IKAWA Home - profiles - Page 8
RoR-apaloosa!
Consolidating some recipes that had been posted in another topic thread comparing Ikawa and drum roasters. These were all designed to emulate an 'ideal' drum roaster profile with a steadily and linearly declining RoR curve. Bean temp (whatever "bean temp" means) closely tracks the chamber temperature, so it comes down to describing a compound curve with only 6 time/temperature setpoints - this made me jealous of the Pro owners who get all the points they need.
I added a few other constraints:
9-minute total roast time, spending about 3 minutes drying, browning, and developing respectively. That's a bit long compared to most of Ikawa's recipes, but not unusual for a drum roast profile. Three minutes for each phase may not be optimal, but hopefully not bad either. I am highly receptive to any and all suggestions for how much time to spend in each phase.
Batch size was only 50g, for compatibility with legacy roasters and to limit the emotional trauma if I ruin a batch. Plus, the swirling, twisting, levitating bean mass at 50g is stupid-cool to watch. Larger batches look lame in comparison, IMHO.
Fan speed fixed at 80% start to finish, because changing the fan speed adds another variable to the heat transfer equation. Makes it all 'math-y', and math is hard - I'd rather go roasting.
I used Panama Elida Catuai natural-processed greens, 'cause I have enough for 145 batches - so I better get busy on it!
Here are the recipes, editable with any version of the Ikawa app:
RoR Light roast
https://share.ikawa.support/profile_hom ... uZWkxYgEw
RoR Medium roast
https://share.ikawa.support/profile_hom ... pMWIBMA==
RoR Dark roast:
https://share.ikawa.support/profile_hom ... laTFiATA=
Here's what they looked like:
The light roast was the trickiest - I found that tweaking the temperature setpoint in the light roast profile at 6:30 up or down just a few degrees would have a radical effect on the roast color. I tilted to the dark end of the light roast range to improve uniformity and keep the time after first-crack to at least a full minute.
If someone with one of those fancy roast meters tries out a recipe, I can use that feedback to tweak the setpoints. Likewise for different greens. I post these recipes in the spirit of getting feedback that will make me a better roaster.
Here are the curves. Note the RoR curves show varying degrees of raggedness only because I've been messing with different smoothing algorithms in Excel:
Consolidating some recipes that had been posted in another topic thread comparing Ikawa and drum roasters. These were all designed to emulate an 'ideal' drum roaster profile with a steadily and linearly declining RoR curve. Bean temp (whatever "bean temp" means) closely tracks the chamber temperature, so it comes down to describing a compound curve with only 6 time/temperature setpoints - this made me jealous of the Pro owners who get all the points they need.
I added a few other constraints:
9-minute total roast time, spending about 3 minutes drying, browning, and developing respectively. That's a bit long compared to most of Ikawa's recipes, but not unusual for a drum roast profile. Three minutes for each phase may not be optimal, but hopefully not bad either. I am highly receptive to any and all suggestions for how much time to spend in each phase.
Batch size was only 50g, for compatibility with legacy roasters and to limit the emotional trauma if I ruin a batch. Plus, the swirling, twisting, levitating bean mass at 50g is stupid-cool to watch. Larger batches look lame in comparison, IMHO.
Fan speed fixed at 80% start to finish, because changing the fan speed adds another variable to the heat transfer equation. Makes it all 'math-y', and math is hard - I'd rather go roasting.
I used Panama Elida Catuai natural-processed greens, 'cause I have enough for 145 batches - so I better get busy on it!
Here are the recipes, editable with any version of the Ikawa app:
RoR Light roast
https://share.ikawa.support/profile_hom ... uZWkxYgEw
RoR Medium roast
https://share.ikawa.support/profile_hom ... pMWIBMA==
RoR Dark roast:
https://share.ikawa.support/profile_hom ... laTFiATA=
Here's what they looked like:
The light roast was the trickiest - I found that tweaking the temperature setpoint in the light roast profile at 6:30 up or down just a few degrees would have a radical effect on the roast color. I tilted to the dark end of the light roast range to improve uniformity and keep the time after first-crack to at least a full minute.
If someone with one of those fancy roast meters tries out a recipe, I can use that feedback to tweak the setpoints. Likewise for different greens. I post these recipes in the spirit of getting feedback that will make me a better roaster.
Here are the curves. Note the RoR curves show varying degrees of raggedness only because I've been messing with different smoothing algorithms in Excel:
Maybe I'm doing something wrong here. I can't seem to open the recipes. I already have the app downloaded.
When clicking on the link I get a page with that opens with a button saying "Tap to open roast recipe in Ikawa Home app." When clicking on that, I get sent to the Ikawa website where it tells me to download the app.
What am I missing?
When clicking on the link I get a page with that opens with a button saying "Tap to open roast recipe in Ikawa Home app." When clicking on that, I get sent to the Ikawa website where it tells me to download the app.
