LPS25HB Temperature Compensation Behavior

[kris2]

Posted on May 09, 2016 at 19:52

My colleagues and I are evaluating the LPS25HB and came across some behavior we don't understand. There are two issues we are trying to resolve right now:

• Embedded temperature compensation behavior is confusing.
• Accuracy isn't what it should be.

Specifics:

• In one of our datalogs (pressureLog_25Hz.xlsx) we see absolute pressure tracking with temperature. We expected temperature to change, but not absolute pressure. Absolute pressure is changing at a rate of about 3 hPa/C. Shouldn't the embedded temperature compensation deal with this?
• In a test we did in a car (car pressure log.xlsx), we saw temperature drop and then increase, but the pressure is almost constant if you draw a trend line through the data. It seems like embedded temperature compensation is doing something this time.
  • We need help determining why is the sensor acting differently.
• The datasheet for the part claims an absolute accuracy of +/-0.2 hPa between 20 and 60 C. So far, the data we have taken from the sensor does not meet the spec. For example, 
  • I a recent test, our local weather stations reported a sea level pressure of 1016.2 mbar.
  • The LPS25HB evaluation kit is told us the absolute pressure was 1026.5 mbar.
  • We are about 160 above sea level, which equates to a pressure difference of -6 mbar, so the LPS25HB should have reported about 1010 mbar. It was off by 16.5 mbar!

Thanks for the help!

#lps25hb

[licciardello.anto]

Posted on May 11, 2016 at 12:21

Hi,

I need further information about your test, 1. pressureLog_25Hz and 2 .car pressure log.

I suppose in both cases the environment was not controlled. Can you confirm?

I would know the settings used in both cases.

Are they the same?

What is the value of ODR? What internal average for temperature and pressure? Was the embedded filtering enabled?

In the first case, if the environment is not controlled, why do you expect that temperature changes, but not the pressure?

I need to know what is the value of voltage supply used for your test too.

Regards

Antonella

My colleagues and I are evaluating the LPS25HB and came across some behavior we don't understand. There are two issues we are trying to resolve right now:

• Embedded temperature compensation behavior is confusing.
• Accuracy isn't what it should be.

Specifics:

• In one of our datalogs (pressureLog_25Hz.xlsx) we see absolute pressure tracking with temperature. We expected temperature to change, but not absolute pressure. Absolute pressure is changing at a rate of about 3 hPa/C. Shouldn't the embedded temperature compensation deal with this?
• In a test we did in a car (car pressure log.xlsx), we saw temperature drop and then increase, but the pressure is almost constant if you draw a trend line through the data. It seems like embedded temperature compensation is doing something this time.
  • We need help determining why is the sensor acting differently.
• The datasheet for the part claims an absolute accuracy of +/-0.2 hPa between 20 and 60 C. So far, the data we have taken from the sensor does not meet the spec. For example, 
  • I a recent test, our local weather stations reported a sea level pressure of 1016.2 mbar.
  • The LPS25HB evaluation kit is told us the absolute pressure was 1026.5 mbar.
  • We are about 160 above sea level, which equates to a pressure difference of -6 mbar, so the LPS25HB should have reported about 1010 mbar. It was off by 16.5 mbar!

Thanks for the help!

[kris2]

Posted on May 17, 2016 at 01:40

Hi Antonella,

Thank you for the questions. We were not testing in an environment where we controlled for pressure, however, we were testing outside and comparing the pressure we were reading against a reliable system at a nearby major airport in real time.

Settings wise, we were using one of the evaluation boards and the ST software. More details:

Easy Configuration (Power Mode - Active, Data Rate - 25 Hz, BDU - Enabled, Pressure Resolution - Avg 512, Temp resolution - Avg 64, Embedded Filter - Disabled).

All tests use Easy Config. All tests powered by the same laptop/USB cable.

Kris

P.S. is there a way to private message you about this? The forum reply and editing tools are really awful.

[licciardello.anto]

Posted on May 18, 2016 at 12:14

Hi Kris,

thanks for reply.

Concerning the first test, the trend of pressure and temperature is not strange if the environment is not controlled.

The sensor is affected from of environmental changes and the pressure and temperature follow these changes increasing in 45 minutes (from 7.44 AM to 08.21 AM as reported in execel file).

In the second test, the drop in temperature is less  0.5 °C in about 1 minute,so the influence on the read pressure is very little.

Concerning the accuracy, you keep in mind that the value reported in the datasheet is refered to the sample not soldered.

You have to consider that there is an offset after soldering that should be in the range 0-3mbar ( working at 1.8 V) and greater than 3 mbar working with voltages greater than 1.8V.

Of course 16,5 mbar is too excessive.

Can you repeat the same measurement  with another sample/device and to check if the measurement/error is repetible?

Regards

Antonella

[davewalker9]

Posted on June 16, 2016 at 12:09

Hi

I might be a bit late replying to this thread, but I enclose some calibration data from a 16 channel sensor we have designedusing16 offLPH25 devices. The unit has been calibrated in an oven at 20 to 60 degrees, between 72000Pa and 126000Pa.

The data shows the test pressure for each channel, and the recorded value for each channel at the different temperatures. Typically we see an offset of several hundred Pascals on each sensors, and a change in gain giving an additional error of 150Pa as temperature changes from 20 degrees to 60 degrees.

Regards

________________

Attachments :

LPS25_Calibration.xlsx : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006HzEZ&d=%2Fa%2F0X0000000bJq%2F3rZWgXuGHBdWp6Zvbgiu1u8DwY2A5xpJITmfJBl1X9c&asPdf=false