• State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 82 subscribers
  • Views 111 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TMS320F28033: question of temperature sensing for the oscillator compensation

Aki Li
Expert 8225 points
Part Number: TMS320F28033

Customer would like to use the internal oscillator for the clock source, and was evaluating the oscillator compensation according to our app note SPRAB84A. They used external ADC reference voltage of 3.184V.  Since they need to get the ADC results of the temperature sensor, they wonder whether there is any difference between using internal ADC reference and external ADC reference.

For the temp sensing result, the below codes were from their one old project, and the engineer would like to understand why the result is required to multiply 3847/4096 before sending for the compensation. Could you please help check and share your comments? Thank you.

  • 0
    TI__Guru 71755 points

    Aki,

    If the customer uses an external reference to sample the temp sensor, they will need to adjust it based on the ratios of the Full Scale range.  The internal reference is set up to give a FSR of 3.3V and when the temp sensor is sampled it is based on that FSR that the voltage to temperature derivation is calculated.

    In their case the same voltage from the temp sensor would appear to be a bigger ADC value since the FSR is smaller.  It looks like this is what their old project was doing, i.e. scaling based on a new FSR(VREFHI).  3847/4096 = 0.939 or FSR of 3.099.  If customer wanted to apply the same theory to the new FSR of 3.184 they would need to change 3847 to 3951(and I think the 4096 should be 4095 since that is max code)

    There will still be some difference from using the internal reference, as our calibration coefficients also accounted for the internal reference drift over temperature(reflected in the gain error of the device ppm/degC).  It will be better than un-compensated, but may not be as robust as the original method.

    Customer can choose to switch over to the internal reference for this measurement, although it may be undesirable since there is a settling time of 1us when changing the reference.

    Best,

    Matthew