Hi GDM528,

Have you ever thought of a more permanent position of the thermocouple?
Maybe in the black plastic exhaust funnel mounted on the glass plate - I assume that mounted there the thermocouple would get exposed to the hot wind from the roasting chamber.
That would just be drilling a tiny hole and running the thermocouple through.

I looked at the small thermocouple from Yoctopuce, and thinking of installing it that way - That way there is no reason to tear down the machine, but you would still be able to get statistics for a roast on artisan

Anyway?
Did you ever test measuring the temperature other places in the Ikawa - Like by the exhaust funnel?


[/img][/img][/img]

After extensive research and some great advice from GDM528, I ordered a 24 gauge Kapton insulated thermocouple which arrived this week.
The wire is 0.5 mm in diameter, and with the insulation its 1.2 mm total.
I am not sure that will be thin enough to place where GDM528 did without stripping the insulation, but we shall see.
If not, I have successfully run a thicker thermocouple through the exhaust port and into the chamber.
This one should be much better as it's much smaller.
Will do some testing, and post my findings and artisan curves this weekend!

Hi, did you get wildly different readings at different placements of the thermocouple?

nicolai wrote:Have you ever thought of a more permanent position of the thermocouple?

Did you ever test measuring the temperature other places in the Ikawa - Like by the exhaust funnel?

I did try monitoring exhaust temps, but it was much harder to interpret the results. The air goes through many expansions and compressions between the intake fan and the exhaust port, and the air pressure fluctuations will affect the temperature reading quite a bit. I suppose over time, one can gain enough experience with the temperature behavior at any given point to correlate with roast results, as I'm sure Ikawa has done. So, it might be prudent to copy where Ikawa has located their exhaust temp sensor - this is where a Pro 100 owner could help out

I consider the actual bean temperature to be the reference-standard for knowing what's going on, so, either thermocouples or IR sensors. IR sensors would require some serious hacking and cost easily as much as 20 pounds of some nice single-origin beans. An ultra-low thermal-mass thermocouple probe placed squarely in the spinning bean mass is much cheaper, and as I asserted in the original post, comes really close to the readings I would've obtained with an IR sensor.

As for a permanent location, if at some point I decide to open up the machine, I have my eye on that bolt in the chamber just under the bean ramp. I had been thinking that bolt was plugging the hole where the sensor would go in the Pro version - but mgrayson posted a photo that shows it actually at the top of the chamber. IMHO I think the thermocouple mounted at about the same level as the bean ramp (halfway up the chamber wall) is better than the Pro's location (apologies to Pro owners).

The thermocouple I taped into the chamber is still in decent shape after dozens of roasts, so it may be quite a while before I need to think about relocating it. Compare this image to my original post:

No probe outside a bean can tell you what's going on temperature-wise inside the bean. Even when the probe is stuck in the bean mass, it sometimes reads the surface temperature of a particular bean, and sometimes reads the temperature of air between the beans.

All probe temperatures are just a proxy for what's happening with the beans, and not an exact one at that. I suspect if you suspend a probe in the bean mass and one above the bean mass where the Pro's probe is, you'll find that the shape of the curves is close to identical, likely with the probe in the bean mass reading a relatively constant amount higher than the exhaust probe. It may well be that exhaust temperature much further downstream won't correlate as well due to mixing with colder air, but I think the Pro's probe is close enough. At any rate, it works well with my roasts.

Peppersass wrote:I suspect if you suspend a probe in the bean mass and one above the bean mass where the Pro's probe is, you'll find that the shape of the curves is close to identical, likely with the probe in the bean mass reading a relatively constant amount higher than the exhaust probe. It may well be that exhaust temperature much further downstream won't correlate as well due to mixing with colder air, but I think the Pro's probe is close enough. At any rate, it works well with my roasts.

Post #42 confirms your theory. The buried-in-the-bean-mass probe has less lag, which might be from the extra thermal mass of the Pro's stainless-steel housing, or more aggressive smoothing in the data processing. Regardless, as you point out, either method can produce the quality data needed to make well-informed roasting decisions. Also shows how well calibrated our machines are with respect to each other (even across different models), which gives me a lot of confidence in sharing and modifying recipes.

Update: Tried my Kapton insulated 24 gauge thermocouple today. I would say results were mixed.
I used the placement similar to GDM528, but didn't strip wires.
First placement was too low (thermocouple was at the depth of the screw). That greatly interfered with bean rotation.
Second placement was better, but I had some beans get stuck near the wire.
Third placement is as the photo below, which is at the height of the bottom of the bean ramp. Rotation wasn't impaired using this approach, so this was the minimum height in my experimenting.
But the other issue is I was clearly getting some air leak through the glass. With the Kapton insulation on the wire, it's 1.2 mm thick. No chaff came out, but I could see that some was getting stuck at the edge of the glass and the foam. I then taped the outside edge near the thermocouple with Kapton, and that solved the issue, but then you can't lift the glass off with ease. Not ideal.
Below is the Artisan curve I was getting with the custom profile I created.
It's not bad, but tomorrow I am going to try removing this thermocouple and using my previous probe as an exhaust probe. I think, in retrospect, that may work better. More testing to follow!

Update 2:
So I just switched thermocouples and now I have an exhaust thermocouple.
This one is Perfect Prime TL1004.
I mounted it as close as I could estimate to where the Pro Exhaust probe is.
The wire then goes down the exhaust tunnel, and then out the exhaust tube. That means no issues with the glass seal.
I will test tomorrow if it interferes with chaff removal or bean ejection, but I don't think it will as I used this thermocouple previously but was trying to position it as a bean probe. Otherwise, the wire path was the same and had no issues with chaff/beans ejection.
It's a 3 mm probe, so isn't the speediest, but hopefully will work well enough!

Iowa_Boy wrote:But the other issue is I was clearly getting some air leak through the glass.

Yeah, the glass cover leaking 270C air is a hades-class showstopper. I only dodged that by completely stripping all the insulation off narrow gauge thermocouple wire. With the cover off, there are four adjustment screws that can raise/lower the level of the foam seal relative to the metal shell. I made sure the foam seal was high enough to absorb the thickness of my thermocouple 'tape', about 0.3mm.

If there's some way to post videos here, I'll try to post a clip of how the beans flow around the probe tip. Early in the roast, as the beans dry and elevate, the probe tip will knock some of the beans into the center, similar to how Ikawa's bean ramp works. Late in the roast when the beans are 'at altitude' there's no apparent effect from the probe tip. I'm starting to think I got lucky with my placement.

Iowa_Boy wrote:I mounted it as close as I could estimate to where the Pro Exhaust probe is.
The wire then goes down the exhaust tunnel, and then out the exhaust tube. That means no issues with the glass seal.

Very nicely done - totally looks "Pro". Is it taped in place - what keeps it perpendicular to the chamber wall?