Is the temperature sensor integrated in STM32F429 calibrated or not?

[diverger]

Posted on September 06, 2017 at 09:59

In STM32F4's reference manual (RM0090), it gives a formula to calculate the temperature from the result converted from ADC.  But in STM32F429xxx's datasheet, I find it gives a table to show the register address store the calibration values:

I know the temperature sensor in STM32F0 series is calibrated. So, I wonder if the temperature sensor in STM32F4 is calibrated too? 

#stm32f4 #temperature-sensor #stm32f0

[Imen.D]

Posted on September 06, 2017 at 18:33

Hello

Zhang.Vincent

,

T

he temperature sensor in STM32F4 is calibrated, andyou have the calibration values in the datasheet.

Kind Regards

Imen

[Tesla DeLorean]

Posted on September 06, 2017 at 18:43

>>So, I wonder if the temperature sensor in STM32F4 is calibrated too? 

But probably not in the sense that the specific device was run over temperature in an environmental chamber.

[john doe]

Posted on September 07, 2017 at 01:20

While it may be 'calibrated,' it is still a relative measurement and not an absolute one.

[Uwe Bonnes]

Posted on September 06, 2017 at 23:40

Clive,

from the datasheet I read that the chip/wafer is actually held at 30 and 110 degree and the ADC values are recorded.

[diverger]

Posted on September 07, 2017 at 03:28

Can you explain more?  Can't we measure the temperature (at least the temperature in the chip) even taking advantage of the two-point calibrated data?

[diverger]

Posted on September 07, 2017 at 03:33

Can you give more explanation? I think by use the two-point calibrated data, we can get a more accurate temperature.

[Tesla DeLorean]

Posted on September 07, 2017 at 03:51

Precision or Accuracy?

Unless you run your chip/board/enclosure up and down across a temperature gradient to understand how your supply voltage and regulation behave, and come up with some sort of correlation, I don't think it will have much value beyond just a gate temp estimate.

[diverger]

Posted on September 07, 2017 at 04:10

Yes, I mean Accuracy.

Actually, I do some correlation, at least for the VDDA. Because the factory value is calibrated at 3.3V. So I measure my VDDA, and do some 'scaling', and try to make it stable.

So, I think the temperature I get will be more accurate. Sure, I know it may only the temperature inside the chip, not outside the chip.

[Tesla DeLorean]

Posted on September 07, 2017 at 17:12

I'm saying outside forces with systemic/dynamic impact will be the primary drivers of accuracy, you should be thermally cycling your system to understand its behaviour. A white-paper/app-note from an ST Lab will cover their experiences, your mileage will likely vary. If this is important to you I recommend you characterize it in your own lab/hot-box.