I've just read all of the pages below to search for similar problems as we encounter and it seems to concern ground loops.
We use the TC-08 on a normal desktop PC, K-couple, measuring in an electric oven. Trouble is above ca 300C the reading drops periodically to <100C. We have tried everything: different copules, connectors, ground cables all over the place connecting oven, PC, USB shielding. Even another TC-08 (we have two of these), no effect. A second K couple on ch 2 behaves the same. Our Picologgers both have no ground terminal.
You said before that running a laptop on battery power solves the problem but we only have a desktop. Is there still one option that we did not consider regarding grounding? Can we use the mass shielding planes on the print inside the pico for grounding the device (of course we looked inside the box too, we're engineers allright).
Maybe use a 1:1 separation transformer in the mains lead of the PC?
We would greatly appreciate your input. Thanx in advance, best regards from holland.
Element Hengelo NL