IKAWA Home - profiles - Page 18

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
GDM528
Posts: 830
Joined: 2 years ago

#171: Post by GDM528 »

jz_coffee wrote:Ristretto baby! I too miss having more liquid in the cup but the taste improvement with this roast level is undeniable. The last 3 days all i have been pulling is 20g in 30g out shots using Damian's Lr3 profile on the Decent and they are pretty much like a dark-chocolate milk shake.

...

Any thoughts on shortening cooling time to 1min instead of 2? I tried that on the last couple of batches just to speed up the process- i felt like with batches this small they cool down pretty fast in my colander.
Much appreciation for the Ristretto revelation. To paraphrase from the movie Jaws: "I'm gonna need a bigger basket". This has me wondering if roasting for Ristretto is a thing...

I too have shortened the programmed cooling time to just one minute - but it always takes longer because the Ikawa's self-preservation algorithms will keep the fan running until the system has sufficiently cooled down (mid-50C range). I drop the fan speed from 80% down to 70% just to create an audible cue that the roast is ending, but if the fan was kept at 80% then it might stop closer to the 1-min mark. The restart time for the next run is definitely quicker and more variable, but I've not observed any significant change in the in-chamber thermal readings once the roast is underway.

User avatar
MNate
Posts: 956
Joined: 8 years ago

#172: Post by MNate »

jz_coffee wrote: I ran 3 batches of this blend, tweaking the CB profile each time:
Are you roasting the types together just for ease of process? With such small batches it's pretty easy to roast one bean type at a time and blend post-roasting. I would think tweaking each bean individually would get better results. Isn't this the typical approach?

I've been meaning to ask some questions on best blending practices but figure I could just reread what I read several years ago. Occasionally I'll think the bean types taste best at different roast levels such that I'll even grind them before blending since they are a few ticks apart on the Niche and it's so easy to adjust. I'm sure there are some best practice theories on this though.

jz_coffee
Supporter ♡
Posts: 27
Joined: 2 years ago

#173: Post by jz_coffee »

:-)Yes the 20g basket has become a good friend. Might even get a bigger one! Thanks for the confirmation on the 1min cooling time. Doing another 3 batches now before the weekend officially is over. The one in the Ikawa as I type is 20g Anerobic Robusta / 20g Monsooned Malabar and 20g washed Ethiopian Yirgz. Time will tell...

jz_coffee
Supporter ♡
Posts: 27
Joined: 2 years ago

#174: Post by jz_coffee »

So I have actually done both and with the Ikawa at least i cant seem to tell the difference with these little batches. It's easier to roast pre-blended so that is what I am doing for these latest batches...

jz_coffee
Supporter ♡
Posts: 27
Joined: 2 years ago

#175: Post by jz_coffee »

Update: received a brief response from David Schomer re: Dolce roasting tips/ recommendations. Nice that he responded although nothing very useful here (wasn't expecting much though).

Hi Jon

To heat the coffee bean through and through you need about a 15 minute total roast cycle. And for the finish try to match the color on the dolce that we roast

Good luck and thank you for your business

David

Sent from my iPhone

GDM528
Posts: 830
Joined: 2 years ago

#176: Post by GDM528 »

Update to post #135: IKAWA Home - profiles

An app-controlled, thermally-agile, small-batch roaster has become my metaphorical cable TV tuner for roasting, and I can't put down the remote ;) I've been constantly tweaking and playing with every roast, exploring the 'roasting space'. Most of what I've learned is unique to me, but I thought I'd share a few notes that might be generally useful to any other experimenters out there.

The stepped "CB" profile has been my favorite platform for experimentation. "CB" stands for Creme Brule, aka 'burnt cream'. My notion was to coddle the sugars (mostly sucrose) in the bean without subsequently burning them all up during the development stage. It has a distinctly different taste profile from a more 'normal' declining RoR roast, with both light and dark attributes. This profile also flips a middle finger at linearly declining RoR and maximizes the horror by stalling... twice!

This chart is from a recent experiment, overlaid with a conventional declining-RoR roast. The stepped roast was tweaked to match the ending temperature and two-minute development time. The roasted beans of the two different profiles look nearly identical:


The link to the stepped (red) profile is here: https://share.ikawa.support/profile_hom ... pMWIBMA==

Odds are high that I'll be trying a different profile next week, but the linked profile is the base I use to tweak from. This particular profile has produced a Full City roast on a half-dozen different origins of greens I've tried so far. BTW, the link is from the Android version of the app, which still sends encoded profile data in the email, which I can share here. It's only a matter of time before Ikawa blocks this.

The thermal break between the browning and development stages makes it easier for me to tweak each stage kinda/sorta independently to better learn the effects of time and temperature. The fast initial ramp adds what I'm calling a 'miasma' phase, where water trapped in the interior of the bean moderates and steams the bean while the exterior is above the caramelization temperature for about two minutes before browning - modeled after the 'wet method' for making caramel. Could be stupid, but that's not gonna stop me from trying and learning.

