IKAWA Home - profiles - Page 14

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
jakline
Posts: 92
Joined: 18 years ago

#131: Post by jakline »

mgrayson wrote:In the meantime, look at the Pro cleaning instructions and, if necessary, order their cleaning kit which, while it has no Urnex products, DOES have felt pads and an industrial toothbrush.
I have a Pro 100 and figured I would get this kit. Two extra felt discs (I already have one) and a brush.

Well, despite giving the serial number of my Pro 100 roaster, they sent me much smaller felt discs meant for the V3. As such, I paid $20 for a brush. I've asked them to fix the description on the kit so others don't make the same mistake.

mgrayson
Supporter ♡
Posts: 645
Joined: 17 years ago

#132: Post by mgrayson replying to jakline »

I made the same mistake and had the same discussion with them. I hope they fix the description and/or make a kit with larger disks.

GDM528
Posts: 837
Joined: 2 years ago

#133: Post by GDM528 »

Here's another roast profile that's intended to add more variety to the taste profiles from the same bag of greens I've been working through - not to replace my previous roast profile (post #117), rather to mix things up a bit. Also an excuse to try some stupid-Ikawa-tricks just 'cause it can ;)

Strategy:
1) Bold beginning / calm interlude / brutal finish. Beans exit with dissociative flavor syndrome.
2) Rapid initial ramp to get the greens into the browning stage with maximized moisture content, aka 'miasma phase'.
3) Hold at a moderate temperature to gently brown the beans without overcooking/burning them.
4) Just as the beans think they're going to get out of this alive, take them where an Italian roast goes to die.



The rapid initial ramp is intended to emulate the pre-heat/bean-drop thing the adults do, and based on some inferences from the curve shapes I think the beans are a hot-steaming mess for the first two minutes of maillarding. After six minutes the beans reach 180C and have a dark caramel color.

The last two minutes are intended to emulate a medium-rare steak prepared with a sous vide and a blowtorch. Get to an attractive color, add some dark roast notes, then GTFO. Resulting roast looks identical to a more conventional Full City roast: 10% of beans show oil patches, and the overall roast has a sexy auburn sheen.

Earlier attempts expanded my tasting lexicon to include "grassfire". Holding the Maillard temperature at 160C and finishing with the air temperature at 290C evoked imagery of a farmer's attempt to burn off the detritus of their field, thwarted by a flash rain shower. The two worst-case polar-opposite extreme tasting notes concurrently in the same cup... Failed hard, left a huge crater - you've been warned.

Bumping up the Maillard temperature and backing down from the highest-possible peak temperature setpoint on the Ikawa significantly moderated the light/dark tasting notes. Also probably not a coincidence that the area-under-curve is within 2% of my previous roast:



Mission accomplished on the taste goals. Sorta like a medium roast in the same manner of a yellow light from green and red bulbs. Weird trickery. Even though the chamber temperature was 10C higher and the final bean-mass temperature was 5C higher, it doesn't have the dark intensity of my 'conventional' roast: less bitter, brighter, but not acidic - with a hollow-middleness that I'm not familiar with. Follow-on experiments will attempt to spread the joy, but I only have six temperature setpoints to work with.

Here's a link to the profile. Editable with the free app, but note that just a fifteen-second shift in the last two minutes can significantly change the (apparent) finished roast level, plus there's that whole 'grassfire' thing:
https://share.ikawa.support/profile_hom ... uZWkxYgEw

GDM528
Posts: 837
Joined: 2 years ago

#134: Post by GDM528 »

mgrayson wrote:PSA:

Something I noticed today - cleaning the roaster is important! I've been getting an ashy taste to even medium light roasts. I turned up the fans, and that helped a bit. Turns out I've roasted about twice the number of batches recommended between cleanings. First roast post-cleaning smelled much better. I'll go back and try the worst offender - Brazil Sweet Blue Daterra - and report back.

Matt
I happened to try cleaning the chamber of my Ikawa Home about 15 minutes after using it for a roasting session - it was still warm but not hot. Using a standard-issue 70% alcohol wipe (moist, but not drippy) I was able to clean out the interior residue surprisingly easily. Just took a couple wipes and some long tweezers to push the wipe into some of the nooks my fingers couldn't reach.

Granted, I'm not using the Ikawa-approved tools and methods, so you can ignore that part and just note that warming up the residue makes it much easier to clean.

GDM528
Posts: 837
Joined: 2 years ago

#135: Post by GDM528 »

Made some adjustments to the "Spike" profile posted in #133. Now crazier than ever, and now my favorite roast profile yet.

