Temperature Probes overheating

[Rocquestar]

I installed two new temperature probes on a new GEM a few weeks ago, and configured them on the dashbox to show temperatures in the two rooms that they were installed, set up a dashboard, etc. A few days later, I saw that the dashbox had stopped getting temperature data from them. When I went to check the connections, I found that the metal probe was hot - as in, too hot to touch. The probe was arranged to read the temperature in the room, so it wasn't in a heat register or on a lightbulb or something- the heat was definitely coming from within the probe.

I disconnected it right away, of course, and went to check that I hadn't crossed the wires somewhere. These were connected using a 'star' wiring topology - a short run to a punch-down-box with 110-style connections, where the two new runs were similarly connected.

After seeing that the wiring appeared correct, I reconnected the hot probe, and the readings resumed as normal.

Then, today, the same thing happened. I disconnected it, then reconnected it, and grabbed a screenshot of the dashboard, showing the temperature - that's in Celcius, so Siri tells me that's about 172F. But with no changes to the wiring, it came right back to the temperature of the room in the few minutes that followed.

The strange thing is that it occurred on two different probes. The only thing common is that they are both connected to the same GEM, and over the same set of wires (obviously, as there's only one 1-wire port to connect to).

Anyone khow what might cause this? Anyone seen this before? There are the 'new' probes with the 3m leads, and metallic ~5cm long actual probe. (I have some of the other, older style probes also, and have never seen this behaviour from them)

Thanks, for any insight anyone may have.

;-D, Rocquestar

Screen Shot 2021-03-19 at 3.24.19 PM.png (228.68 KiB) Viewed 4528 times

[Teken]

I installed two new temperature probes on a new GEM a few weeks ago, and configured them on the dashbox to show temperatures in the two rooms that they were installed, set up a dashboard, etc. A few days later, I saw that the dashbox had stopped getting temperature data from them. When I went to check the connections, I found that the metal probe was hot - as in, too hot to touch. The probe was arranged to read the temperature in the room, so it wasn't in a heat register or on a lightbulb or something- the heat was definitely coming from within the probe.

I disconnected it right away, of course, and went to check that I hadn't crossed the wires somewhere. These were connected using a 'star' wiring topology - a short run to a punch-down-box with 110-style connections, where the two new runs were similarly connected.

After seeing that the wiring appeared correct, I reconnected the hot probe, and the readings resumed as normal.

Then, today, the same thing happened. I disconnected it, then reconnected it, and grabbed a screenshot of the dashboard, showing the temperature - that's in Celcius, so Siri tells me that's about 172F. But with no changes to the wiring, it came right back to the temperature of the room in the few minutes that followed.

The strange thing is that it occurred on two different probes. The only thing common is that they are both connected to the same GEM, and over the same set of wires (obviously, as there's only one 1-wire port to connect to).

Anyone khow what might cause this? Anyone seen this before? There are the 'new' probes with the 3m leads, and metallic ~5cm long actual probe. (I have some of the other, older style probes also, and have never seen this behaviour from them)

Thanks, for any insight anyone may have.

;-D, Rocquestar


Screen Shot 2021-03-19 at 3.24.19 PM.png

Can you provide a picture of the actual sensor tip and sensor wires.

[Rocquestar]

Hi Teken-

Sure can. (If I misunderstood what you meant by “sensor tip” let me know)

The small 110 block is the start part of the 1-wire topology- lower wire is to the GEM, the upper two are to the two probes.

[Rocquestar]

Yikes, those were big photos!

I actually see that there was another spike this afternoon (after the first one today) which corrected itself - but both probes on that GEM were affected.

This suggests to me that there's a short in the wiring somewhere.

(Pink and yellow are the affected ones, and they're the two on the same 1-wire wiring.)

Screen Shot 2021-03-19 at 5.47.57 PM.png (58.01 KiB) Viewed 4521 times

[Teken]

I would remove everything and insert just a single sensor to the GEM and monitor what happens.