Just as a further check on this coffee, roasted according to the IKAWA recipe, I just tasted the Espresso Filter Medium +. The coffee is fine. My charge weight was slightly high at 104.5 g. I pulled my first shot as a guesstimate in my Cafelat Robot of 16 g coffee in, 51 g out, using water just off the boil. I slowly poured extra water and allowed it to overflow as suggested by Sam Law as a simple preheating method. The coffee was ground in a Niche Zero. The shot ran a bit fast, so I backed off pressure.

Here are my tasting impressions: At the start, there's a nice balance between sweet and sour. The next sip as it was cooling had very subtle sweet notes I would describe as strawberry and orange. As it cooled, a tobacco note emerged, followed by a nutty anisette aftertaste that changed to candied orange as it lingered on my tongue. The flavor notes according to the app are orange, nutty and baking spice, so my shot turned up the first two of those flavors. Maybe the baking spice appears in a lighter filter roast than I did. For my next espresso, I'll try 14 g of coffee to get more flavor clarity. Here's the inlet temperature profile on the IKAWA app.



For some background, here's what I've been roasting on since early 2014. If your roast is off, you waste much more coffee, and it doesn't have the guard rails of a manufacturer-provided profile.

Sorry for the trifecta, but I think this comment is worth noting. According to posts by Jim Schulman aka another_jim long ago, using less beans in an air roaster can stall the roast at first crack because the roast goes slower with fewer beans. The OP is using a 75 g charge instead of 100 g. Also, if the roast is stretched, that can cause the baking defect, eliminating the sweetness and adding unwanted papery and other flavors. Here's a link to Jim's post in the FAQ section of this forum. I am quoting Jim because I'm a much less experienced roaster and haven't done much roasting on a fluid bed device like this IKAWA.

The cheapest and simplest way to improve your roasts

Here's another quote of Jim's from that same section of our site. He was writing in 2010.

another_jim wrote:On most pure air roasters, a smaller batch size translates to a slower roast, since with less resistance from the beans, the air flow increases and the temperature drops. Hot Air/Mechanical Agitation roasters, e.g. Z & D, Gene, SC/TO roasters, etc, work like drums in this respect, with reduced batch sizes translating to faster roasts.

The thread this comes from gives some basics about roasting. We've learned a lot since then, but this is a good starting point.

How to Home Roast

drgary wrote:The thread this comes from gives some basics about roasting. We've learned a lot since then, but this is a good starting point.

How to Home Roast

In support of "learned a lot since then":

Thanks to a teardown that was posted a while back Ikawa Home teardowns I can see there's a rpm-aware motor speed controller chip running the fan. So 80% fan speed should be the same rpm for all IKAWA Home units - and very likely all Pro100 units.

The IKAWA Home (and Pro100 set to inlet control) has a fantastically good thermal control loop for the air temperature going into the chamber. You could flummox the system by suddenly changing the room air temperature, but my testing has shown the IKAWA can course correct in about thirty seconds. So, IMO, the IKAWA has come a very long way from earlier air roasters.

Given the above, everyone's IKAWA Home with greens from the same batch should yield the exact same result - and that is what makes the OP's thread so interesting. I suspect this highlights how precision is not a panacea for a tasty cup of joe. I've sampled my own roasting profiles to several coffee afficionados (none had trained palates) and received unique (and often bipolar) tasting notes from each of them.

drgary made a key observation however, regarding charge size. The IKAWA will push a fixed (by the profile) amount of heat energy through the chamber - it doesn't adjust for batch size. I've observed up to a full roasting-level shift from 50g (darker) to 100g (lighter). I don't recall the owner's manual issuing any warnings of that effect - I'm guessing they presume everyone is running 100g.

Can't rule out that the OP might have a miscalibrated unit. One way to test that, is for each owner to run the same batch size of the same profile of the same greens, note when first-crack begins, and compare finish color.

