IKAWA Home Roasting System Review - Page 2

Behind the scenes of the site's projects and equipment reviews.
User avatar
Team HB

#11: Post by baldheadracing »

Baldheadracing's IKAWA Home Roaster Review

This review will be a bit dis-jointed as pretty well everything about the IKAWA Home Roaster has been covered in the main review by Gary. One item Gary did not cover was the Android app, so I have made some comments about using the Android app. Click to skip to the details.

The rest of my review covers my experience with the IKAWA Home Roaster. The review includes:
1. What I would use the roaster for - as a smaller-capacity roaster to complement my 'big' roaster;
2. Roasting with the IKAWA versus the iRoast2: Given my tastes, the IKAWA scores a narrow win;
3. Back-to-back roasting with the IKAWA Home Roaster: Yes, it can;
4. My favourite part of roasting with the IKAWA: Smelling sensory milestones; and
5. Should you buy, and some minor reviewer nitpicks.


1. What I would use the roaster for

I think that the market for the IKAWA Home Roaster can be divided into three overlapping groups:

The first group would be people who are not currently roasting. They most likely do not have a huge interest in learning the craft of roasting, but they would like the convenience and delight of freshly roasted coffee in their homes. They would like guidance on the "best" way to roast a particular green coffee, and be able to chose green coffee from a wide variety of quality curated beans. They might also like to try making adjustments to that guidance. This first group should very much benefit from using the Apple iOS app.

The second group would be people who would want the IKAWA Pro/ProX Roaster but do not want to spend that much money for a home roaster. They might be interested in modifying their Home Roaster to see something representing bean temperature. They would probably use the Home Roaster as their do-everything roaster, as the roaster can handle back-to-back roasting. This group might well prefer the Android app to avoid the monthly subscription fee for the Graph Editor in the Apple iOS app.

The third group would be people who would use the IKAWA Home Roaster as a complementary roaster to their larger roasters. For example, 300 g is the mimimum charge for my one kilogram roaster. Thus, I have to use another roaster for smaller charges, for example, for sample roasting. I wanted to see how the Home Roaster would do in this role - as delivered. I did not use the Graph Editor due to the monthly access cost for Apple iOS users.

2. Roasting with the IKAWA versus the iRoast2

The roaster that I currently use for smaller charges is the long-discontinued iRoast2. The fluid-bed iRoast2 had a capacity of 75g-150g. It was somewhat programmable, and had the reputation for producing an even roast, both within and between beans. The reputation seems due to the patented roast chamber design, which has not been correctly copied in the clones of the iRoast2 that are currently on the market.

For comparison, I used both roasters in push-button mode. As IKAWA's sample green coffees were 300g, three 100g roasts were blindly compared: the recommended espresso roast in the IKAWA, the recommended filter roast in the IKAWA, and a roast in the iRoast2 that was aimed to try to get the same roast colour as the IKAWA filter roast.

There has been renewed interest recently in roast colour. The current trend is perhaps most attributable to Morten Münchow's statement:
"Test consistency with your roaster and explore the different scenarios that could lead to inconsistency," he recommends. "The key parameter for consistency is flavour, which is 80% driven by variation in colour and 20% driven by variation in timing."
- source: https://perfectdailygrind.com/2021/09/h ... ng-coffee/

Thus, if Morten Münchow's 80% proportion is reasonable, then reasonably-done roasts with the same roast colour are going to taste much more alike than different. In comparing roasts of similar colours between the Home Roaster and the iRoast2, the primary and probably valid conclusion was that I could not reliably discern between roasts. However, there was a pattern of the best IKAWA roasts tasting better than the iRoast2 roasts so the IKAWA Home Roaster wins. Click to skip to the details.

3. Back-to-back roasting with the IKAWA Home Roaster

Some experienced roasters will know that it can be frustratingly difficult to get a traditional drum roaster to produce identical back-to-back roasts. The IKAWA claims to solve that problem. I ran six roasts consecutively alternating between espresso and filter profiles. I could blindly distinguish between espresso and filter cups, but I could not taste differences within the three espresso roasts or the three filter roasts. Click to skip to the details.

4. My favourite part of roasting with the IKAWA
My favourite part of using the IKAWA Home Roaster was that I could easily smell the "Sensory Milestones," as popularized by Willem Boot. I stood over the roaster and smelled the exhaust rising from the Home Roaster. I could smell all of the milestones, with one following another. This is something that I found difficult to do in a small drum roaster, as constantly pulling the trier also affected the roast.