How I treat the roast matters. I discovered this when an inadvertent too-fine grind of this roast led to a 60-second shot that was actually quite nice, sweet with just a hint of bitterness and dark-roast notes. I also pull somewhat past 2:1, to e:1, where e=2.718 , aka Euler's number, aka 'naturale' (it's a math-nerd thing) So I'm learning not to judge a roast until I've adapted my brewing technique to it, which may involve deviating from canon.

The green trace in the chart is the spinning bean-mass temperature. I included the empty-chamber curve (blue) to show how the roast profile inflects when the bean temp reaches the chamber temp. RoR actually goes slightly negative just before I open the trapdoor to Hades.



Most epic change came from increasing the mid-roast temperature to 180C. From what I've read, the sugars in the coffee start caramelizing at 170C - but slowly, hence an extra 10C to speed things up. Less significant was a tweak to the blast-furnace phase to increase the gradient - plus it looks even more violent now!

Speaking of gradients, what does the bean look like after being torched for 2 minutes? Better than my more conventional roasts, with virtually no scorching, tipping, cratering, etc. Perhaps it takes longer than two minutes to incinerate a roast. Overall external roast color looks nearly identical to a more conventional Full CIty roast, including a small percentage of the beans showing oil spots - but this does not taste like a Full City roast; it's more like a post-roast blend of light and dark roasts.

I was curious if the inside of the bean might reflect the sudden temperature changes, so I carefully cracked a bean in half and put it under a microscope. Sorry for the blurriness, but let's see you try and crack a roasted bean in half such that the exposed face is completely flat...



At first glance it looks like outer part of the bean is actually lighter, but if you squint you can see a very thin darker skin. Still interesting how the interior of the bean is darker than the outer layers, as if there was more than one roast going on in there.

Here's the recipe if you'd like to try this experiment yourself. Takes just 50g to give it a go, and the beans won't see it coming: https://share.ikawa.support/profile_hom ... uZWkxYgEw

Pretty sure this profile would rain chaos upon the Pro models, as the sudden temperature changes would throw an ET-controlled machine into apoplectic spasms. Apologies for that, but I think there is a way to set a Pro V3 model to inlet control...

User avatar
Peppersass
Posts: 3690
Joined: 15 years ago

#136: Post by Peppersass »

GDM528 wrote:Pretty sure this profile would rain chaos upon the Pro models, as the sudden temperature changes would throw an ET-controlled machine into apoplectic spasms. Apologies for that, but I think there is a way to set a Pro V3 model to inlet control...
Indeed, there's a way to create inlet profiles for the Pro models. All you have to do is enter the Edit mode, hit the exhaust/inlet icon to select inlet, then enter the time and temperature points. If you save the profile while inlet is selected, iIt'll be marked as an inlet profiles. When an inlet profile is loaded, the Pro PID reads the inlet probe instead of the exhaust probe.

(I've tried the edit steps, but haven't actually tried to create an inlet profile to run on the Pro...)

mgrayson
Supporter ♡
Posts: 645
Joined: 17 years ago

#137: Post by mgrayson »

Th pro models take inlet temperature profiles. My go-to is such a one (from royal coffee). It generates a perfectly decreasing RoR which flattens out at about 4C/min near the end. I changed the fan speeds a bit to handle 80g loads, which is what I usually use. This is the Colombia Aguazul Pink Bourbon.



Matt

GDM528
Posts: 837
Joined: 2 years ago

#138: Post by GDM528 »

mgrayson wrote:Th pro models take inlet temperature profiles. My go-to is such a one (from royal coffee). It generates a perfectly decreasing RoR which flattens out at about 4C/min near the end.
Why does your inlet temperature dip at the one-minute mark - is it programmed to do that? Bean drop is T=0 on the graph?

User avatar
MNate
Posts: 957
Joined: 8 years ago

#139: Post by MNate »

mgrayson wrote:Th pro models take inlet temperature profiles. My go-to is such a one (from royal coffee). It generates a perfectly decreasing RoR which flattens out at about 4C/min near the end. I changed the fan speeds a bit to handle 80g loads, which is what I usually use. This is the Colombia Aguazul Pink Bourbon.

image

Matt
Are the Pro profiles shareable with the Home? Why don't I know this...? If so, and you don't mind, I'd love to try out this profile as I have that bean as well (or similar, I think I got mine from Burman). I have settled on the M+++ for it, but know there is room for improvement.

GDM528
Posts: 837
Joined: 2 years ago

#140: Post by GDM528 replying to MNate »

I've installed the Pro app (sans machine), which allows me to browse profiles in tabular format, from which I transcribe by hand into a Home profile.

Yeah, kinda clunky process, and it gets really complicated if the Pro profile has much more than six setpoints.