Saved by the voltages, I was in

The voltage could be a source of differences. I have made a more final 3d printed case to go with the pump and I'll get some more data soon. This should get me again with TVOC readings.

Esteve wrote:Saved by the voltages, I was in

Then it might be easier for you to take your equipment somewhere down to sea level, maybe

That seems doable, if I manage to get a tidy setup as Gary's I might give it a try.

Did 3 more roasts of the same profile, the FC seems very consistent. Today the base humidity was higher, so all the results are shifted. As GDM528 posted before the MRoR is unaffected.

I always follow the same protocol, an empty run followed by 3 back to back roasts. This time the first roast's humidity was very similar to the following ones. In the past I have observed that the first one tends to be higher (shifted up).

Curves:



Also, since we are here for the end drink, just say that profile is super tasty All my gear is towards lighter roasts, so I am sure a better result could have been achieved, but I am quite happy with it. Also, it only rested like 9 days...

Profile used:

https://visualizer.coffee/shots/8f5398e ... 53254986ee

Result:

Endurance Test!

Set a new personal record for consecutive, back-to-back runs with my IKAWA Home: two runs empty, followed by twelve runs of greens with no pauses. Each run was an eight-minute curved ramp that produces a linearly-declining RoR. First crack set to roughly six minutes and ends at a City+ roast level.

After two empty-chamber runs, ran two runs of a Sumatran green, and logged the moisture release. Then I ran six different greens for another nine runs. the twelfth and last run was the same Sumatran green.



So, by the twelfth run the moisture curve shifted less than ten seconds later than the first two runs. I suspect the primary cause could be related to the cooling effect of the silicone tubing used to route the exhaust gas to the moisture sensor. The silicone tubing will warm up gradually, causing the air reaching the sensor to also warm up, which in turn will affect the humidity reading. The sensor is supposed to correct for air temperature, but the correction could be less than perfect.

FWIW, the humidity readings are very solid, compared to the VOC readings:



Interestingly, the gap between the first and second run is larger than between the second and the twelfth run. In their defense, the gas sensors are quite sensitive, and will pick up on the history prior to the current run - that's why the first run (green line) is so different, as the prior run was an empty chamber. Likewise, the last run of the Sumatra was preceded by a different green with a different VOC output.

This is your lung on roasting exhaust.



The right-hand filter is after about 20 runs, 65g batch size. No chunks of debris, just a gradual discoloration - and only on the receiving end of the filter. The other end of the filter is pristine, so the filter seems to be doing a proper job of removing the residue that might otherwise gum up the expensive air pump and gas sensors downstream.

I'm using cigarette filters, fitted in-line with the silicone tubing: https://a.co/d/bRriAQr
These 'slim' filters happen to be a very close match to the diameter of the silicone tubing used to sample the exhaust. That's helpful for the design of the adapter needed to accommodate easily-replaceable filters while also forming a tight seal to the silicone tubing.

Until I see a teardown, this has me a bit concerned about the humidity sensor in the Pro100x. Their cleaning procedure only uses a pipe cleaner to clear out physical debris out of the sampling port, but that's not the same thing as gaseous chemicals in the exhaust air...