What am I missing?
Can you use profiles developed on the Ikawa Pro in the Ikawa Home model? Or reproduce them somehow without the advanced editor subscription?
Is your Ikawa app signed in to their servers?Cwilli62 wrote:Maybe I'm doing something wrong here. I can't seem to open the recipes. I already have the app downloaded.
When clicking on the link I get a page with that opens with a button saying "Tap to open roast recipe in Ikawa Home app." When clicking on that, I get sent to the Ikawa website where it tells me to download the app.
What am I missing?
Sorta. You can download and run the Pro app in 'Guest' mode. In the Pro app you can open a recipe for editing, which allows you to view the exact time/temperature/fan setpoints in a tabular format. Then you transcribe the setpoint into an editable (what they call "legacy") profile. Yeah, that's a pain, but at least it's possible. The recipes I've been posting are editable with the free version of the app, which can be turned into a completely new profile.dbdnt wrote:Can you use profiles developed on the Ikawa Pro in the Ikawa Home model? Or reproduce them somehow without the advanced editor subscription?
I've read some discussion of the airflow rates being different between models, which unfortunately can significantly affect the way heat transfers into the beans. Perhaps someday the HB collective can figure out how to correct for the airflow discrepancies. In the meantime, I focus on the time/temperature settings to try and understand the overall strategy of the roast profile - learn from the masters.
- Peppersass
- Supporter ❤
I don't know if this will help, but here are curves from two batches of Klatch Panama Elida Catuai Natural I roasted in my Ikawa Pro V3 at the end of September, 2021. I used a shortened version of Ikawa's Espresso Roast 1 for Pro roasters.
The first is a light roast, measuring 27 on my RoastVision, close to the middle of the light range:
I always mark 1C when I hear the very first pop. I do this based on a Rao article about marking 1C based on a spike in RoR, which usually happens before 1C gets rolling. In other words, waiting for 1C to get going often results in marking it too late. I've had much better results marking 1C early than marking when it gets rolling.
As you can see, DT is quite short, 45 seconds, and DTR is only 9.2%. I suppose Rao would be horrified, but remember that we're dealing with a tiny fluid bed roaster, not an industrial drum roaster. After 89 roasts with the Pro V3, I've concluded that DT is a much more useful number than DTR. Also note that the exhaust temperature increases about 11F after 1C. I've read that a rise of at least 10F is important for proper development, and my experience so far confirms this. My tasting notes on this roast are "Bright, fruity, best of the three batches." (that I roasted that day.) Wonderful coffee.
The second roast is medium-light, measuring 21 on my RoastVision, which is near the medium end of the medium-light range:
This time the DT is quite a bit longer, 1:15 vs 0:45, and DTR is in the range of what people report for light roasts in drum roasters. The temperature rise after 1C is 17F. Not surprising that this one came out darker. My tasting notes are "Bright, juicy, hint of blueberry. Very slight astringency on finish."
Note that time to dry, which I define as 300F, and time to 1C is about the same for both roasts. I've had a lot of success manipulating roast color/development by manually dropping the roast after a pre-decided DT or temperature rise after 1C. Lately, I've settled on the latter as a more reliable method.
Neither RoR curve looks good from a Rao standpoint, but we're dealing with a much different roaster than his customers use. Also, the scale used for RoR is much too compressed (I've written to Ikawa about that.) The last part of the curve from about 3/4 of the way through Maillard looks more Rao-like, with the medium-light roast, but it flattens out a bit at the end, possibly the cause of the slight astringency on the finish. If I wanted to perfect this roast, I'd probably have to gradually reduce heat just after 1C.
A comment about that: While the Pro series allows an unlimited number of edit point, I'm not convinced that you need more than six. Unlike most roasters, a curve point tells the roaster what you want the temperature to be at a certain time. So, if you want a gradual decline between the start of 1C and drop, you only need to add or edit a point at the drop time. That'll create a downward slope between the start of 1C and drop. It might be necessary to do more than six temperature changes on a gas drum roaster to get a desired gradual decline in heat, but I don't think that's necessary for a PID-controlled roaster. You should be able to get a pretty smooth curve with six or fewer points.
Yeah, it's problematic that you can only program the inlet temperature in the Home, and my curves show that the PID is doing a lot of little tweaks to follow the exhaust temperature curve. But I suspect by adding an exhaust probe to the Home you can come up with six points for the inlet temperature curve that get pretty close to the desired exhaust probe curve.
The first is a light roast, measuring 27 on my RoastVision, close to the middle of the light range:
I always mark 1C when I hear the very first pop. I do this based on a Rao article about marking 1C based on a spike in RoR, which usually happens before 1C gets rolling. In other words, waiting for 1C to get going often results in marking it too late. I've had much better results marking 1C early than marking when it gets rolling.
