T100 Electronics version 0.2 test

[charlespax]

Experiment

I tested Electronics version 0.2 against a Fluke 54B II thermocouple thermometer. You can see the experimental setup in the image below. The thermocouple connected to the T100 and the thermocouple connected to the 54B have their ends wrapped around each other, are submerged in hot water, and allowed to cool.

A USB microscope monitors the 54B while the T100 temperatures are output to the terminal. The screen was captured in a video, which is sped up by 16x. Later in the video the 54B displays the ambient air temperature measured by a second thermocouple in the air.

Analysis

In this experiment the T100 consistently measured approximately 3.5 C above the temperature measured by the 54B. The T100 is also measuring a junction temperature approximately 3.5 C above the 54B junction temperature. Presumably, the 54B junction is approximately air temperature as it has been sitting idle in the room for hours.

A closer look at the T400 connection gives a clue to the temperature difference. In this experiment the actual and measured junctions are not thermally connected well, which can result in a temperature differential between the the true junction and the measured point. It is reasonable that the MCP9800 is measuring an accurate temperature of itself while the true junction temperature is closer to the 20.7 C air temperature measured by the 54B. Taking this into account, the T100 is measuring temperatures fairly close to the 53B.

[charlespax]

Opened an issue to retest (https://github.com/PaxInstruments/t100-electronics/issues/45). Closing this issue.