Much time has passed and now I've installed TCs in the back plate of my Hottop B. Tonight was my first test run. Am still using the Amprobe data logger but wonder what I'm doing wrong. The temperature readings seemed fine until I dropped beans into the roaster. Then the readings went haywire until I ejected the beans and they went smooth again. This looks like I've got some very basic setting wrong. Is the mistake obvious to someone reading this? My hunch is I've got some setting wrong on the data logger, maybe confusing the types of TCs.* If not, what should I troubleshoot? At least I followed Henry's basic instructions for using the Hottop with its internal temperature readings rather than try and steer in that snowstorm, so I didn't ruin my test beans!



* Added: I just fiddled with the Amprobe data logger and made sure both probes are set as K type. The next troubleshooting step will be to immerse TCs in hot and cold water while plugged into Artisan and see if the readings and display correspond with what's being measured. Next I would test the cable, etc.


...split from Data Logging to a Native Mac Environment Running Artisan? (Yes!) by moderator...

Hi.

Strange.

Check the temperature reading of the meter display. See if that matches the software

If the temperature matches the software then:

Are you using an AC power adapter with the meter? The thermocouple tips may be touching some metal and creating a ground electrical loop that throws off the readings. Maybe a loaded drum is heavier and making the tips make metal contact. Or they could be bending the wires. Also, try unplugging the laptop power supply and see if that makes a difference (part of the ground loop).

Cheers

Thanks, Arpi. The tips making metal contact may be the issue. Does that completely mess with their ability to read temperature? I didn't have the meter plugged into an AC adaptor.

I was able to isolate that problem to the cheap thermocouples that come with the Amprobe. They were working fine in another roast. Then they started to act up again. I jiggled the wires near their connectors to the probe. They settled down for about 30 seconds and started going haywire (sorry) again. So I changed to my heat gun/bread machine equipped with different K-type TCs. The roast wasn't successful -- too light -- and I had the BT and ET probes reversed, but the tracings were completely smooth.

Not so fast. Maybe I haven't isolated the problem to those TCs. I'm suspecting it's a grounding problem. Rafael mentioned the possibility of grounding issues above and I'll have to try and disconnect the laptop from its power source and see if that issue continues. I saw a good discussion of grounding issues here:

Choosing thermometry equipment for home roasting

I roasted some Guatemalan Decaf in the HG/BM tonight using different probes and still got occasional issues with the readout suddenly jumping. At least it settled down again and gave useful readings. Here's the graph:

Hi.

Strange readings. The error looks repetitive every ~ 30 seconds. Is there a moving part with rpms? If you do a 10 minute test with the thermocouples out of the roaster and no errors happen then the problem is in the thermocouples. Maybe the beans are moving too much the tip of the thermocouple causing the tip to touch metal or something. It is better to use thermocouple probes (stainless steel) instead of weak thermocouples that move (for BT).

Some suspects are:

Bad thermocouples (most probable)
Bad meter
Low meter batteries

Cheers

Hi Rafael:

For this test the thermocouple in the beans was a probe, not a wire. The thermocouple doing ET was a wire but a thicker one and its tip had no chance to touch metal. The bread machine agitator moves periodically but that movement did not correlate with the jumps in the reading. The BT probe is threaded through a hole on the outside of the case and a hole in the inner bread pan. It would tend to slide a bit but when I would push it back toward center that didn't correlate with the jumps either. The data logger is nearly new but I can certainly check its batteries. Once when I turned up the heat gun the display jumped, but that was only once. If there is a grounding problem how would I test that? Unplug the laptop?

I'll start by plugging in the data logger with the probes attached and will run it with Artisan with the laptop plugged in. If it's jumpy the next stage would be the same test with the laptop unplugged. Next I guess would be a battery test for the Amprobe.

I did the first stage of the test just now, running it for 10 minutes. There were no jumps in the display using the same TCs I did for yesterday's roast. The slight changes in readings occurred when I warmed different TC tips with my fingers. Once I tried to gently wiggle the plugs where they connect to the data logger and this had no effect. At the 7 minute mark I briefly turned off the data logger and turned it on without pausing Artisan. That didn't show up in the display. I also unplugged the laptop and plugged it in. No change. Once early on I touched the end of the wire TC to the metal on the laptop and there was only a slight change in temperature reading. Using different TCs and different roasters (HG/BM and Hottop) I've had this problem only when running the roaster.

It could be also induced voltage on the thermocouple wires (antennas picking noise). Wrapping the thermocouple wires on aluminum foil (like inside roaster) should help reduce induced voltage picked up by the meter. I think you'll have to do like a process of elimination till you find what creates the 30 second errors.

Cheers

Thanks, I'll try that. I think it may not be the issue though because I haven't seen others requiring shielded cables. The only thing that's different than most others is I have an Amprobe data logger. Most people don't use a Mac either but given the number of those out there I doubt its grounding functions differently than any popular notebook computer. I wonder if somehow the data logger is picking up induced voltage and needs to be grounded? How would I effectively ground it?