Ethiopia - Natural Guji & Natural Yirgacheffe Gr. 3



Hi everyone,

New to the IKAWA space and I’m wondering if someone has a roast profile they like for a Ethiopia Natural?
Ideally light roast.

Here is some info about the beans below…


Varietal: Kurume
Alt: 2100 masl
Process: Natural


Varietal: Bourbon, JARC varieties, Local Landraces, Typica
Alt: 1800-1900 masl
Process: Natural

Please share any insight or questions.



'Natural flt ’ freshly roasted by @ikawahome, here’s the recipe

I use the profile for the filter


Thank you for sharing! Any way to share that profile from the IKAWA Home to the IKAWA Pro100?


profile sent is for ikawa home only


‘Natural Ethiopia Medium Roast’ profile - shared #IKAWAPro roast profile @ikawacoffee


You can use this profile on ikawa pro


Thanks so much! I’m going to play around with the settings to make this a lighter roast. If you or someone else has any tips on editing this profile to make it such it would be greatly appreciated !!

Thank you again!


I tried it first by letting it run all the way though, I like how you had the cooling fan much higher, lately that’s what I do (86-90% cooling fan). During the crack it smelled like the blueberries so I tried another roast to stop when the roasted beans smelled like berries. They are resting and I’ll add some tasting notes in a few days.


Nice! @deven.patel411 please share the profile you used after the testing period. I’d love to try it out! Is it more on the lighter end of roast?


Glad to hear that, I look forward to seeing your opinion.


So in my climate here in Florida goes through temperature fluctuations sometimes around 20F/7C, I roasted the Natural FLT on a cold day and hot day and both produced a moisture loss of 15.2% and 15.3% for me, that is on the higher side of where I want my Natural Ethiopians to be (~13%). Because I shoot for fruity/floral notes and not deep milk chocolate. Though 13% for me is that mixture, fruits and chocolate. I do assume that if the roast isn’t perfect ie baked, then the cocoa will but on the hot cup and the cup will cool into berries. If it is a perfect roast then the flavors will swirl on my palate either hot or cold.

Recently also I finally noticed that my probe above the beans was inaccurate (loose) and so it would shift without me realizing it. So I got a new one, and it reminds myself that it is never overstated to revisit the basics, to make sure the new things and the old things are calibrated and if not, where does an adjustment need to be made.

The profile itself is a good starting point for sure, I always felt that these high arching profiles zoom through yellow too quickly but now I will try and alter it so it doesn’t (for me) and want to see if that will also works for you. My thoughts are that: a very fast speed through yellow (which should retain and save moisture for later flavor development) the heat has dried out the flavor so even though a 15% moisture loss would clearly hit milk chocolate it is not, the berries, which is a sign that there was enough time and heat for this berry note. Is still very light in flavor, which makes me think that more time in a lower heat develops the things necessary in order to be produced by/during the first crack. I hope that makes sense for everyone reading this, if it does not please say so.

My initial thinking on how I will alter the profile based on the above, is to shift the charge about 30-45s important tangent I do not think there is anything more important than the specific ROR of each segment, calculating that RoR and using Maths to then determine the hypothesis of what the next segment should be, not only of RoR but also of time. In a way this is all chemistry, we have a specific amount of molecules that react to temp in a clearly defined and repeatable way. The only question I think for us is what beans (based on their physical characteristics like processing, density etc) and apply heat to alter the chemicals in the bean. So like throwing darts at a board, you (or I) say…ok I want to charge at “x” RoR for “y” amount of time to get to temperature “z”. In order to reach a final temp/time range I will try dividing by the golden ratio of the previous segment…and so on. The charge can look like a spike or it can not, but the beginning segment is the charge and whatever that is, in a way will help determine the profile. Within limitations, charges can be too high or too low, yellow takes place in certain ranges based on the profile shape and beans, yellow/brown transition, 1C, moisture loss are all fairly predictable markers for the shared similarities of the beans. How different or similar is based on the bean. tangent over so I will shift the charge, bring down the starting point of the yellow and increase its ROR and then run the profile, and see what alterations of that I should do and if not look at the last segment, roast that and then determine if the overall profile adjustments are good or if I need to restart.


So these beans are really, I don’t know if the word is difficult but it’s always coming so close and then the flavors dip out and it over roasts. I only have about 100g left and I’ll try more with the Natural FLT. But it doesn’t seem to matter since these beans don’t want to be roasted by any profile I throw at it.

