Slope of internal temperature sensor

[kbroger]

Posted on April 10, 2012 at 11:25

Hello,

we are using the microcontroller STM32F100C6T7 in our design. While trying to do something with the internal temperature sensor we noticed that the slope we measured doesn't correspond with the specified slope in the datasheet at all. Does anyone know something about this?

[frankmeyer9]

Posted on April 10, 2012 at 13:13

I hope you are aware of the fact that this internal sensor measures the die temperature, not the ambient temperature. As the uC heats up as soon as you turn it on, the measurement are changing (say, increasing), even under constant ambient conditions.

If you want to check absolute figures, heat the uC up to a point above it's usual die temperature under ambient conditions. We do this by heating the whole device to 60..80°C, and keep it there for at least 15 minutes, before measuring.

[raptorhal2]

Posted on April 10, 2012 at 22:31

What slope are you experiencing ? It is easy to get the slope sign incorrect.

If you believe you have done everything correctly, post the code and perhaps we can spot something. Cheers, Hal

[rosarium]

Posted on April 11, 2012 at 06:59

For what purpose you want to use temp sensor??? to measure ambient temperature you can use STLM75 / STLM20 etc.

[r2d2]

Posted on May 16, 2012 at 13:47

hey there, i got the same problem with average slope. all datasheets show that it is 4.3 mV/C, but then calculations do not match

But when you say, that slope is 43 mV/C then it matches...

so my experience say, that its a typo error in datasheets which is missed suprisingly for a very long time