I had a chance to tailor a profile closer to the graph I posted earlier. I plateaued the temperature at 450F and let it ride for about 1min at 450F. First crack started around 7:22 and 407F. It ended at roughly 8:20 @ 422F. Second crack started at 9:30 at 448F. Rolling 2nd crack started at 10:00 450F. I dropped the Sumatra at 10:20 and 450F just as Happy Mug suggested. (20s after rolling 2nd or 40s after start of 2nd.) I was able to clearly hear 2nd crack now that I shifted the temperature back up.

Something I noticed going higher with the Ikawa is that the beans come out quite warm. I measured them at 115F after being ejected. So I broke out my old cooling box setup from the Behmor days to finish the cooling before jarring them.

Milligan wrote:You can still find the longer portion on their site under the general profiles called Morten Munchow https://www.ikawacoffee.com/pro-sample- ... rofiles/#!. Unfortunately, Ikawa seems to have removed the fast profile.

Thanks for the link. That led me to yet another profile from Rob Hoos, where he presented a "Masterclass" on creating roast profiles.
https://www.ikawacoffee.com/for-profess ... mperature/
What was interesting about the post: after much discussion a profile for the Ikawa Pro was presented with the asssertion that it would "closely mimic" a traditional sample roast from a drum roaster. Cool. Then, at the end of the article - without any discussion or explanation - they post a revised profile from Rob that's radically different...

So apparently there's a lot of ways to get from A to B, and with enough marketing language it's all good. Whatever floats your boat. This might explain why I treat my Ikawa like a cable TV tuner where I can change channels every ten minutes - and I do

Milligan wrote: Something I noticed going higher with the Ikawa is that the beans come out quite warm. I measured them at 115F after being ejected. So I broke out my old cooling box setup from the Behmor days to finish the cooling before jarring them.

I saw the Tim Wendelbow profile in the link you posted spends three minutes in cool-down, roughly double the cool-down in Ikawa's curated Home profiles. I tried opting for the shortest possible cool-down time to get the beans out of the chamber asap so I can cool them down faster... However, during actual operation my machine overrode my programmed one-minute cool-down and kept running until the inlet temperature was in the same temperature range you measured. I've observed that same behavior regardless of my ending temperature. I appreciate the machine is protecting itself from overheating, but now I'm wondering if Tim Wendelbow is on to something.

Mine seems to cool to 112 on the upper sensor.

After taste testing an 8-minute roast versus a 10-minute roast to get to a Full City dark level, I'm a bit perplexed. Same greens used for each roast, and each were similar curves: fast step followed by linear slope to a drop temperature in the 230-235C/ 445-455F range. Development times went from 2-ish minutes for the 8-minute roast to 3-ish minutes for the 10-min roast.

I was thinking the increased development time would reduce the acidity - but instead I noted slightly increased acidity with the longer roast time. The longer roast time was still good, but it didn't go in the direction I was expecting.

I got a few theories for what happened:
1) Simply scaling up the total roast time also increased the browning/caramelization time. That may have reduced sweetness enough to unmask the acidity.
2) I didn't spend enough time above whatever temperature I need to hit to break down the acidity. Is there is such a thing as 'acidity reduction temperature'?
3) An espresso shot tuning problem. Perhaps I need to extract a long roast differently than a short roast?

Any other theories/explanations?