Favorite profile? Marshall Etheo? Spikes?



Hi everyone,

Got my Ikawa home yesterday and started roasting.
Roasted the 50g Burundi it came with, and also 2 different batches for Etheopian Aricha Naturals I had which has strong notes of berries.
Apologies if my questions are random and noob.

1) What are your favorite profiles or profile styles?
I tried the marshall etheo yesterday (the version which goes up to 290c)
Dropped the beans around 5:30~5:50 and looked really nice for espresso.

It seems like many people like the marshall etheo or its variations, but what are your favorites?
Does it really make a difference when it brings the air temp to 290c? (hitting first crack around 265c) Because all the other profiles I see only goes up to 250c ish, and with that it seems the first crack is very gentle, hard to notice.

2) The spike and difference it brings
I see all of the original Ikawa profiles have spikes in the beginning, and Marshall Etheo does not have that. If anyone had tested same profiles with and without the spike in the beginning, curious to know what difference it brings in result and taste, and in what cases they’re beneficial.
Seems like it acts similar to most roasters charging hot, but does it really work as intended, and make difference in cup?

3) Roast time (5 min? 10 min?)
The marshall etheo I tried finished roasting around 5:30, which seems really quick and some other profiles I see take it to 9~10 mins. What’s the difference in roasting really fast and slow, and what advantages each of them have?


Hey welcome! I am also a noob to ikawa, and have had similar questions, looking forward to the answers! I will add one more.

Why do some profiles raise temperature all the way to the end, and some start to fall?

  1. I’d say my favorite profile is based on many factors like the bean itself; what variety/varietal it is, the processing method used, the brewing method… these factors and more determine what will be the best profile for the bean, suited to your taste preferences. I like the Marshal Etheo profile style for naturals, specifically Ethiopians. Can you post the profile that goes to 290c? For the profiles that go to 250c I think these are best suited for an Aeropress or pour over style, because they retain a lot of acidity and in an espresso, it can be overpowering. Going hotter can reduce the sharp acidity and also bring out some flavors better suited for espresso found in profiles with more development. But the issue is complicated and can vary based on what you want to do with the coffee.

  2. The spike is a good way to add energy into the bean early on. In my experience it gives the extra push a profile might need to produce the flavors of the profile/bean. To be more specific, if you have a profile that goes light you’d get things like acidity, developing that acidity turns into juicy-ness, developing that juicy-ness will develop into sugar like notes, either texture wise or flavor wise. Like sugars or syrups, molasses moving past these we get into secondary flavor formations and resting allows for tertiary flavor formations. Sometimes a profile is great when it really makes the beans pop with flavor, right out of the roaster. Sometimes flavors will pop out after resting the resting allows for those tertiary flavors to form. I haven’t had much experience roasting dark and besides some caramel notes, haven’t had a profile that really popped out (flavor-wise) like Marshal Etheo did for an Ethiopian DP. I don’t doubt that a profile style like Marshal Etheo can have a complex palate like the over arching profiles that spike and start around 200c and finish in the 240ish range. But for the most part, in my experience, the Marshal Etheo profile gets a really nice resounding flavor when the profile is altered to the bean so a decent cracking is heard, I tend to drop in the 5-15s range. But I am re-evaluating this because since the updates and bean changes from season to season and a roaster repair, I need to reevaluate things things like what is a good fan speed, ROR speed through the roast, and length of profile. This is a key reason why taking notes/observations/thoughts, when you roast is a very good idea.
    How I determine how I adjust the fan: is by how the beans move in the beginning and what is the shape of the profile, if the beans taste burn or overshoot the flavors, I tend to raise the fan higher at the end. Anything over 80 is going to make a 13% moisture loss bean, shoot out before cooling. I also think about how I want heat to be applied to the beans, and this is the fan’s job. It determines how much space are between the beans. Leading to how the beans are exchanging their cool moisture removal with the hot surrounding air. How compressed the beans are against the wall of the chamber also effects the heat exchange with the system and the beans. And that is directly linked to a higher fan so it is inevitable that a higher fan is going to heat the beans faster, but doesn’t mean they are going to necessarily loose more moisture sooner because his usually ends up leading to a louder crack which is just another variable that has to be adjusted for. This is usually automatically taken care of in higher temp profiles because the steam released is transferring energy (like somebody wobbling on the tight rop), the higher sustained heat or a higher fan can help reduce the bumps that the bean feel and hopefully not effect the rate of chemical reactions taking place.
    ROR: is based on color and crack markers, those are usually very consistent and will happen at specific temperatures and not about length of time, when it comes to Marshal Etheo style profiles. For those profile types I like adjust them by altering how much time I want to spend in green/yellow/brown and when I want to hit 1C without blowing though the bean or under-developing the bean. I previously stuck to around 15F/min or around that in the final segment. But higher temp profiles disregard this as they blow through yellow and are usually brown in 2-3 minutes. The duration after is much longer and the ROR is much flatter throughout the profile. Which is not a bad thing, just different. I have less experience working and altering those profiles so I can’t really give a good assessment. So far I just picture the bean temp and use brewing to asses if I should go longer or shorter/hotter or colder on the drop temp, etc.

