• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 78 subscribers
  • Views 272 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.

MSP430F2274-EP: Temperature sensor accuracy and calibration

adevries
Intellectual 945 points
Part Number: MSP430F2274-EP
Other Parts Discussed in Thread: MSP430F2274

Hello,

I am interested in using the built-in temperature sensor on the MSP430F2274-EP, and after reading the datasheet section on it, I had a couple questions. Please find them below.

1. Does the MSP430F2274-EP come with any temperature sensor calibration data in Information Memory Segment A?

2. I understand the temperature sensor offset voltage is +/-100 mV, but does the offset voltage vary with temperature? E.g. if a given MSP430F2274 had an offset voltage of 56 mV at 0 C, would it also be 56 mV at -55 C and 125 C?

3. Would a single point calibration be able to compensate for the offset voltage? If so, what would the process for a single point calibration be?

4. Does TI have any application notes or other information that would help me calculate the total error of ADC measurements? I know the internal voltage reference has certain errors, the ADC has offset and linearity errors, and the temperatures sensor also has offset and gain errors, but I don't know how to take all of this information and calculate the total possible error for ADC measurements and temperature measurements I take.

Thanks

  • 0
    Guru 26200 points

    Hello,

    check this document: SLAS614D–SEPTEMBER 2008–REVISED MAY 2011
    sections:
    22.2.8 Using the Integrated Temperature Sensor
    24.2.2.1 Temperature Sensor Calibration Data

    and related. 

  • 0 in reply to Tomasz Kocon
    Intellectual 945 points

    Hi Tom,

    Document SLAS614D is the datasheet for MSP4302274-EP and doesn't contain the sections you referenced. However, I'm assuming you meant to refer me to the sections of the user guide. I've read those sections and still have the questions I posted above. Section 22.2.8 "Using the Integrated Temperature Sensor" doesn't address any of my questions. Section 24.2.2.1 "Temperature Sensor Calibration Data" would seem to suggest that the MSP4302274-EP does contain temperature sensor calibration data, but I don't believe that's actually the case.

    Thanks

  • 0 in reply to adevries
    Guru 26200 points

    Sorry a for wrong reference.
    It should be 569SLAU144J –December 2004–Revised July 2013.

    Ad 1.)
    The temperature sensor is calibrated using the internal voltage references. Each reference voltage (1.5 V and 2.5 V) contains a measured value for two temperatures, 30°C±2°C and 85°C±2°C and are stored in the TLV structure at the respective SegmentA location (see Table 24-4)

    Ad 2.)
    TCsensor may vary from 3.44 to 3.66 mV/°C for dies from the same wafer.
    Of course, the offset would change with a temperature. The sensor is a semiconductor device. 
    The die working at a full clock speed could have 10-20 degC more than an ambient temp. 

    Ad 3.)
    Until TCsensor is unknown, a single point calibration is impossible. 
    It is a pure math.

    Ad 4.)
    Check this reference: ti.com/analogrefguide
    and
    training.ti.com/ti-precision-labs-adcs

  • 0 in reply to Tomasz Kocon
    Intellectual 945 points

    Hi Tom,

    1. I also read that in the user guide, which originally made me think there were calibration values for the temperatures sensor. However, the table that describes the locations (Table 24-4) states that the calibration data is for ADC12. The MSP430F2274-EP only has an ADC10 module. Additionally, if the calibration data was present, the labels should be defined right next to the DCO calibration labels in the msp430f2274.h header file. However, I don't see them defined there. I'm looking for confirmation this means they are not defined. This is the calibration data I see:

    2. I understand that TC_Sensor may vary from 3.44 to 3.66 mv/C, and I understand that the microcontroller will heat up based on the energy its consuming. That's not what I'm asking. I'm asking about V_Offset,Sensor. The datasheet says the value can be between -100mV and 100mV across all recommended temperatures and voltages. What is *doesn't* tell me is if V_Offset,Sensor varies as a function of the ambient temperature. 

    In other words, if I could determine a given MSP430F2274 had an offset voltage of 56 mV at 0 C, would it also be 56 mV at -55 C and 125 C? Or could V_Offset,Sensor change to another value at hotter and colder temperatures?

    3. That's not true. This forum post talks about using a single point calibration for greater accuracy: https://e2e.ti.com/support/microcontrollers/msp430/f/166/t/873088 . I understand that a single point of calibration won't give an exact value for TCsensor, but it will allow for significantly more accurate temperature measurements compared to an uncalibrated sensor.

    4. Thanks, I'll look into those resources and see if they answer my question.

    Thanks

    Thanks

  • 0 in reply to adevries
    Guru 26200 points

    Hi,

    ad 1)

    adevries said:
    The MSP430F2274-EP only has an ADC10 module.

    Left shift by 2 bits your ADC(raw) value.
    Check header files for other MSPs from the MSP430F2xx family.

    ad 2)

    adevries said:
    I'm asking about V_Offset,Sensor.

    Datasheet does not provide it. 
    V_Offset is a non linear function of a temperature (temp of die, not ambient).
    Without a temp of die you cannot calculate a temp  of V_offset and vice versa.
    Would you like to calculate exponents on this tiny chip?
    You have this sensor for less than one cent. 
    Do not expect to much. 
    It is not a measurement tools, it is for informational purposes.

    ad 3)
    Proposed one point method is worse than the two point calibration method presented in the User's Guide.
    The post pointed out by you, says that the CAL data in TLV is unreliable.

**Attention** This is a public forum