The previous version of the "CB" stepped profile had very fast and large temperature ramps. I've observed that will occasionally flummox the Ikawa's control algorithms, causing it to settle into a much lower temperature than the actual program. That has taught me two things: slow down the ramps to 30sec, and always watch the roast progress on the app so I can cancel and restart the roast if it ever gets confused.

The temperature window for the browning phase is only about 10C wide. Drop the setpoints below 210/200 and the roast turns grassy and vegetal, and going up to 225/215 setpoint will burn the caramels and get too close to first crack. On the other hand, there seems to be a lot of latitude to the duration of the browning phase. Cutting the time in half adds sour notes. Increasing the time another minute or so is smoother but more bland (baked?).

I've found that setpoints above 270C adds a 'burnt for no good reason' taste that I might appreciate more once winter sets in. For now, I've lowered the development temperature and bumped the development time to stop the roast about two minutes after first crack. If I go any longer the beans look more uniform in color, but I fear that does too much damage to the caramelized sugars. If I stop at one minute after first crack, the roast gets too acidic/sour for my tastes.

Many thanks to jz_coffee, who showed me that lower brew ratios are better suited to the stepped roast profile. I don't understand the science/chemistry of why, but it does add an interesting twist to how I roast and brew.

User avatar
MNate
Posts: 956
Joined: 8 years ago

#177: Post by MNate »

GDM528 wrote:Could you post your list of 8 coffees in the Ikawa Home Profiles thread? I see the three bags in your photo - but you folded over the part that names the source! - lol. You're doing what the Ikawa is good at: polyroastery.
Sure. I may have mentioned these before:




Of these the Max Pérez was our favorite. I ended up roasting each at a M++ but different Ikawa-bean based profiles that I thought matched the bean.

In the end I think my roasts matched the general idea of the descriptions and we favored, as we usually do, those that started with chocolate, maybe had some fruit, and we didn't like as well the more acidic descriptors. I thought they were all good though.

These pictured were the Guatemalas.

I haven't actually roasted the Brazil yet. I tend to do Brazil as a base to blends. And I haven't tried the Ethiopian, it's resting at the moment.

ira
Team HB
Posts: 5476
Joined: 16 years ago

#178: Post by ira »

I adjusted the starting temperature to clean up the beginning of the roast. No clue if it helps, doubtful, but it certainly looks better. I also smoothed it out a bit and then adjusted the end so I could just let it finish on it's own. So now it seems to taste pretty good and be really easy to use.

Cwilli62
Posts: 208
Joined: 5 years ago

#179: Post by Cwilli62 »

Here is the profile that I started using as a general go to and it has been solid for my taste preferences with most of the coffees I've tried it with (but not all). I had started with someone else's that was a complete bust for me (didn't even hit 1st crack) and then tweaked it.



The issue for me is that I'm still having trouble figuring out how adjustments on any given profile with the Ikawa will alter the end result in the cup. So I'm finding it difficult to confidently "play" with profiles. I've read a couple of roasting books and some articles...but these people are using drum roasters. And I will admit that my grasp on coffee roasting is fledgling at best.

Does anyone have references for learning a bit more about fluid bed roasting? Or am I maybe missing/overlooking some key correlations between standard roasting knowledge and the Ikawa that can help inform decisions on creating/altering profiles for the Ikawa?

GDM528
Posts: 830
Joined: 2 years ago

#180: Post by GDM528 »

Cwilli62 wrote: The issue for me is that I'm still having trouble figuring out how adjustments on any given profile with the Ikawa will alter the end result in the cup. So I'm finding it difficult to confidently "play" with profiles.
Thanks for posting that profile. Could be unique to my circumstances, but I found that large, sudden (few seconds) initial temperature ramps would occasionally confuse the temperature control algorithm into acting as if the temperature was higher than it actually was. It would ramp up but stop 20-50C short of the target temperature and follow the programmed profile - but offset well below the programmed curve. It happened less than 10% of the runs - but ironically more common whenever I tried to impress a guest with my fancy roaster. I've since migrated to ramping no faster than 30 seconds, and even then, I still watch over the roast with the app for every run.

Absent a book on the topic, I found it instructive to deliberately botch a few roasts so I could better recognize the relationship between curve adjustments and taste. Part of Ikawa's small-batch-magic is I only have to suffer through two or three bad shots for my mistakes. I was having a hard time figuring out what to do after tasting two different - but both decent-tasting - roasting experiments.

Tweaking relatively smoothly-curved profiles seems more challenging to me, as the points interact with each other because of that RoR thing. Each of my steadily-declining RoR profiles had to be wholistically created, just to change the drop temperature - not conducive to idly tinkering around. It would be super-handy if Ikawa could add a 'group-shift' function to their app. In the meantime, I use a stepped profile that allows me to change just one or two setpoints and get a demonstrably different roast result.