• State Resolved
  • Locked Locked
  • Replies 1 reply
  • Subscribers 30 subscribers
  • Views 394 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Facing issue to measure body skin temperature using TMP006

hiteshgambhava
Intellectual 475 points
Other Parts Discussed in Thread: TMP006

Hi,

We are using TMP006 temp sensor to measure body skin temperature but We are not able to get stable and reliable results from sensor. We followed layout guidelines in PCB properly as mentioned in TI document.

We have made sure about the Conical field of view to just see the object for all the experiments. The distance from outer surface of plastic enclosure (PLA) to the top surface of sensor is 1.6mm. We calculated the +/- 70 degree viewing angle point of view, required opening diameter as around 5.8mm. We have kept 8mm diameter opening above sensor. We haven't use any polyethylene material in between sensor and object. Sensor directly seeing object.

After all these points taken care, we are still facing accuracy and reliability problem. What could be issue?

 

Thanks.

Hitesh Gambhava

  • 0
    TI__Expert 3370 points

    Hi Hiteshi,

    THe TMP006 is responsive to all forms of thermal energy - radiation, conduction and convection. To ensure optimum performance, it is necessary to ensure that the thermal energy flow to the TMP006 is dominated by radiation and not by conduction or convection.

    First step is to verify the noise performance of the system. Collect approx 50 samplesof the sensor voltage, register 00h at 1 sec conversion rate (reg02h, bits D11-D9). Plot the data vs time to verify there is no long term system drift; if system drift is suspected then fit a line to the data of the form y=a x Vsensor) + b. Then generate a set of drift corrected sensor values

                 V sensor_corrected = V sensor_measured - (a x V sensor_measured + b)

    THe data set V sensor_corrected should be centered around zero (mean = 0) and have no slope when plotted against time.  Next calculate the standard deviation of V sensor_corrected. The standard deviation should be less than 1uV.

    If the standard deviation is >1uV, then it is likely that conduction or convection effects are present. To check for convection effects, try placing a sheet of polyethylene over the hole in the case and remeasure the sensor noise as described above. Conduction may be due to compenents on the PCB which are generating significnt power or if a fan is present in the product,

    I hope this is helpful in improving your system performance,

    Regards,

    Werner Metz, Ph. D.

    Systems Engineer, IRThermal Products