Perfect Ehiopia Natural Roast and My New Favorite All-Around Profile


I’ve done hundreds of roasts on the Ikawa Home since my last post! Looks like I may not be able to take advantage of that one year warranty ;). I’ve been waiting for something worth posting, and this is it. This profile produced the best cup out of about 50 different profiles attempted with this Hambela Hassan. It is such an incredible coffee (the highest scoring coffee I’ve seen at Sweet Maria’s in about 14 months - a 93.6), with so many radically different flavor possibilities, layers, and nuances, that even my more experimental attempts (4 min. roast, 11 min. Kochere Debo style, 11 min. Andino style) all came out above average for specialty coffee in my opinion. This one, however, ranks among the top ten cups of coffee I’ve ever had from any roaster.
The Andino style approaches proved to be my favorite as they often do (although shakiso made a great full city roast with lots of popping fruit and dark chocolate). This Hambela lines up a little differently with the Andino style profile than other beans typically do. For example, most other coffees show more signs of paling on the initial spike, often right at the highest point at about 1:10, and initial yellowing starts shortly thereafter with distinct yellowing effects often noticeable around 1:30. The Hambela, however, starts to pale slightly at about 1:35 and show the first distinct signs of yellowing as late as 2. All traces of green are gone at about 3:55, and the appearance of maillaird smells and coloration becomes distinct at about 4:45. FC happens shortly after the rising inlet temperate tapers off - around 7:05. The batch crack will typically have ended by about 8:40, but the end point extends another 30 seconds and catches a lot of straggler pops. I have tried about 30 variations of this overall approach, and they all make a big difference - for instance lowering the end airflow point by one percentage point produced a more developed cup that lacked the brilliant acidity of this one. Deleting point two on the temperature plot had an even more drastic effect on development. The temperature and time settings for points 3 and 4 do a lot to change the progression of yellowing and maillard as well as acid breakdown. Of course the development phase is very finnicky too - raising point 5 by two degrees Fahrenheit resulted in significantly more development in flavor. This roast is designed for a full doser batch in a slightly warmed machine.

Since getting such great results from the Ethiopian bean, I tried it on my gesha (from earlier posts). It is resting now, but based on the aroma from over a hundred batches of gesha, it will be one of the best. For the gesha I stoped it at 8:54.

Myanmar natural profile?

Woah, nice! That’s what I’m talkin’ 'bout!!! I’m getting in another round of beans, should be here on the 22nd, this post has got me pumped I’m gona try it for sure!


So good - well done! I’ve jus received a nice shipment of speciality GBs and will give it a try. What would be really cool is if we could each try the same beans. Maybe we start something here?? Like @deven.patel411, I’m using 3rd wave water so we could really compare and have something to talk about.


Yea the idea has been floated across a couple of times but hasn’t taken hold yet. I would really need like to but I forgot to say anything before I ordered this time around. Usually someone has to keep an eye out for a good bean, and tell everyone here so we can all purchase soon because (especially at SM) where I lost out on my pound of Sumatra, because of delaying my purchase.

I’ll list my recent purchase I got today in a thread. Listed here:Recent Greens Purchase


Update: better cups produced! This profile in particular is promising, although the roast is very fresh so I don’t want to guarantee anything yet. My original profile in this post ended up creating some baked flavors after four days of rest or so, and also when roasted in different conditions (the weather here in Connecticut has been changing wildly throughout the course of this process). So, anyway, I think this approach offers some good solutions - the higher charge temp., earlier pt. 4, that exact fan speed, and that exact end temperature of the development phase all seemed fairly critical. Although there seems to be some wiggle room with most of the points as well, which I see as a good sign. My earlier attempts were exactly the same, but with pt. 5 extended out to about 10:10 (same temp.). It was a fantastic cup of coffee, retaining nearly all of the origin character but with a medium roast body and a deep lasting sweetness that stuck to my tongue long after each sip. Today I tried shortening it to where you see on the picture below, and, well, it smells fantastic! I know smell can be misleading, but, hey I wanted to post again and to let people know that I’m pretty confident in the changes I’ve mentioned above, except maybe the 9:15 stopping point - if you want to use the profile now and be guaranteed a more drinkable cup (less chance of underdevelopment), I’d try extending pt. 5 out a bit but not raising it. I’ll post again with tasting notes.


This profile is lacking a link to download into the app. The earlier version of the profile can be downloaded. Something must have gone wrong when you uploaded this one. If you can fix it, I’d be grateful. Thanks.