Below are descriptions of the milestones of smell from my notes of Willem Boot's course video. Note that each milestone smell may only last a few seconds; coming in slowly, then strongly, and then fading away into the next stage. (The temperatures below are from the course and are for a typical conventional gas drum roaster, not the IKAWA Home Roaster.)

Freshly-cut grass or wet grass: 200F-240F
- water in the beans is changing from liquid to gas; moisture is evaporating at a fast rate.

Hay: 290F-320F
- the Maillard reaction starts. There is a savoury smell. This is an indication that sugars are being caramelized.

Baking bread stage: 330F-355F
- the fragrance becomes even more savoury. There is not yet the smell of coffee. Almost all free moisture has been evaporated and the beans have a light cinnamon brownish colour

- the Aroma-point may happen just before the start of first crack. For the first time, the fragrance/aroma smells like coffee - but only for 15-20 seconds.You may not smell anything like the A-point again until you cup.

First crack
- First crack is a consequence of prior events; it is not a milestone in and of itself. The speed of first crack may be an indicator of roast quality; a fast first crack might mean the roast has gone out of control; a slow first crack might mean not enough heat had entered the beans. There are also variations due to varietal and processing; for example, first crack might be hard to hear with Brazil pulp naturals, or decaffinated beans, and easier to hear with higher-grown and/or washed coffees.

5. Should you buy?

The IKAWA Home Roaster is a well-built, well-designed, 100g capacity roaster. It consistently roasts tasty coffee, regardless of one's roasting experience. It doesn't make a mess with chaff all over your house, but you still have to find a way to deal with the steam and smoke of roasting. The price is in line with potential competitors. There is a 30-day "Try it" guarantee. If the price is palatable, then the arguments in favour are compelling.

However ... about those apps. The Apple iOS app offers functionality for beginner and intermediate roasters, but has a monthly charge for what advanced roasters would want. The Android app has nothing for intermediate roasters, but includes what advanced roasters would want. There is no timeline for updating the Android app, which might be a good thing for advanced roasters.

At the beginning of this review, three groups of potential users were given:
- the first group was people who are not currently roasting;
- the second group was people who want the Pro version of the IKAWA Home Roaster; and
- the third group was people who would use the IKAWA Home Roaster as a complementary roaster.
Although I did not compare to existing roasters in the same price bracket, all three groups should give the IKAWA Home Roaster strong consideration.

FYI, comments are welcome here: IKAWA Home Roasting System Review Discussion

Some minor reviewer nitpicks:
- different jars were supplied; one for chaff collection during roasting, and the other for bean collection after roasting. There is no label distinguishing between the two jars; one just has a wider cork band. A label or colour-coding could be desirable;
- there are two jars, but only one gasket/lid was supplied. A second gasket/lid would be desirable;
- some IKAWA roasters used to come with two standard-sized coffee sample trays. The Home Roaster does not come with these trays. The loading zone/cup of the Home Roaster seemed perfectly sized for loading the Home Roaster with those standard-sized coffee sample trays (I tried it). I used Mason jars to hold green coffee; to avoid spillage, I ended up using a metal Aeropress-sized funnel in the loading zone/cup to load green coffee;
- early on, one charge of green coffee did not load properly; the beans ended up on one "side" of the roast chamber and thus the beans did not start to spin once the roast started. For subsequent roasts, a gentle shake of the Home Roaster after loading settled the beans evenly in the roast chamber before starting the roast;
- in a few roasts, a few beans were ejected into the chaff during the cooling cycle (not during roasting proper). I could not identify a reason. The only common feature was the green coffees for these roasts had little chaff. The ejected beans were not quakers or semi-quakers, were not under- or over-sized, and the ejected beans did not taste like defects when chewed; and
- roasts were not completely cooled. The usual standard is four minutes to go from drop to room temperature. Beans were still quite warm to the touch after the pre-specified cooling cycles were completed. However, the cooling cycles can be lengthened with access to the Graph Editor.

Note: Thank-you to IKAWA for supplying Home-Barista reviewers with an IKAWA Home Roaster and their curated green coffee.
-"Good quality brings happiness as you use it" - Nobuho Miya, Kamasada

User avatar
Team HB

#12: Post by baldheadracing »

The Android App

If you go on the IKAWA Home website, you may notice that there is almost no mention of an Android app for the IKAWA Home Roaster - but if you go onto the Google Play Store, there is an app. What is going on? Does the Android app work?

The short answer is that the Android app works, but it doesn't have the ease-of-use features found on the Apple iOS app. Both iOS and Android apps work well for beginners; both apps work well for advanced users, albeit at a monthly cost on iOS; but the Android app has nothing for intermediate users, whereas the iOS app does.

Way back in the spring of 2015 when the 50g version of the IKAWA Home Roaster was on KickStarter, it was advertised with both an Apple iOS and an Android app.

Fast-forward a few years to the development of the 100g version of the IKAWA Home Roaster, and IKAWA found that 95% of their users were on Apple. (The Apple iOS app supports iPhones, iPads, and Macs with Apple Silicon.) As such, development of the Home app was done first for Apple iOS. The Android app remains the 50g version, as evidenced by its support for Android 5.1 Lollipop, which was released in 2015. IKAWA has no timeline for when the Android app will match the functionality of the iOS app.

I tested the Android app with a Google Pixel 6 Pro phone running Android 13 with the February 2023 and March 2023 Security and Play system updates. During the review period, the app did receive a few minor fixes, but not any new functionality. I encountered zero crashes or communication hang-ups, no Bluetooth confusion, hiccups, etc. No issues at all. The app worked perfectly and quickly 100% of the time.

There are a few quirks to using the Android app:

1. In comparison to the iOS app, there is no way to adjust recipes by "Roast Degree" or "Development Time." There are no "Roast Stages" indicated in the app (Drying, Yellowing, First Crack, Development Time). You do have similar functionality to iOS' "Graph Editor," but there is no monthly charge as iOS has. As such, the Android app is not as new-user-friendly as the iOS app. In my testing, I did not use the Graph Editor functionality; I picked from amongst their supplied recipes.

2. App permissions. Android permissions have changed greatly since Android Lollipop. Upon installing the app and before opening the app, I would suggest going into "Android Settings," selecting "Apps," then selecting the "IKAWA Home" app, and checking two things:

a. Under "Permissions," give the app all the permissions that it asks for, especially "Location," including "Use precise location," and also "Nearby Devices." While it may not seem to make sense to give the app "Location" permissions, apps for older versions of Android sometimes need precise "Location" to use Bluetooth reliably, and your enjoyment of this roaster is predicated on reliable Bluetooth communications.

b. Under "Open by default," check that "Open supported links" is turned on.

3. Open link permissions. Another security feature in current versions of Android is control of opening external applications. When you select a recipe from the IKAWA Home website (more on this below), nothing may happen. If nothing happens, then there are a couple places to look:

a. Open the Google app. At the top right, tap your Profile picture or initial and then "Settings" and then "General." Turn off "Open web pages in the app."

b. In your default browser - I'm using Firefox - open the browser, go into the browser's "Settings," and turn on"Open links in apps" or similar.

4. IKAWA's Recipes - QR code scanning. Every bag of greens from IKAWA has a QR code, so you scan the code in the app, and your browser opens up to a description of the coffee and some links to choose and download the recipe that you want to roast to the app. While this is what happens in the iOS app, what happens in the Android app is the QR code just goes to a listing of all recipes at https://ikawahome.com/pages/ikawa-home-recipe-library-app-android-reference

You then have to scroll through the listing to find that particular green coffee and then choose and download your recipe. You may find it more convenient to first open the full listing in your browser and then search and choose the recipe to download, which will then open the app.

Note that there are two kinds of recipes listed. After downloading, the recipes developed for the 100g Home Roaster will show the Roast Level and Development Time indicators in the app. Other recipes will not have Roast Level and Development Time indicators. These are called "Legacy Recipes" in the iOS app.

5. The recipes that come in the app are 50g recipes. Unlike the iOS app, the Android app includes a bunch of recipes that are not on the IKAWA Home webpage given above. To quote IKAWA, "There is firmware on each roaster that is specific to the roaster variant which allows for 50g roast recipes to use 100g roast recipes and vice versa. The 50g roasts will be slightly lighter in colour, as more mass holds and conducts heat better than less mass. The smaller 50g dose will hold and conduct less heat than a 100g dose. The difference is consistently just about 3 points on the Agtron roast colour scale. So a roast recipe that is created on a 50g machine, will have a slightly darker roast colour when used on a 100g machine. But the difference is rarely perceptible in taste, unless the user is using some extremely dark roast recipes."

I used one of the included 50g profiles when roasting 100g of a Kenya coffee; more on that in the next section.

6. Sending a recipe from the app to the roaster. If you can't send a recipe from the app to the roaster, then check that the roaster's bean loader is in the closed position and the app is showing "Ready to Roast." Sometimes the bean loader seems to need a bit of spin to learn where it is. In addition, it is good practice to confirm that the correct recipe is loaded into the roaster. While I encountered no Bluetooth issues, Bluetooth functionality in Android can have hiccups.

7. Finally, some super-secret knowledge that you will find nowhere else :-): To delete a recipe, swipe the recipe's title in the "Recipe Library."
-"Good quality brings happiness as you use it" - Nobuho Miya, Kamasada

User avatar
Team HB

#13: Post by baldheadracing »

2. Roasting with the IKAWA Home Roaster versus the iRoast2

IKAWA provided five green coffees:

As 300 g were provided for each coffee, I did one 100 g roast using the IKAWA filter roast suggested for that coffee and a second 100 g roast using the IKAWA espresso roast for that coffee. I then did a third roast with the remaining coffee using the iRoast2. I tried to match the roast colour of the IKAWA filter roast with the iRoast2. The profile used in the iRoast2 was from Sweet Maria's and recommended for sample roasting. A RoastVision was used for colour measurement two or three days post-roast, with multiple measurements averaged and rounded to 0.5.

There was a blind cupping session a few days after roasting. Beans were ground on a Baratza Vario grinder with Ditting steel burrs. The grinder was modified for single-dosing and "hyper-aligned." In the three-roast scenario above, cupping was done with two cups for each of the three roasts, giving six cups in total. The test was to taste the six cups blindly and try to form the original three pairs, and, if possible, order the pairs in taste preference. Overall, I could not tell the difference between roasts of the same green coffee. That does not mean that there were no differences; just that I could not reliably discern the differences. My sensory skills are not good; I have a Scentone kit that regularly and frustratingly proves how bad I am at picking out tasting notes. I did, however, have some taste preferences when cupping:

Colombian: iRoast2 tasted slightly better
Honduran: no preference
Nicaraguan: Home Roaster tasted better
Kenyan: no preference
Ethiopian: Home Roaster tasted much better

All roasts were also pulled as espresso regardless of whether the roast was labelled as "espresso" or "filter." The espresso results are not presented as the shots were not blindly compared. In my experience, usable results for blind-tasting espresso demand a sous-vide bath or similar very fine equalization of temperature within and across shots, as well as much larger sample sizes than cupping, for example, eight double-shots per coffee. There just wasn't enough coffee for a comparison. All espressos were pulled on a Linea Micra, with 9 bar indicated pressure, 14g in a VST-15 basket, and a Coffee-Sensor 0.2mm puck screen. A Weber Key grinder was used at 85 rpm. The darker roasts (Columbia, Ethiopia) were pulled at 200F 1:2; all other roasts were pulled at 205F 1:3. I did not notice anything untoward in any of the roasts. I would be happy to drink any of the espresso roasts.

Here are the detailed results, in order of tasting:

First coffee: Columbia San Lorenzo
- Espresso Medium +++ profile. RoastVision 19.5 - medium, 16.1% moisture loss
- Filter Light-Medium +++ profile. RoastVision 21.5 - medium-light, 15.7% moisture loss
- iRoast2 Sweet Maria's sample roasting profile. RoastVision 21 - medium-light, 15.4% moisture loss
I was able to match the Home Roast filter and iRoast2 roasts pretty well :).

Cupped three days post-roast:
- first (ranked tastiest) pair of cups were paired correctly (same roast) - there was a very slight preference for the cups from the iRoast2;
- second pair correct - the two filter cups from the Home Roaster; and
- third pair correct - the Home Roaster Espresso Medium +++ roast tasted noticeably darker from the other two roasts.
Not surprisingly, the darkest roast ranked lowest in cupping.

Second coffee: Honduras Finca Cual Bicicleta Oscar Omar Alonzo
- Espresso Light-Medium++. RoastVision 25 - medium-light, 14.1% moisture loss
- Filter Medium +. RoastVision 28 - light, 13.2% moisture loss
- iRoast2 Sweet Maria's. RoastVision 24 - medium-light, 14.6% moisture loss
I was way out on this one. Note colour matching is not as easy as it seems as the decision to end the roast has to happen before the final colour is reached, as coffee continues roasting during cooling.

Cupped two days post-roast:
- first pair wrong - one of each of the two Home Roasts tasted best;
- second pair correct - the two iRoast2 cups; and
- third pair wrong - the other of the Home Roast cups.
Some days your taste buds work better - like the first coffee. Other days ...

Third coffee: Nicaragua Julio and Octavio Peralta Finca La Argentina
- Espresso Medium ++. RoastVision 19.5 - medium, 14.1% moisture loss
- Filter Medium ++. RoastVision 23 - medium-light, 13.2% moisture loss
- iRoast2 Sweet Maria's. Roast Vision 23 - medium-light, 13.5% moisture loss
Another good colour match between the IKAWA and iRoast2 roasts.

