---

I don't see the reason to push sample time more than 10x faster than the slow temperature change rate due to the TC/chip resolution.
Explanation: an average 10 C/m RoR led us to one C degree change every 6 seconds.
The MAX/TC resolution is 0.25 C degree.
So we have 4 unique temperature values in a 6 seconds windows, a change every 1.5 seconds, as average... what sense have to acquire them 15 times faster, at 100 ms rate ?
If you see this as an oversampling, I find it a bit excessive...

Anyway, the thread is about control