IKAWA Home - profiles - Page 6
-
Auctor
- Posts: 432
- Joined: 3 years ago
This is a really interesting take. I haven't really experimented properly with the Ikawa yet, but based on what i've seen across different varieties, the curve shape itself (convexity and concavity) can vary significantly (in addition to development time). Is it your thought that the shape is less important?nicolai wrote: When I bought it I was expecting I would really focus roast profiles, but after I bought my roast meter I almost always use the same profile and just change the development time based on the roast level I want to achieve.
-
nicolai
- Posts: 26
- Joined: 3 years ago
We are moving in to the world of controversy, but Yes, that is what I am saying.
You of course need a decent drying phase up to FC - on air Beds around 5 mins and then development phase to the roast level you want.
I did the same bean developed with different profiles to almost the same roast level (Agtron 85) and at least my untrained palette was unable to taste significant differences in cupping.
It is probably the 80/20 rule - 80% of the cup quality comes from using good beans and roasting to a level you desire and which fit the beans. And the 20% comes from the roast curve? Maybe it is even 90/10
Again - maybe dumbing I at least experienced that quality greens and roast color is crucial.
I bought a roast meter from espresso-vision for the same reason.
You of course need a decent drying phase up to FC - on air Beds around 5 mins and then development phase to the roast level you want.
I did the same bean developed with different profiles to almost the same roast level (Agtron 85) and at least my untrained palette was unable to taste significant differences in cupping.
It is probably the 80/20 rule - 80% of the cup quality comes from using good beans and roasting to a level you desire and which fit the beans. And the 20% comes from the roast curve? Maybe it is even 90/10
Again - maybe dumbing I at least experienced that quality greens and roast color is crucial.
I bought a roast meter from espresso-vision for the same reason.
-
nicolai
- Posts: 26
- Joined: 3 years ago
And btw. I cannot take credit for that thinking.
Science supports it that roast development (color) is much more important then development time.
Here is some pretty heavy statistics talking about the same thing done by the coffee roasting company coffee-mind.
https://www.mdpi.com/2306-5710/6/2/29/htm
The content is also covered in this podcast with Tim Wendelboe.
https://www.podbean.com/ew/pb-bqkps-10d81d6
Science supports it that roast development (color) is much more important then development time.
Here is some pretty heavy statistics talking about the same thing done by the coffee roasting company coffee-mind.
https://www.mdpi.com/2306-5710/6/2/29/htm
The content is also covered in this podcast with Tim Wendelboe.
https://www.podbean.com/ew/pb-bqkps-10d81d6
-
mgrayson
- Supporter ♡
- Posts: 665
- Joined: 17 years ago
In one attempt with a non-Ikawa bean, the development time was about half what I was expecting and the result was a very good medium roast. ¯\_(ツ)_/¯ Obviously, people use different profiles for *some* reason, and I hope to learn what and why.
-
nicolai
- Posts: 26
- Joined: 3 years ago
Of course - I am not saying it is completely irrelevant, but the primary reason you got a good result was because you nailed your roast level and had a good input bean for what you were trying to achieve.
I am aLao really hope to understand more, but it is like building a house, the foundation has to be good.
I am aLao really hope to understand more, but it is like building a house, the foundation has to be good.
-
GDM528
- Posts: 853
- Joined: 2 years ago
They seem to work, but there were changes to the thermal mass of the system and the addition of the little 'ski ramp' in the chamber could produce a different result I suppose.mariowar wrote:Thanks, i also saw these ones with my Android, however, I wonder how recipes for a different roaster and different charges would work on the Ikawa Home 100...
Am.I missing something?
-
GDM528
- Posts: 853
- Joined: 2 years ago
Can't think of what else beyond air temperature and volume I would want to control. And at least for now, dynamically adjusting the air speed is too complicated for me, so I just peg it at a fixed speed for the entire roast: 80% speed for a small 50g batch size, which nicely elevates the bean mass.nicolai wrote:I have not really seen the graph editor, but there are really only two parameters that the software can control - air speed and the temperature of the heating element?
Do you guys miss any of the knobs from the more advanced roasters? (If you ever had one)
Not technically a 'knob' but the biggest thing missing from the Home version is a temperature sensor for the beans in the chamber.