  3. Is just a variable that is part of and reliant on all the other variables. It’s like a rocket ship with so much fuel, trying to shoot into outer space. Except we are trying to apply heat to loose moisture at just the right rate, without finishing the beans with no protection so they burn up and loose all the flavor. So more like flying a plane while at the brink of being out of fuel and unsure where you can land at the same time.

I could be wrong, but I view the issue also can be found in the answer above but also simply: Some beans (for some reason idk) loose moisture faster and “run away” so it is easier to crack them at a higher temperature and drop the heat as the bean come to finish. Technically the beans are still rising in temperature during the time. But the heat that needs to be applied to them needs to be reduced so they don’t heat up too much when coming to the optimal zone to finish and stop the roast, or else they will taste burnt or dead.


This is the Marshall Etheo i found here and tried which goes into 1c in 4:30 ish and finished roasting around 5:30.

The Marshall Etheo you are mentioning is the MELFS3.4?


Does “light” here mean with or without the spike? Or do you mean light roast in general?
Was curious what effect the spike, “applying more energy in the beginning” bring to the whole roasting process.


I meant to say light in regards to roast color of the bean, outside color of the bean isn’t necessarily correlated to the beans total development so alterations of the fan and heat length can develop a bean more than what the outside color would usually indicate or relate to from a different roasting machine.

The spike is like charging hotter in a gas drum, both a non-spike and spike profile still are being “charged” in the first minute to minute and thirty seconds of the roast. But the spike is adding in energy into the beginning of the roast, allowing for a lower overall finishing temp. If a profile used tastes a little underdeveloped in the cup, a 10F/3-4C can help develop those underdeveloped flavors. Maybe you will have to increase the temp more or less, this is something you’ll have to test and taste the results to know.

Too hot of a spike and too hot of an overall profile can lead to tipping (burnt flavor) which will effect the tasting results of some burned and could lead to baked out flavors. Which could be good (or terrible), because if I bake the roast a little, the pastry or wafer like flavor into the roast, it wouldn’t be a terrible cup to drink, think strawberry pop tart or chocolate pastry. If I were to spike within a minute, versus extending the tip of the spike to around 2 or 3 minutes, this will effect how much heat is applied to the bean. And also how much pressure the heat is applying to the bean, the hot air is one force pressing against the bean, while the beans defense is the moisture it is releasing to try and stay cool and protect itself. If I alter the pressure of heat applied or increase fan (an increase a lower pressure environment) I can weaken the counter force of moisture loss (by altering via lowering the evaporation temp of water) coming from the bean, so I can push heat deeper into the bean at a faster rate. Convection is a tricky business because if you lower the pressure and increase the speed of air around the object you can heat the inside much faster than the outside of the bean. Resulting in a much faster roast with a much lighter outside.

We can alter the chemical/flavor results at the end because I am compressing or lengthen the window for the reactants to react how much moisture is left in each segment will change the flavor is by altering heat in certain segments because then I am limiting or creating more reagents to react and produce or reduce more or different flavors. The spike changes how much time I will need to roast and thus alter the window I can roast in to produce the flavors I want. A longer spike will remove more moisture, dig heat deeper into the bean and alter when you want yellow to end and also alter how long brown to development to last and at what rate. I am being vague because there are hundreds, maybe thousands of different coffee greens that you could potential roast that differ in it’s ‘origin’ season/growth, variety, and processing, etc. that will alter how much time or how much heat is needed. A first crack temp can alter by a couple degrees up or down, for the same bean from the same farm but from a different year, or it could hit at exactly the same temp from last year, but you will have to be aware and know this info so you can make adjustments from year to year, or adjust from bean to bean. The further a bean varies from the original bean the more the profile will vary. (That last is an assumption that is slowly becoming a fact in my eyes) There are many variations and I only roast for my own consumption and learning, I don’t roast purely to test theories so I am moving slower than if I were to be running a lab doing profile experiments and cupping all day, so my understanding is though my own experimentation and reading what others have wrote.

This is the closest version of the Marshal Etheo profile that I have, lost and recovered from memory. This profile has a cooling fan of 85% which you might want to change. I remember vaguely that originally 1C happened around 9:30ish but I’ll have to dig through some notes for two years ago.

‘Marshal etheo 1fan80%’ freshly roasted by @ikawahome, here’s the recipe


Wow, thank you so much for all the info. This is a lot to digest and try out. :+1: