I didn't play much with the fan speed. It gently lowers throughout the roast as the beans lose density. I'm curious to try GDM528's method of higher loft on the beans. I may have forgotten to mention it, but a washed Costa Rican bean was used for my testing. I'll play with a natural this weekend.

Tried a few of the stock Ikawa app and Library profiles for a Brazilian Natural with not so great results.
The RoR Medium profile from post #71 have gotten me the best result so far.
I have noticed a little bit of tipping for these beans vs using the same profile for a Costa Rican with a much nicer outcome, but my taste buds aren't developed enough to pick it up in the cup. Will try experimenting with the setpoints for the next batch.

SquidMan wrote:I have noticed a little bit of tipping for these beans vs using the same profile for a Costa Rican with a much nicer outcome, but my taste buds aren't developed enough to pick it up in the cup.

The beans in your photo look pretty nice IMHO, with nary a hint of tipping to my eye. Was the tipping observed using the RoR profile in post #71? If yes, these profiles are similar, but one minute shorter, which may reduce the 'burn time': IKAWA Home - profiles

I try to keep the batch size small enough to keep the greens largely off the center dome in the chamber bottom. I suspect direct contact with the center dome can overcook the greens that spend too much time sliding around it. Furthermore, some of IKAWA's recipes will spike the temperature at the start of the roast, when the greens are their heaviest, which seems like a recipe for tipping. So, with that in mind, I've adjusted my batch size to get three shots from each run: Batch size = (3 x espresso dose size / 0.85), so for 18g shots I run 65g batches.

You raise an interesting question, however: is IKAWA tipping a really bad thing? I see a few dark areas on some of the roasted beans, but they're not burnt. Maybe that simply adds a few darker roast taste notes to the overall finish level. FWIW there's no part of the chamber that's any hotter than the air temperature that's roasting the beans, so excessive direct contact with the surfaces might equate to a few of the beans being overdeveloped, which I associate as bland - but inoffensive.

GDM528 wrote:The beans in your photo look pretty nice IMHO, with nary a hint of tipping to my eye. Was the tipping observed using the RoR profile in post #71? If yes, these profiles are similar, but one minute shorter, which may reduce the 'burn time'

Yes, the first batch was roasted using the RoR profile from post #71. Pulled a few shots on a Lelit and a Flair with decent results, Flair shots were sweeter as I had more control over the pressure. The machine shot was like biting into a 90% dark chocolate bar which isn't a bad thing, but definitely better in a latte.

drgary wrote:You could increase the fan speed in Graph Editor, which would cause more of the heated air to flow through the bean mass rather than stall there (at least in theory). And, reducing sudden and strong temperature increases would tend to eliminate tipping. Here are just a few of the beans that look tipped to me. It doesn't look extreme.

I tried out the stepped 185 260 drop profile and decided to make a tiny tweak to the fan settings. I noticed after preheat the first 35-40 seconds or so there wasn't much bean movement so I bumped it up a bit, which flung some beans out so I settled on 90% while tapering back down to 80 throughout the roast. Will definitely go back and do a few experiments with the fan settings to sort out some of the defects?(not too sure) I see on some of the beans that might have been sitting against the chamber walls too long, at least that is what I assume is happening.

SquidMan wrote:I noticed after preheat the first 35-40 seconds or so there wasn't much bean movement so I bumped it up a bit, which flung some beans out so I settled on 90% while tapering back down to 80 throughout the roast. Will definitely go back and do a few experiments with the fan settings to sort out some of the defects?(not too sure) I see on some of the beans that might have been sitting against the chamber walls too long, at least that is what I assume is happening.

First off, I am in awe of how good that roast looks. Just looking at the beans will make the coffee taste better. Can you be more specific on exactly what greens you're using?

I settled on 80% fan speed such that I lose an average of one bean per roast. I'm presuming that the IKAWA is ejecting low density greens, which I probably don't want in my cup anyway. Post roast I'll typically toss out another 2-3 beans for looking distinctly off-color from the overall roast - which might imply I could have set the fan speed higher, but I'd rather reject bad beans visually.

Batch size matters, and I've found that a given batch size and green type will have a sweet spot between lofting the greens and ejecting too many of them. So for me, it's 80% @ 65g. I was targeting first-crack at the six-minute mark plus two minutes in development, but there may be other combinations of greens, batch size, and fan speed that can hit the same timing (presuming that's the goal).

I think loft is most important once the bean temperatures exceed 190C/375F (app will show 230C/446F), where the sugars start seriously burning. I generally don't see the beans fully lofted until the latter half of the browning phase, around four minutes in. The greens sliding around on the bottom of the chamber doesn't seem like a huge concern to me, as the temperatures up to that point are too low to significantly burn the greens. That's one of the key reasons I stopped using IKAWA's Home profiles that spike the temperature early in the roast.

GDM528 wrote:First off, I am in awe of how good that roast looks. Just looking at the beans will make the coffee taste better. Can you be more specific on exactly what greens you're using?

I suspect my phone is doing a little processing to make the beans look nicer. For this current batch I am using beans from SM specifically Costa Rica Honey Process Familia Ureña. Last batch I was a Brazilian Minas Gerias from Hacea.

So far I've tried 100g/75g/65g/50g batch sizes. I have found 50-65 grams to be quite similar so I've stuck with 65g(I kinda like how the beans form a nice donut shape while they spin in the chamber). With 75 grams and above I noticed some of the beans I have used cover the middle dome at the bottom of the chamber and can stay there for a while even later into the roast, that may or may not contribute to some of the more developed spots that may looked scorched in previous roasts.

Thank you so much for these! The second one is an exhaust profile and ends at 210C, which is usually no darker than medium-light. But then I don't really know the effect of airflow. I'll certainly try them both!

I've been experimenting with the inlet profiles posted in these threads, adding a steeper post-FC to reach 227C exhaust with Polar Xpresso and Yemen Mokha Hajjah in 8 minutes. I've kept the airflow high. Perhaps that has been a mistake, as I'm not thrilled with the results. Only 5 days post-roast, so hard to evaluate yet.

The inlet profile came out fairly dark at RoastVision 13. The exhaust profile fairly light at 24. This was Yemen Mokha Hajjah, 75g on a Pro100. I roast for espresso, but I've been surprised by roasts of all light/dark levels. I'll report back after I try them.