Gary's work is very helpful in measuring temperature in the IKAWA Home and understanding its internals. This will help develop profiling independently. But are the profiles he and others are trying more effective at developing roasts that distinguish flavor nuances than IKAWA's roasting professionals, consultants and customers over a decade?

Enok has asked for roasting principles. I'm suggesting that running an air roast at the same speed as a much larger drum roaster will not necessarily yield the same result in the cup. Also, charge size is important. Roasting plans differ for darker roasts, where it may taste better if bean temperature rate of rise increases somewhat leading into second crack. The start of a roast does well with higher heat than later. Different bean densities and wet versus dry versus anaerobic or other processing require different treatment, as do bean size, varietal, and so on. That dizzying array of variables makes learning to roast from the ground up very challenging.

I believe that IKAWA has done much of that work for us. The profiles they provide are probably designed on their Pro platform and translated for the Home app and device. They're trying to provide controls for getting good roasts and tuning them without needing all of that knowledge at first.

I was surprised to read Jim's observations about air roasters but I trust that they come from experience and a tuned palate. Also, coffees cupped by different people will yield similar observations, whether or not they like what they're sampling. This is especially true with well-trained professionals.

IKAWA may lean away from darker roast profiles in order to avoid excessive smoke and fire risk, but I don't know that and need to test IKAWA's dark roast capability. Added 3/3: It's fine. They have a legacy profile that worked well on my different greens, and Gary offered a profile that took the beans to an Italian roast.

I'm suggesting trying and modifying IKAWA's provided profiles first, because they're developed with some of these complex considerations in mind. I have had several of their coffees roasted with the provided profiles and found that they tasted very good and as described.

At the same time it is useful to understand how their roaster works and how to control it.

If others report that they're not getting sufficiently dark roasts, that's especially worth exploring. I like a variety of roast levels and can produce dark roasts, including Neapolitan style roasts, on my propane roaster. So I will test dark roasts on the IKAWA.

To the OP:
There's no guarantee that the greens you are trying are any good. Any operation can have a hiccup in the supply chain, especially for an organic product (although a famous batch of early GS/3's shows that improper storage can damage anything).

Look at one of the "Greens Alerts" or other roasting threads to find a specific coffee from a specific supplier - one where other people are roasting and getting results that you can compare to. I never developed a sense of, e.g., "washed Colombian". Coffees are individual to me, probably because I only go for coffees that others say are exceptional in one way or another.

Disclaimer: Just my 2 cents. I don't really know what I'm talking about.

Matt

Another note on what Gary wrote above: Smaller batches do seem to create darker roasts. This seems counter to what Jim wrote but he wasn't absolute and instead attributed this to "most" air roasters. The IKAWA may be better sealed than roasters he was writing about and thus more able to contain heat rather than have it dissipate quickly when the heated air vents past fewer beans. Gary would understand the mechanical and computer aspects of the roaster far better than me, and there may be other ways it is affected by a smaller charge than 100 g.

I'm doing more roasting and tasting to get more experience with this roaster. BTW I was able to get good dark roasts with 100 g charges. I'll write more as I learn more.

Also, if the OP is using Brazilian beans for the test roasting, these tend to have low density compared to most other greens. A good place to start may be one of IKAWA's recipes for Brazilian beans. I used one of those recipes to roast a Haitian Blue Mountain coffee, which is also low density, and chewing a bean right off the roast suggested that it was still at a light-medium roast level but well developed. I look forward to trying that very mild, low acid coffee when it's had more time to rest.

After more roasting today with smaller charge weights, the roast temperature does start to decline from the programmed temperature curve. But the roaster's software is smart enough to notice this and adapt, bringing temperature back to the pre-programmed curve. It will take a few days before I taste the smaller charge roasts and compare them to the 100 g charge using the same recipe to see whether the smaller charges result in a darker roast.

Having spent some time with a Home demo unit and doing 2 - 3 dozen roasts, kudos to Gary aka GDM528. He knows this roaster's thermal performance very, very well.