As you can see, DT is quite short, 45 seconds, and DTR is only 9.2%. I suppose Rao would be horrified, but remember that we're dealing with a tiny fluid bed roaster, not an industrial drum roaster. After 89 roasts with the Pro V3, I've concluded that DT is a much more useful number than DTR. Also note that the exhaust temperature increases about 11F after 1C. I've read that a rise of at least 10F is important for proper development, and my experience so far confirms this. My tasting notes on this roast are "Bright, fruity, best of the three batches." (that I roasted that day.) Wonderful coffee.
The second roast is medium-light, measuring 21 on my RoastVision, which is near the medium end of the medium-light range:
This time the DT is quite a bit longer, 1:15 vs 0:45, and DTR is in the range of what people report for light roasts in drum roasters. The temperature rise after 1C is 17F. Not surprising that this one came out darker. My tasting notes are "Bright, juicy, hint of blueberry. Very slight astringency on finish."
Note that time to dry, which I define as 300F, and time to 1C is about the same for both roasts. I've had a lot of success manipulating roast color/development by manually dropping the roast after a pre-decided DT or temperature rise after 1C. Lately, I've settled on the latter as a more reliable method.
Neither RoR curve looks good from a Rao standpoint, but we're dealing with a much different roaster than his customers use. Also, the scale used for RoR is much too compressed (I've written to Ikawa about that.) The last part of the curve from about 3/4 of the way through Maillard looks more Rao-like, with the medium-light roast, but it flattens out a bit at the end, possibly the cause of the slight astringency on the finish. If I wanted to perfect this roast, I'd probably have to gradually reduce heat just after 1C.
A comment about that: While the Pro series allows an unlimited number of edit point, I'm not convinced that you need more than six. Unlike most roasters, a curve point tells the roaster what you want the temperature to be at a certain time. So, if you want a gradual decline between the start of 1C and drop, you only need to add or edit a point at the drop time. That'll create a downward slope between the start of 1C and drop. It might be necessary to do more than six temperature changes on a gas drum roaster to get a desired gradual decline in heat, but I don't think that's necessary for a PID-controlled roaster. You should be able to get a pretty smooth curve with six or fewer points.
Yeah, it's problematic that you can only program the inlet temperature in the Home, and my curves show that the PID is doing a lot of little tweaks to follow the exhaust temperature curve. But I suspect by adding an exhaust probe to the Home you can come up with six points for the inlet temperature curve that get pretty close to the desired exhaust probe curve.
That does totally help, especially since I have the same greens - thank you!Peppersass wrote:I don't know if this will help, but here are curves from two batches of Klatch Panama Elida Catuai Natural I roasted in my Ikawa Pro V3 at the end of September, 2021. I used a shortened version of Ikawa's Espresso Roast 1 for Pro roasters.
Can you post a link to the Pro recipe? I can read it into the Pro app and convert the time/temp table into a Home recipe.
That initial temperature spike to speed drying is interesting. Up to now I've had the impression the spike was a no-no for natural-processed greens, which I thought were more prone to burning/scorching than washed.
There's a Ikawa v Bullet thread that I think is trending to the conclusion that Ikawa owners don't need to sweat the RoR curves - which opens up a lot more potential for doing some crazy/fun profiles.
- Peppersass
- Supporter ❤
Here's a link to the Ikawa Pro Espresso 1 profile that I started with.
If you'd like a copy of my edited version, PM me your email address and I'll send it to you.
If you'd like a copy of my edited version, PM me your email address and I'll send it to you.
GDM528 wrote:Is your Ikawa app signed in to their servers?
Yeah, I'm signed in. On the Android app. Does it matter that I don't have the roaster in my possession yet? I was trying to get comfortable with how to use the app and share profiles/recipes before the roaster arrives.
If you are able to convert this to a home recipe, would you be willing to share it?GDM528 wrote:That does totally help, especially since I have the same greens - thank you!
Can you post a link to the Pro recipe? I can read it into the Pro app and convert the time/temp table into a Home recipe.
That initial temperature spike to speed drying is interesting. Up to now I've had the impression the spike was a no-no for natural-processed greens, which I thought were more prone to burning/scorching than washed.
There's a Ikawa v Bullet thread that I think is trending to the conclusion that Ikawa owners don't need to sweat the RoR curves - which opens up a lot more potential for doing some crazy/fun profiles.
Would love to test it out with my new exhaust probe setup (see other Ikawa thread).
To convert it, it seems likes the first profile point would be the Pro preheat temp, and once the Ikawa Home reaches that first point, then you would dump in the beans (rather than starting with the beans in the chamber, then spiking the temp). That seems like that would be the best way to most closely mimic what the pro profile was doing (at least in the beginning)
I agree I would have thought the spike would scorch/burn natural process beans.
Peppersass, do you adjust preheat temp or the height of the spike for different bean types?