dutch_dk wrote:The behavior has been normal since I made the ufinished grounding.
Your solution with grounding is correct.
I have had the same experience with several USB TC-08´s and they are very sensitive to even small voltage differences between input terminals and computer chassis/ground.
Even isolated thermocouple (2 mm steel tubing type, type T) made occasionally troubles.
Remember that the Seebeck voltage is only 39 µV/K around 0 deg.C
We have taken the consequences and provided all of the Labtops used for TC-08 with approx. 4 m soft cable from the chassis and a crocodile clip and instructed the user to always connect the measuring object with the computer.