Im really happy to see this Gary! your profile looks beautiful too! Looks like now you can finally start experimenting with profiles.

quick side note: did you check to see that the thermocouples were calibrated correctly before installing them in the roaster? When I bought a bunch of thermocouples off ebay when I got the quest, and after hours and hours of trying to figure out my weird readings(not spikes, but inaccurate temps) I realized I never checked the probes in boiling water. So I did. And ALL of the cheap thermocouples I bought were total crap, not a single one was giving me 212 or even close. (also my analog thermometer that came with the quest was off, so I know I always check thermometery before applying it)

Ah, well. I think I calibrated them. But there are so many tiny screws in that Hottop and the TCs are now JB Welded in there. I'll live with imperfection and dial in my profiles to the readings I'm getting. After your note though I'll check the TCs used with my HG/BM. At least I can work with calibrated TCs for that before electrically isolating the BT probe. The ET probe is isolated if I simply have it sit in the vent so that the insulation is touching the vent instead of the wire tip.

Dustin, you've also raised questions about eBay cheap probes versus others that may be better quality. So here are some at least. What sources of better probes do you recommend? How much more do they cost? Which ones on eBay are okay? How do you shop for these? If they're off temp, is there a way to calibrate the data logger or Artisan to correct that?

Congrats.

To calibrate the meter in boiling water, I think there is a tiny hole next to the power connector (but not sure, you may need to open the meter). I would just leave like it is. Also, make sure your settings in the meter match your thermocouple type.

Cheers

jedovaty wrote:However, if I were in your shoes, I'd calibrate it because that's just me

Nah! I've got a Prestina to rebuild, and a Duchessa that may get a booster spring, a Coffex to rebuild, another roaster to electrically isolate with probes I will calibrate. I'm just glad I don't have a display anymore that looks like a seismograph in an earthquake.

drgary wrote:Ah, well. I think

Dustin, you've also raised questions about eBay cheap probes versus others that may be better quality. So here are some at least. What sources of better probes do you recommend? How much more do they cost? Which ones on eBay are okay? How do you shop for these? If they're off temp, is there a way to calibrate the data logger or Artisan to correct that?

Thats a really good question, and maybe should have a thread of its own because it would be a valuable resource. The probes I bought off ebay that were total garbage were ones like these below. Im speculating they off because they way the wires were connected to the probe itself.

http://www.ebay.com/itm/1M-3-3ft-K-Type ... 4168bbfd65

http://www.ebay.com/itm/1M-3-3ft-K-K-Se ... 53ee730111

All the bare(or naked thermocouples i have bought have been great on ebay. Also any thermocouple ive purchased that was just a probe and a handle works well. Its just the probes that have a screw in/on attachment that have let me down.

The whole thermocouple debacle made me stear clear of "probed" thermocouples, and ive been using naked ones since then on my quest. Though recently Ive been wishing I could find a probed thermocouple for the new favorite bean temp spot.... But once again Ill probably have to put something together myself.

As ive stated to Gary before, I understand why people say "it doesn't matter what the temp says, it just matter what that tells you". For lots of reasons I dont personally agree with that. For instance in this scenario you wont ever be able to replace a probe on the hottop with out having to relearn what all the temp reading mean again. Also if your first crack is happening at 340(like it was before) and your second crack at 355 thats not a very big spread to determine a drop temp. And of course just being able to talk to others about your profile.

Obviously to each there own, and I probably obsess a little to much about temps in general, but any variable you can quantify seems worth it to me.

Back to thermocouples... I would be much obliged if someone could point me in the direction of a fast accurate thermocouple that is not insanely expensive.

Thanks, Dustin. Since those links will quickly expire, here are images of the types of probes you're describing.





I agree it would be optimal to calibrate these probes. I'll do that when I replace them. When I do I'll have the built-in sensor in the Hottop to use as a reference point to track my roast as I recalibrate my expected probe temperatures. For now I just don't have the time to fuss with all of those screws in the Hottop, expose the probes I have in place and test them. There's just too much else to do. As I learn to use the Hottop I'm going by Henry Chang's guidelines, and he was using only the built-in sensor.

However I'll have a chance to calibrate the very different probes in my HG/BM, which I'm now getting ready for plugging into Artisan. Neither of them resembles the probes you've pictured. And, the probes in my Hottop aren't like those either. They resemble the one with the longer probe but there's no screw-in attachment and the ends terminate in a yellow plug with two blades.

I will be able to compare roasts of the same beans in the different roasters with different probes. That may help me interpolate approximate temperatures for the probes in the Hottop. For instance 1C for the same beans may give slightly different temperature readings. Knowing the 1C temperature on a calibrated probe will help me interpolate for the installed ones in the Hottop.

Thanks for posting those photos Gary. I would strongly recommend everyone to stay away from ones that look like those(on ebay).

I always think the readings will not be the same across roasters, and even among roasters of the same manufacturer. The environment, air flow, location of thermocouple within "bean mass", drum rotation, etc are vastly different even with the same roaster at different location. The Mini500 profiles are different in my garage vs Gary's front yard, even with the same charge weight and temperature.

We tend to think first crack occurs at the same temperature for the same bean. Or does it? I have been noticing with the heat turned down at 188c, first crack will occur later, at 194c, vs if the heat is not turn down, then it occurs at 192c. I even noticed the temperature readings are a little bit different before and after cleaning of various tubings. And this is for the same beans in the same roaster! Therefore I believe as long as the profiles are reproducible, and the readings from the thermocouples are consistent, the absolute accuracy does not really matter.

Today I did two roasts with the HG/BM and one with the Hottop. I tested the probes for that and the BT probe is 3 degrees cool in boiling water. The wire probe inserted at the vent is dead on. There was a little noise on the HG/BM but I noticed that some metal on the BT probe had bared when it rubbed against the entry hole through the outer case, where I've got to do a better job of deburring. Even then I turned the meter off and on again and the noise disappeared.

Here's the probe I'm using. I recoated it with JB Weld tonight. The lump on the shaft limits it from entering too far into the mixing bowl where it can get hit by the agitating vane.



I've put ferrites on the lead, as you see here. Yakster has asked if I could test without ferrites to see if they're making a difference. I didn't have time for that today but will do it soon.



I've also found the sweet spot for loading that roaster at about 12 oz. With the HG/BM I pre-load the beans. The heat gun is hot enough and loading through the chimney scatters beans enough I don't think I'm hurting anything. Here's how my Guatemalan water processed-decaf looked in the cooling tray. There was barely any chaff.



Chaff collects underneath the mixing bowl. This is what remained after my first roast today, a mix of leftover Salvadoran, Guatemalan and Sidamo greens.



The Hottop is now working like a charm, so I dared to roast some really good beans -- this year's Klatch distribution of Gedeo Worka. I looked at others' results and guesstimated a profile, pre-heating to 340 ET, loading 210 gm greens, taking it up fairly fast and not extending too far after 1C, ejecting the beans at 10:30. Just a few hours later the beans are starting to smell good. When you look at the tracing you'll see end of 1C displayed about 20 seconds before it actually ended. Also I started recording about 90 seconds into drying. 1C on these beans is lively. To calibrate the probes I'm thinking of ordering an extra from the same source and testing it. They're cheap enough I'll save time disassembling the roaster and will have an extra when one of the installed probes fails.