Cupped five days post-roast:
- first pair wrong - one of each of the two Home Roaster roasts tasted best;
- second pair wrong - the other of the Home Roaster cups;
- third pair correct - the two iRoast2 cups.
The IKAWA was better, but my tasting skills failed again.

Fourth coffee: Kenya Othaya Farmers Co-op AA Karuthi
I made an error with this coffee and chose the wrong profile for the first roast. The first roast was actually a 50g roast curve that was one of the curves supplied with the Android app. Note that for that roast, the indicated temperature could not match the target curve temperature, presumably because 100g of beans cannot follow the temperature rise of 50g of beans.
- Filter Kenya Kimoini AA (50g profile for a different coffee). RoastVision 26 - light, 11.1% moisture loss
- correct Filter Legacy roast - RoastVision 24 - medium-light, 12.4% moisture loss
- iRoast2 Sweet Maria's. RoastVision 24 - medium-light, 12.0% moisture loss
Another good colour match.

Cupped six days post-roast:
- first pair wrong - one of each of the two Home Roaster roasts tested best; and
- I could not distinguish between the four remaining cups.
I cupped all three coffees again three weeks post-roast and could not distinguish between any of the six cups.

Fifth coffee: Ethiopia Buie Bora Yirgacheffe
- Espresso Dark++. RoastVision 14 - medium-dark, 15.5% moisture loss
- Filter Light+. RoastVision 31 - very light, 9.8% moisture loss
- iRoast2 default profile (I forgot to change profiles). RoastVision 18.5 - medium, 15.2% moisture loss
This was the last set of roasts for that day and I just got the colour matching wrong. This points to one of the strengths of the IKAWA Home Roaster - it does what it was programmed to do; there is no manual intervention needed during roasting. Of interest is the Home Roaster Espresso dark roast and the iRoast2 roast had very similar moisture loss while being apart on roast colour. This could indicate that one of the roasts was off-kilter, and this was shown during cupping.

Cupped seven days post-roast:
- first pair correct - the Home Roaster light+ roast had a sparkling acidity;
- second pair correct - the iRoast2 roast was bland compared to the first pair; nothing exciting;
- third pair correct - the Espresso Dark roast was easily identified.
These roasts were easy to distinguish apart, unfortunately.
-"Good quality brings happiness as you use it" - Nobuho Miya, Kamasada

User avatar
Team HB

#14: Post by baldheadracing »

3. Back-to-back roasting with the IKAWA Home Roaster

To keep my caffeine intake reasonable, these roasts were done using a decaffinated coffee - a washed Colombian (https://royalcoffee.com/product/3427097000010940063/) As there was no profile for this coffee in the IKAWA Home Roaster library, profiles for a Colombian "Rainbow" decaf were used instead, with no alterations or curve adjustments. This simulates what one could do without access to the Graph Editor.

Six back-to-back roasts on the Home Roaster were done, and alternated between Espresso and Filter. However, roast information is presented by profile:

Espresso Medium++
- first roast: RoastVision 17.5 - medium, 14% moisture loss;
- third roast: RoastVision 17 - medium, 14% moisture loss;
- fifth roast: RoastVision 17 - medium, 14% moisture loss.

Filter Medium +
- second roast: RoastVision 22 - medium-light, 13% moisture loss;
- fourth roast: RoastVision 21.5 - medium-light, 12% moisture loss;
- sixth roast: RoastVision 21.5 - medium-light, 13% moisture loss.

Blind cupping easily distinguished between Espresso and Filter roasts. However, the consistent colour and moisture loss within roast type is impressive, and no difference could be tasted between the three filter roasts, nor between the three espresso roasts.

Two iRoast2 roasts were also done, albeit with a 150g charge instead of 100g:
- Sweet Maria's profile: RoastVision 23 - medium-light, 12.7% moisture loss;
- Default profile: RoastVision 20.5 - medium-light, 13.3% moisture loss.

The two roasts from the iRoast2 were blindly cupped with the Home Roaster's Filter Medium+ roast.
- first pair wrong - iRoast2 default and Home Roaster Filter Medium +;
- second pair wrong - iRoast2 Sweet Maria's and iRoast2 default; and
- third pair wrong - iRoast2 Sweet Maria's and Home Roaster Filter Medium +.
There was no clear result other than demonstrating how to get everything wrong :mrgreen:.
-"Good quality brings happiness as you use it" - Nobuho Miya, Kamasada