-
GDM528
- Posts: 853
- Joined: 2 years ago
Does shape matter? I created two profiles to test that question.
"RoR". There's a school of thought that the optimum roast profile has a steadily declining Rate of Rise - in effect the beans are always trying to catch up with the surrounding temperature. So I ginned up a set of temperature setpoints that would mimic the kind of profile a drum roaster would produce. Some of Ikawa's own curated profiles do the same, but not at my level of obsession to get a linearly declining RoR.
Graph Decoder Ring:
Red = Ikawa program setpoints.
Blue = Temperature of empty chamber (first run).
Green = Temperature of chamber with a 50g batch of beans inside (second run).
Purple = RoR of bean-mass temperature.
Also marked the timing of first crack start and end.
Must say I'm feeling kinda smug about how straight that RoR curve is. Yet another magic moment in the Ikawa's roast control.
"Stepped". This profile 'deconstructs' the roasting curve into separate stages for drying, browning, and development. The concept here is to divide and conquer, separating each step so they can be individually optimized without interacting with the other steps. For example, the first drying step can be adjusted to 'level the playing field' for whatever beans I load in such that they go into the browing phase completely dry. Likewise, the browning stage can be moderated to minimize burning of natural-process beans. Could be a dumb idea, but it does serve well as an example of a profile that has complete disregard of RoR.
Same decoder ring as previous graph:
The Stepped profile was Q-matched to the RoR profile, meaning both profiles delivered the same volume of heat (area under curve) to the beans for all temperatures above 160C. Below 160C, the RoR profile delivered about 4% less heat because of its more gentle ramp - but theoretically the effects on bean chemistry below 160C are supposed to be minimal.
There was a 30 second difference in the first-crack timing such that the RoR profile may have gotten a bit more development time. My theory for the difference is the more gentle slope of the RoR curve preloaded more heat into the beans going into first crack. If I wasn't trying to match total heat delivered, it would've been prudent to add another 30 seconds to the Stepped profile.
Both came out looking like a City roast:
My big takewaway here, is they look virtually the same - despite radically different roasting profiles! The only thing they had in common was the total heat delivered (area under curve). Even the 30 second difference in development time doesn't seem to have shifted the color. If I stare at each batch for a while, I think the the RoR roast is a teensy bit more uniform.
Of course, just because they look the same doesn't mean they'll taste the same. I'll let them rest several days before testing - but I don't consider myself to having any viable 'tasting' skills. So, I'm hoping that some of you with beans to spare and a better-trained palete can dedicate two 50g doses of beans to science, and the question of does shape matter:
RoR profile:
https://share.ikawa.support/profile_hom ... uZWkxYgEw
Stepped profile:
https://share.ikawa.support/profile_hom ... uZWkxYgEw
Note that both recipes were designed for a small 50g batch size, but the total heat delivered would still be the same between them, and you may end up with a slightly lighter roast if you use a larger dosage. Both should also be fully editable with the free iOS and Android apps, if you're inclined to tweak them - just note that you'll change the Q match if you're not careful. Also, your beans may levitate differently, and some may eject into the chaff as the roast proceeds. You can edit the fan speed in the free app to raise/lower the fan speed/bean mass during the roast. I've settled on a flat 80% fan for the entire roast, and rarely see a bean fly into the chaff from a 50g dose. If you do adjust fan speed, I suggest keeping the fan speed flat throughout the roast to keep the total heat delivered the same between RoR and Stepped.
Thanks in advance to any brave explorers who give this a try!
"RoR". There's a school of thought that the optimum roast profile has a steadily declining Rate of Rise - in effect the beans are always trying to catch up with the surrounding temperature. So I ginned up a set of temperature setpoints that would mimic the kind of profile a drum roaster would produce. Some of Ikawa's own curated profiles do the same, but not at my level of obsession to get a linearly declining RoR.
Graph Decoder Ring:
Red = Ikawa program setpoints.
Blue = Temperature of empty chamber (first run).
Green = Temperature of chamber with a 50g batch of beans inside (second run).
Purple = RoR of bean-mass temperature.
Also marked the timing of first crack start and end.
Must say I'm feeling kinda smug about how straight that RoR curve is. Yet another magic moment in the Ikawa's roast control.