@jboutte88, I did a single roast to try it out. I only have 2lbs so this time around I’ll use about 250g to test around your profile. The pros so far was that I was getting some clearer flavor notes but not ones I was aiming for, a lot more in the realm of dry cacao and molasses, which were on the bag for tasting notes :+1::+1:. A slight hint of fruit but didn’t have the voluptuous, curvy mouth feel I like in my wom…coffee…there were a couple cons, nothing terrible the two cups I made really were very drinkable, but I think through my own experimentations I have been put off by molasses(so not any fault really). But I am sensitive to burnt taste, in the toasting/roasting of nuts and cookies and such. And there was some, certainly not overpowering but present.
I will try and work with the next five roasts and finding something more akin to my personal preference and write something up.

Currently I have 12 batches of Costa Rican (Brazil2 7(v.2) is looking promising )going so hopefully I can do something soon.


New favorite variation. Hits all the notes and really tatstes of them too. Sweet and delicious!


You know what, I got home early. And I fancy a roast, I think I’ll try it. I mean it isn’t the whole point of the roaster :smile::smile:. I will report back findings maybe I’ll find a gold nugget in my cup :wink:.


Did three roasts, if you want the details of the roast I can list them. Just a short summary; I thought the profile was magnificent. I drank two cups both of the Shakiso one stopped just before first crack (based on smell) and the second after the end of rolling. Definite fruit, some berry. If I do a hard slurp insane berry/mango. Some interworkings of various meats. Bacon on the grind. Slight shrimp(?) I know that sounds weird but it’s ok. The different flavors keep rolling in my mouth and it’s kind of hard to let the cup cool. Slight tannin dryness but still gushy on the sides. On the hot cup the pre1C was creamy and fruity, tomato but super smooth, like a good light tomato soup (not creamy). I am actually very happy how this roaster can produce juicy cups like this. As this rolling cup keeps chugging along the room temp cup keeps swirling in different flavors, some slight cracker more like biscuit note keeps passing over. Overall very nice profile!


I’ve just tried your profile on some natural beans from Guji, Ethiopia, which I bought from Has Bean in England. I bought 250g roasted and 250g green. Yesterday, I had a V60 cup of the lightly roasted Has Bean ones (a week post-roast). It was really delicious. Today’s roast looks similar to Has Bean’s own roast in colour. I’ll give them a day or two de-gas and then brew them.


With 5 days rest, the whole bean smells like sandalwood, the ground bean smells like a dry sweet fruit and beef Jerky. Wet beans smel very good, hits/droplets of some full sweetness. The real true note that is the entry free into the realm of go(o)d coffees…fruit/flavor on the hot cup. Some sort of fruit berry on be tongue. This is an authentic 9/10 pt profile and in my opinion I’d def rate this in the top tier of coffees I’ve drank off the Ikawa. A nice profile for sure. For the contrast: the coffee lacks definition of flavor with indistict flavor notes, this was a hint of astringency, some biting in the back of the throat but it is fading. This could definitely be fuller. A hard slurp on the room temp cup unveils a nice coconut/lime that is creamy in flavor but not texture. Truly an excellent profile that I can tune to my particular bean. I have roughly a pound left that I will spit betwixt the Hassan and this other profile listed below to fine tune with.


To clarify what I wrote above was for the Hassan profile, but after 10days or so I cannot say that all the goodness carried over for my aeropressed cup. Unfortunately I thought the tomato in the Hassan was going to go away but it didn’t. I did have that “origin flavor” which to me, tastes like potential for blueberry. But even after rest this flower did not open up. Maybe the profile is better for espresso users but I cannot relate my experience to that.

The profile MELFS3.4 (v.5) dropped at 8:30 (a bit too late imo), listed above, did turn into out quite well (aeropress 197*F in 143mL) I’ve been pondering for a while because of my own bias but really the cup is how I like it, and I am the last one to give myself credit. Full, rounded flavors (I do get some notes from the label, specifically: passionfruit) that are distinct, although it does require some slight tuning I feel like I’m in the ballpark with the MELFS3.4 profile. Though I’m unsure how to alter it (lol).

Also on a side note has anybody noticed that if you roast a sweet potato plain in the oven it somehow has a reminiscence of blueberry?


I’ve roasted quite a number of African naturals now from Ethiopia, Kenya, Rwanda, Tanzania and still like the Marshall Etheo style profile. Never had anything but consistently good results from this.

