name for it! I like that too!
There are a few people in Japan who do 2-step roasting, but it is rare to find a case where the first step advances the caramelization as much as yours. Very unique, but this seems to make a lot of sense.


By the way, most people in the world don't know this, but there is a roasting machine in Japan that is partially compatible with Ikawa. Panasonic and Ikawa have partnered to make this roaster. It looks just like the Ikawa, but there are some minor differences. 50g and inlet control only. Unfortunately, it was discontinued last year. I use that roaster. The sensor position of the inlet is different from that of the Ikawa. I imagine that the sensor is located somewhere on the back side near the chamber and receives feedback on the temperature of the beans that come in contact with the chamber. I imagine it would be somewhere behind the stirring prongs. So you would need to set the temperature about 20-30 degrees Celsius lower than your profile. Also, the speed of the temperature rise and fall response is much slower than the Ikawa inlet. It is similar to the behavior of exhaust temperature. Thanks to this, it is possible to hit the points as if you were controlling the BT.


Fan speeds are also not interchangeable.
You will have to adjust your profile to fill in the differences. But it's also a fun process.
I can't wait to drink a good latte, so I'll work on adjusting my profile.
Thanks!

GDM528 wrote: ...Small-batch air roasters have a propensity for driving out moisture really fast, hence getting to caramelization temperatures as quickly as possible, such that some amount of water is still present for a minute or so.

This is something that I've been trying to dial in on my roaster recently. I agree fluid beds seem to have a drying affect on the beans. I would guess the humidity level in the location of the roaster has a lot to do with it too. I imagine that a fluid bed being run in Florida with 98% humidity will act quite a bit differently than the 25-30% we've had here in the last couple months. Although certainly the same could probably be said for drum roasters. I want to try two back-to-back roasts at some point with the same profile and bean, but run a humidifier blowing into the roaster inlet and see if there's a difference in weight loss. Someday.

I want to start playing with faster drying times, but I'm also battling off-flavors in certain coffees that I think are coming from too much heat. One is severe tipping on a low-grown Honduras, and a high-grown Yemen, the results of which are almost undrinkable in both. Then there's also a heavy base note in a white honey process Costa Rica that I think might be the dried mucilage layer taking on too much heat, but that's pure speculation.

Can I ask you what kind of dry times you're running these days? And if you're still taking BT readings, I'd be very curious what temps you usually hit at 1:00, 2:00 and 3:00.

Erichimedes wrote:Can I ask you what kind of dry times you're running these days? And if you're still taking BT readings, I'd be very curious what temps you usually hit at 1:00, 2:00 and 3:00.

From the moisture readings I'm taking now, it looks like I may never completely dry the greens. But looking at the difference between inlet temperature and BT, it appears the beans are tracking the inlet 1:1 after about 2-3 minutes. Beyond that point the remaining moisture may be inconsequential in cooling the beans, so maybe that qualifies as 'dry'.

Some illustrative examples:
IKAWA Home - profiles
IKAWA inlet profile calculator

I don't think that air roasters unduly dry the coffee. It's a matter of adjusting parameters, so that the beans heat evenly through. Consider that drum roasters also often have adjustable airflow, and the application of heat can also be mostly convective. Added: I recently tried three roasts at maximum air on my 1 kg propane perforated drum roaster. That simulates an air roaster, and my coffees were not dried out. More of a concern was adjusting the flame to avoid too high a temperature that could unevenly roast the coffee. This showed up when I roasted a dense Kenya compared to two decafs, which have low density by definition, since the decaffeination process breaks down the cell structure of the coffee. I was experimenting with a concept that was posted by Rob Hoos, who says that he now automatically sets his drum roaster fan at 100%.

Also, I amskeptical about heavy moisture in the environment affecting the moisture content of the coffee, considering that you are applying so much heat at the inlet of an air roaster that you would drive away water vapor from the air.

At least those are my initial thoughts on the prior post. I may not be correct.

drgary wrote:I recently tried three roasts at maximum air on my 1 kg propane perforated drum roaster. That simulates an air roaster

Are you sure about that? I thought air flow was a lot higher for fluid bed.

I quite like the more medium to dark roasts I get on the Kaffelogic* compared to the Quest** so I'm not sure whatever drying's going on is an issue to begin with.

*fluid bed
**drum

"Simulates." So, I'm not sure of that but I'm sticking with my theory. Simulate and replicate have different meanings.

1. My roaster is unusual for its model because it has a perforated drum.
2. It required a higher burner setting to track with my usual drum roasts.
3. My usual fan setting is 30. This was 100. And no, I don't measure the actual airflow.

The point was to use my roaster like a convection oven, moving the heat more uniformly through the roaster. The vent from my cyclone chaff collecting fan got quite hot.

I expect to have an IKAWA roaster in the Fall, so can't compare them. I will be trying another air roaster with agitation in a month or two.

drgary, thanks for chiming in.

Rob Hoos sets fan speed (highly equipment-specific) to 100% for every roast, or just decafs?

What does "dry" mean, really? Absolutes can be hard to achieve and may exact a penalty on other aspects of the roast. My humidity measurements seem to indicate it could take 20 minutes to completely drive out all the drive-out-able water in the greens. How is 'dry end' marked in a drum roaster?

Rather than dry the greens during the roast, I tried dehydrating the greens prior to roasting. It kinda-sorta roasted normally, save one fascinating quirk in the BT curve. The resulting espresso was kinda bland. So that got me thinking that perhaps water during the roast can be my friend, once harnessed.