"Stepped". This profile 'deconstructs' the roasting curve into separate stages for drying, browning, and development. The concept here is to divide and conquer, separating each step so they can be individually optimized without interacting with the other steps. For example, the first drying step can be adjusted to 'level the playing field' for whatever beans I load in such that they go into the browing phase completely dry. Likewise, the browning stage can be moderated to minimize burning of natural-process beans. Could be a dumb idea, but it does serve well as an example of a profile that has complete disregard of RoR.
Same decoder ring as previous graph:
The Stepped profile was Q-matched to the RoR profile, meaning both profiles delivered the same volume of heat (area under curve) to the beans for all temperatures above 160C. Below 160C, the RoR profile delivered about 4% less heat because of its more gentle ramp - but theoretically the effects on bean chemistry below 160C are supposed to be minimal.
There was a 30 second difference in the first-crack timing such that the RoR profile may have gotten a bit more development time. My theory for the difference is the more gentle slope of the RoR curve preloaded more heat into the beans going into first crack. If I wasn't trying to match total heat delivered, it would've been prudent to add another 30 seconds to the Stepped profile.
Both came out looking like a City roast:
My big takewaway here, is they look virtually the same - despite radically different roasting profiles! The only thing they had in common was the total heat delivered (area under curve). Even the 30 second difference in development time doesn't seem to have shifted the color. If I stare at each batch for a while, I think the the RoR roast is a teensy bit more uniform.
Of course, just because they look the same doesn't mean they'll taste the same. I'll let them rest several days before testing - but I don't consider myself to having any viable 'tasting' skills. So, I'm hoping that some of you with beans to spare and a better-trained palete can dedicate two 50g doses of beans to science, and the question of does shape matter:
RoR profile:
https://share.ikawa.support/profile_hom ... uZWkxYgEw
Stepped profile:
https://share.ikawa.support/profile_hom ... uZWkxYgEw
Note that both recipes were designed for a small 50g batch size, but the total heat delivered would still be the same between them, and you may end up with a slightly lighter roast if you use a larger dosage. Both should also be fully editable with the free iOS and Android apps, if you're inclined to tweak them - just note that you'll change the Q match if you're not careful. Also, your beans may levitate differently, and some may eject into the chaff as the roast proceeds. You can edit the fan speed in the free app to raise/lower the fan speed/bean mass during the roast. I've settled on a flat 80% fan for the entire roast, and rarely see a bean fly into the chaff from a 50g dose. If you do adjust fan speed, I suggest keeping the fan speed flat throughout the roast to keep the total heat delivered the same between RoR and Stepped.
Thanks in advance to any brave explorers who give this a try!
-
mathof
- Posts: 1487
- Joined: 13 years ago
I have taken up GDM528's invitation to use his stepped and RoR profiles. I have the previous 50g Home roaster, tricked out with a thermocouple placed in much the same position as his and run through an Amprobe TMD-56 to a MacBook Pro running Artisan. Here are the results of today's roasting (I used Guatemala Poaquil beans):
As you can see, the AUC of the two roasts differs markedly, with the stepped profile measuring 240C*min and the RoR profile 313*C min. This is reflected in the colours of the roasted beans. The former measures 78 on my Tonino (interpreted by the Tonino as City) and the RoR a stage darker at 70 (interpreted as Dark). With these measured differences, they will certainly taste different.
As you can see, the AUC of the two roasts differs markedly, with the stepped profile measuring 240C*min and the RoR profile 313*C min. This is reflected in the colours of the roasted beans. The former measures 78 on my Tonino (interpreted by the Tonino as City) and the RoR a stage darker at 70 (interpreted as Dark). With these measured differences, they will certainly taste different.
-
mgrayson
- Supporter ♡
- Posts: 665
- Joined: 17 years ago
Great to see these experiments!
Pardon my unfamiliarity with these plots, but where do the numbers (I assume they are cutoffs to the AUC) 167.4 and 160.3 come from? The relevant numbers look like 150 and 148.
Pardon my unfamiliarity with these plots, but where do the numbers (I assume they are cutoffs to the AUC) 167.4 and 160.3 come from? The relevant numbers look like 150 and 148.
