Hi,
I am reading the internal temperature in LM3S9DN6 according to the errata sheet (Stellaris LM3S9DN6 RevA2 Errata) and disregarding the two first samples. At first I set it up to do 3 samples using ADC1 and SS2 but found there was too big a difference in readings for it have stabilized. So I added the fourth sample in SS2 and could see that my suspicion was justified. The fourth sample clearly showed that the readings were not yet stable.
I have done several measurements and they all give the same result (+/- normal fluctuations).
Example:
1 sample: ADC = 2162
2 sample: ADC = 2126
3 sample: ADC = 2105 (~32 celcius)
4 sample: ADC = 2088
Values are decreasing by less and less but certainly have not yet stabilized as stated in the errata sheet, and a temperature of 32 degress celcius seems low. When I touch the chip I can feel that it is warmer than a finger (probably around 37 - 40 degrees). So what can be wrong?
Is sampling done too fast? If so they ought to have mentioned this in the errata sheet...
I have tried to lower the sample rate to 125ksps and it seems to have improved it slightly - but only slightly. I still do not have stable readings (sample 3 and 4 still up to 20 apart - although 20 is not much you have to consider that to the total sum until it settles is perhaps 100-120)
The interrupt priority of the SS of the ADC is set to 0 and there are no other SS that uses a higher priority (as stated in the errata sheet).
All SS in ADC are used with priorities ranging from 0 to 3. In ADC1 the two first SS are used together with comperators and also have priority 0. The fourth SS in ADC1 is not used.
Kind regards,
Christian