Why does this thermocouple measurement work so well? (EEVblog #512)

[necessaryevil]

Why does this thermocouple measurement in EEvblog #512 at 14:35 work so well? The sensor is just touching the regulator, so most of the sensor is exposed to the ambient.

[electr_peter]

Video is cut. Thermocouple heated up "in the backstage".

[necessaryevil]

Yes, but most of the thermocouple is exposed to the ambient air, so I would expect a much lower temperature.

[electr_peter]

Thermal resistance of thermocouple <-> heatsink - very low.
Thermal resistance of thermocouple <-> air - very high. Air flow is also very low.
Comparatively high thermal resistance to air has almost no effect on heat transfer from heatsink to thermocouple. It is like connecting 1Ohm and 1000Ohm resistors in parallel - circuit would not see any significant difference between [1Ohm] and [1Ohm | 1000Ohm] load.

Also, thermal mass of thermocouple is very low, so it would heat up very quickly to almost the same temperature as the metal heatsink.

[necessaryevil]

Hmmm... Thanks! I figured out that the small thermal mass would help, but I didn't think about the thermal resistance difference.

Good explanation with the parallel resistance analogy.