STM32L temperature trimming values 90° or 110°

[hartmann]

Posted on October 10, 2012 at 09:25

Hello,

currently I'm evaluating the internal temperature senor of the STM32L151, but in the docs are different values for trimming values given. Sometimes 90° is stated somtimes 110°, so which value is correct? STM32L151xx datasheet -> 110° for upper factory trimming value STM32L151xx errata     ->    90° RM0038 reference manual -> 90° regards thomas

[raptorhal2]

Posted on October 10, 2012 at 16:05

In fact it doesn't matter. The relevant advice from the user manual is:

To reduce the temperature sensor error, the user can measure it at ambient temperature (25°C) to redefine more accurately the average slope (avg_slope) and the offset. The temperature sensor has an offset that varies from one processor to another, so each processor needs to be calibrated. If your design is for a production run, an external temperature sensor with less variation in offset might be more suitable. Cheers, Hal

[sheldon]

Posted on October 16, 2012 at 01:54

I just received an email from ST stating:

I have preliminary information from our factory, and it appears STM32Ls high density such as yours have correctly programmed engineering data only starting from revision Y (newer than your revision Z).

All previous revisions are susceptible to bad engineering data programming.

(I'm using STM32L152VC)

[raptorhal2]

Posted on October 16, 2012 at 15:28

Now we just have to figure out the algorithm difference between ''

correctly programmed engineering data

'' and

''bad engineering data programming''.

STOne - What is the design slope and offset for Revision Z ? Is there a way that Revision C can be used to measure internal temperature ? If all else fails, with Rev C use actual readings with the processor in a plastic bag in ice water, then at room temperature, then in your expected usage environment with a thermometer nearby. For production, you should switch to Rev Z. Cheers, Hal