I’ve had some good results from other profiles but not consistent across different coffees. So I always use this as a reference when roasting naturals.

Not really a fan of the initial spike on naturals. I prefer a slower steadier start with these.


@jboutte88 I ran your Hassan U3 profile on a natural Guji ( Ana Sora from HasBean in the UK - a 90pt coffee)

Brewed in the V60 12 days post roast, it is very good. Sweet, sugary, light and juicy with a good blueberry/strawberry fruit and a mild acidity.

I got 1C at 7.01 but just a solitary crack and stopped at 8:30. Probably could have done with running to the end of the profile. I have enough to run another batch and a batch of @deven.patel411 suggested tapered Etheo.

The standard Marshall Etheo profile gives more of a blueberry/lemon flavour which is a closer match to the roaster notes. Possibly a slight preference for this one. Hard to call which is a nice problem!

A very good profile and thanks for sharing.

I note points 2 and 3 are very close together with a short acceleration between them. What was your rationale with this part ( to complete yellowing?)


Hi, everyone,
I’ve been reading and absorbing, trying to figure out a common sense of roasting. It’s great to have a forum where home roasters can share their experience with each other. At the same time, I keep trying out all the profiles you guys share here. There’s one thing I found it hard to ignore. As the above picture shown, is it normal for the Costa Rican to look “pale and wrinkled” on profile MELF 3.4? The beans dropped at 8’42 which is the set end point. I didn’t hear any obvious noise of FC.
Oh, btw, the fan of my machine stopped working at the end blowing process. Should I be worried ?


Both on Costa Rican beans;
Both dropped at 08:42
Neither has obvious FC popping sound as I’ve listened to as hard as I could.


It can be normal for some beans to not have a very obvious 1C but that has been noticed more in Sumatrans and not Costa Rican’s but with each profile profile, applying enough energy to be built up in the beans will cause audible cracks, not hard to hear. The pale and wrinkled is “normal” (in some cases) going for very light roasts. With not a lot of outer bean development vs the inner bean. You can see this by cutting open a bean (prefereablly with a sharp knife to make a clean cut) so you can see how dark the inside go vs the outside and you’ll also see where the outter and inner bean color meet. Ideally the color difference depends on what you are trying to accomplish whether that is a wider contrast or completely even.

The profiles made were for Ethiopia region, the Hassan was for Dry Processed and the MELFS was for Wet Processed. I assume the layers of similarity goes from Variety -> Region -> Processing, so the more similar the less profile alterations needed. Even the same bean (I have seen) from the same farm year to year will require some tweaking of the profile. Making judgement calls like the beans aren’t as developed or tasting as you like, or not hearing 1C and how it tastes in the cup are results that indicate how you should alter the profile. The Hassan will always make beans look more developed than the MELFS because the area under the profile’s curve is how much energy is being put in the bean and the Hassan has a more area meaning more energy.

Ikawa from what I understand from previous posts, is aware of the problem and working on it.


I want to comment on the color, but it’s hard to know if I’ll be right. Color is useful, but only after you’ve tasted every color of the spectrum for every bean. If those taste astringent or really aggressive on the tip of the tongue, it’s probably underdeveloped. Based on my experience, I wouldn’t be surprised if the hassan u3 is baked.

Crack is probably a more useful factor than color for knowing that you’re doing the right things chemically. It doesn’t need to be that loud or make an audible sound for every bean to come out great. You can have a crack that sounds merely like whooshing and puffing sounds of gas escaping and have amazingly juicy results in my experience. Lately I have been aiming for a first crack that consists of a little bit of that typical popping, cracking sound but mostly much quieter cracks. What I end up looking for more than color is size and shape. I’d like to see quite a lot of expansion without too much forceful rupturing of the seam; in other words, i want the beans to get much bigger, but also not change shape that much. The face of the beans shouldn’t have become too convex for example.

I’m no longer a fan of the Andino style profiles including the ones from this post - they are mostly “baked” I believe. they’ll taste great for a few days and then lose flavor, or just be slightly less sweet and flavorful than they should be. I tend to notice a “cardboardy” thing with a lot of coffees roasted this way. Interestingly I noticed it on the Andino Colombian bean itself when the bean was older but not when I first got it. Maybe the Andino approach makes sense for fresher coffees but that’s pure speculation. Lately I’ve been using profiles more like the Colombian I posted recently. I’ll post some variations in a second - you might find some promising ones.