Xmega16E5 internal temperature sensor

Go To Last Post
4 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Anyone successfully getting meaningful results from the internal temperature sensor? I'm using code which worked fine on a xmegaA, but not on the E series. I'm getting readings about 2x what it should be. I suspect the calibrated temperature numbers in the production signatures are wrong, looks like they are based on a signed conversion. My TEMPSENSE is 0x04B7 but the A series was about double this number. I know that early E5's had an errata (before date code 1324), but mine are after that. Could anyone confirm these calibration numbers.

Thanks
Colin

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Do a search, there has been a lot written on this subject. The temperature sensor needs calibration at room temperature and even then are not brilliant. I just about manage to temperature compensate a watch crystal with on, but doing it again I'd just use an NTC thermistor.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I realize this thread is over a year old but I'm experiencing the exact same issue: https://www.avrfreaks.net/forum/xmega32e5-internal-temperature-sensor-problems

I'm using both the HOTTEMP and ROOMTEMP values from the production signatures row.  Did you ever get this resolved?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I opened a case with Atmel and got to the bottom of this.  See the solution here: https://www.avrfreaks.net/forum/x...