The FLT profile seemed to always stall at 300F on my BT probe so I was trying to shorten that stage from the top of the spike to the second point. The green part (when I pushed the 1st point to 1:30, turned bright green and smelled decent so I really wanted to keep that and a yellow with 20F/min ROR smelled like pineapple which lets me know that I am close, to get blueberry I should lower the RoR and lengthen the time but that’s for the Marshal Etheo style, but with this style of profile I kept overshooting and was trying to get a decent crack (not that it is really an indicator for these styles) but none of them really got the berry note I was looking for so the original profile is still the best in that regard. My thinking was to lower the overall heat of the profile so I can push more moisture to the later part of the roast but either my approach is off or these beans are just tough to roast. The profile I roasted today followed a fairly nice declining RoR on the BT probe and smelled decent, cracked well, tasted ok, but now 20min after resting, taste like popcorn. I don’t recommend eating beans but if your going to have some water to swish and spit, eating a couple beans without spitting them out is a lot of caffeine. Interestingly enough I am chewing a bean currently and it just popped a flavor note like this amazing Yemen I have, but it is just getting my hopes up, I know its going to flatten put to popcorn again *sigh. I dono, to me the roaster lately (for a long while) has felt different, back in the day some one said the roaster is nimble like a Ferrari and I disagreed and said it was like a jacked up Ford Fiesta with the 240hp or like an Evo (there is nothing like a Ferrari but a Ferrari)…nimble, but now the roaster feels like a boat, it overly slides from one segment to the next (too much buffering) (and my app is super glitchy when connected to the internet, very annoying) maybe I am being overly sensitive because my coffee sucks lately and I can’t figure it out. :man_shrugging: But whatever the case may be so far it’s always been close but no cigar.

Here is the best attempt so far:
‘EDP NFLT10’ freshly roasted by @deven.patel411 , here’s the recipe

The little dip at the top is from looking at Chris Korman’s profiles on the Royal website, where the have the Crown Jewel analysis. This was a first attempt and it came out ok.


I didn’t even realize this but as I was adjusting, I was getting closer to one of the Ikawa Espresso profiles (the link above). I kept asking why am I baking out this roast, and for this style and Marshal Etheo the answer is shorter and hotter lol. Pic of my adjust profile below.


So six days post roast and the above profile does have more fullness but it is definitively baked which I think came from an 83%flat fan being far too high. Which I should have known, especially for this style of profile. But there was some sweetness popping vaguely in the background. I also think the NFLT and the NFLT3 were too high in the last temp segment, not only for naturals in general, but specifically my particular Ethiopian bean which seems a little less dense than normal and slightly drier. This style of profile I would naturally assume (pun not intended) that we were roasting a washed processed bean, guessing that it would be a Central American bean. However I hypothesized that technically any profile can be used, and adjusted (based on milestones and taste) to a shape within that style, to fit that bean. This is that exercise, though I am out of beans. Since I fell prey to roasting too many attempts without enough patience to taste it in the cup. My apologies. My next attempt would have been this:

‘Natural flt .2’ freshly roasted by @ikawahome, here’s the recipe


14days post roast, these beans smelled good like a *$ like char but in a pleasant way. I tried it a 15g in 135mL 2min steep in an aeropress, swirled not stirred. As per James Hoffman, the opposite of James Bond now that I think if it. Not in a bad way, just seems right. Anyways, 1C pop at 484 6:35, rolling at 494 7:16 and hit 2C at 512 8:19 which is where I dropped it. Moisture loss was 14.7%. Maybe by day 21 I’ll try and pull a shot or two…

‘Sacrifice to the Void #8’ freshly roasted by @ikawahome, here’s the recipe

Though if I had more beans to alter it I would try something like this…


So I roasted the Marshal Etheo O.8 on April 14th, and I finished it off in my aeropress, as per ujus, 115mL with 15g at boiling, 2min steep. It was ok, decent for a dark roast a bit thin in the middle on the aeropress. I wonder if as an espresso it would fill out the mouthfeel more. As I understand the adjustments/temp ranges better and curtailing my overdevelopment before/during the Millard. It’s nice to see how the popcorn notes on a fresh roast turns into a malty milk chocolate. It’s still not good, compared to hitting milk chocolate and fruit notes that plume and waft off a fresh roast, not even close. But it’s closer to drinkable now than before.