Other Parts Discussed in Thread: TMP116,
I'm currently evaluating the application of the TMP117 for an industrial process. The accuracy is quite good but I am a bit underwhelmed by the response times from my initial tests. I'm taking advantage of the fast conversion cycle of 15.5ms and no averaging in continuos conversion mode to read the device at 64Hz, this works well but during large temperature deltas the device takes a long time to reach the actual temperature being measured. The test is done with a thermal bath, which I would imagine would be the best-case scenario for minimizing thermal resistance. In my tests, the device took more than 10 seconds to go from 0ºC to 100ºC, in contrast, a thermocouple took less than a second. In the "SNOA986A - Precise Temperature Measurements With the TMP116 and TMP117" document, in section 9 Soldering to PCB, it is stated that the "it is possible to calculate the sensor thermal response time [...] response time values are 0.39 and 0.72 seconds", which leads me to believe I'm probably missing something here. Is this behaviour expected or what steps should I take to minimize the thermal response time? This is critical for my application since it is similar to a flash pasteurization process which I would like to monitor.
Some more details:
- I'm using the TMP117EVM PCB "tip" with a separate controller, just so it's clear what thermal mass we are working with.
- I'm writing 0x00 to the [01h] Configuration Register for continuous conversion mode, 15.5ms conversion cycle and No averaging.
Thank you for your time!
