cough cough seasoning beans cough cough
Another little tidbit that I am -completely- ignoring is where he spends a significant time explaining why it really doesn’t matter, within reason, how long it takes to get to first crack. Like almost 14 minutes to first crack doesn’t matter.
I’d buy that.
I’m probably going to run one more low drum speed test and shelve the whole thing until I hear back from Aillio support.
So… if your setup gets you a consistent, say, FC at 8:30 why not just go with that and apply the rest of his philosophy from there? 
Support responded and gave me some tests to run.
I conducted a wattage test at the request of Aillio support.
I measured the wattage in ascending order from P1 to P9
P9 had a lower starting wattage than P8 and dropped steadily down to P6 power level.
I cooled the roaster down and did the same test but started with P9. Started at 1534w and steadily dropped down to P8 wattage. Power level was still declining when I ended the test.
Anyone bored with a wattmeter want to compare notes while I wait for support?
I’m confident that the continual drop in wattage at P9 is outside of spec.
Interesting… one would think P9 wouldn’t overlap with P8 like that.
I don’t think it’s supposed to? P8 and P7 were more powerful than P9. The wattage continued to drop until I ended the test.
I redid the same test but started at P9. Initially it was 1534w which is where I would expect P9 to live but just as before continued a steady decrease down to 1290w before I ended the experiment.
Would be good if someone else with a wattmeter can run the same test as you for comparison. I would have expected the low end of P9 to be higher than the high end of P8 based on all your other data points. Let us know what support says too…
What’s the point of P9 being full power if it’s less than the previous two?
That is really interesting…hopefully Aillio can give you an idea of what’s going on soon.
I wonder if there’s some thermal effect related to the induction board? Did you gather the roaster diagnostic data as the P9 test ran?
Nothing was out of tolerance.
I just ran a few more tests
Can’t get a shorter path to the fuse panel than this. Isolated 20a circuit.
Confirmed the wattage drop of P9 from before. Starts after 4 minutes of roasting. 1500w to 1100w in 10 minutes.
Just out of curiosity, trialed P8. 1350w to 1185w in 10 minutes.
Saw your post on the other thread in the meantime about augmenting the cooling to the power board as well…
Can you describe the test procedure a little more so I can give it a try and get some data back to you? I think I’m reading this but just wanted to make sure:
To repeat the experiment more faithfully, what fan/drum setting did you use for this?
Hey thanks for jumping in!
You have described my process accurately.
Fan 2, drum 9.
This is done without any beans.
So for craps and giggles, I did another test run. This time with RT and a desk fan blowing under the roaster. I was kinda hoping maybe there was a component overheating and causing the drop in power.
No preheat, No beans, P9, F2, D9
Wattage held steady above 1500w for 3:30, then continued it’s steady decline throughout the roast. FC marks the point where wattage had dropped into the 1100w territory.
You can see where the graph changes from a linear-ish temperature increase to a less aggressive rise that correlates with the decrease in wattage.
Assuming this isn’t a safety feature or intentionally coded PID to decrease power over time, no wonder I’m having trouble hitting first crack in the same time as others if my power output drops steadily after 3:30 minutes.
Just spinning my wheels until support gets back to.
Support got back to me early this evening and they suggested updating to the beta firmware 602. Reading the changelogs, I was somewhat hopeful.
Updated the firmware and ran another P9 F2 D9 empty drum test.
Starting wattage was 1500w+. After 8 minutes, the wattage had dropped to 1465. Significantly less than before.
Out of idle curiosity because heat is always bad for electronics, I turned on my “oh crap the IGBT is about to overheat again” fan. Within about 15 seconds, the roaster had returned to 1500w of power draw. The additional fan is on a separate circuit so it wasn’t that added load.
I’m cautiously optimistic. I’ll run some actual roasts tomorrow and see how they compare.
I put my wattmeter on my bullet: P9, D9, F2 (Firmware 591)
Minutes — Watts ---- PCB temp
0 --------------- 1480 --------- 30
1 ---------------- 1500 -----------32
5 ---------------- 1365 ---------- 33.5
7 ---------------- 1260 ---------- 34.6
9 --------------- 1223 ---------- 36.7
10 -------------- 1176 ----------- 40.6
I put my small underbelly fan directed at the circuit board, but didn’t see the watts rise or the PCB get cooler. (it is a weak fan)
Did it stay there for how long?
Yup, that basically matches my trend with v591 prior to updating the firmware to v602.
Edit: Placing a fan under the bullet also had no effect for me with v591.
I kept it there for a minute or two. It was late and the bean temp was approaching 300 and I don’t want to run it hotter than that for no specific reason.
