MSP430FR5994: MSP430 USING PROBLEM

Hi,

the question for me is, whether your customer is using the factory calibration values for the temperature sensor, and if, whether these values have potentially been overwritten unintentionally?

The reason for my question is, the sensor does not by itself deliver the absolute temperature level, and thus needs to be corrected in terms of offset and potentially also gain. The main portion is the sensor offset.

Best regards

Peter

HI

Thanks for your reply!

I confirmed with my customer that they used the factory calibration values for the temperature sensor, and how do we confirm whether  the calibration value in the chip has been overwritten unintentionally?

thanks again!

Hello,

well this is unfortunately not possible in the sense of getting a 100% proof. If the content would be FFFFh or a repetitive pattern, one could guess that something went wrong.

I'd suggest the following. I'll check with our engineering, what the calibration accuracy and parameters are, including the to be expected results, and then come back to you on this.

Best regards

Peter

Hi,

a few things to add, which might help you making some further steps. In the datasheet of the MSP430FR5994 you can find Table 5-28. Below this table in the foot notes, there are some comments on the to be expected accuracy, when using the factory calibration values. Note 2 speaks about +/-3degC of tolerance at 30degC and at 85 degC.

Furthermore there are some indications on the required sampling time. Thus I would suggest to check the settings and conditions the customer is using to measure the values of the temperature sensor. Especially a too short sampling time could be a root cause of wrong measurement results.

With ADC measurements there are also some HW related aspects. For the specified performance there are also some requirements on the supply capacitors. Please check whether the specified supply capacitors are applied.

On top of this, if the devices are of the same Lot Trace Code one could try a comparison of the calibration values of the units with the unexpected temperature offset with the values of "good" units. Without overwriting the original calibration values, the customer could on one hand document the values and the potential differences and try also applying the calibration values of "good" units to the units with huge deviation, again without overwriting the original calibration values.

As stated before, I am checking on further data and recommendations on the ADC settings with our test engineering, but this will take time until next Monday due to public holiday in Germany.

As a starting point you could use the code example from our Code Composer Studio Resource Explorer msp430fr599x_adc12_10.c, which shows how to use the temperature sensor with the factory calibration values.

Best regards

Peter

Hello,

could you please give us an update on the status on your side? Many thanks in advance.

Best regards

Peter

Hello,

We are double checking the problem now! and we found something interesting that the factory calibration values in problem chips are different from the good ones!

BR!

Hello,

many thanks for the update. Maybe when looking into the calibration values of the temperature sensor, please take a look also at the other calibration values in the sense, whether they exhibit also some "unexpected" or deviating values in comparison to good units.

Best regards

Peter