temperature compensation for pressure

[hemnath]

I have pressure sensor and calibrated for 0 to 10 bar at 25 deg C. Also i have a temperature sensor for temperature measurement. Since the pressure value gets changed with varying the temperature. Now i want to do temperature compensation.

Temperature coefficient of sensor which I am using is 0.02%F.S./ ℃

With this coefficient and measuring the temperature using temp. sensor. Is it possible to derive the equation to compensate for different temperature?

Or should I have to calibrate the pressure sensor at different temperature?

Please guide me.

 

[KlausST]

Hi,

0.02%F.S./ ℃

If this is the real drift, then this is the formula, too.

Where do you see the problem?

Are you sure this is the real drift, or maybe this is the max expectable drift?

Klaus

 

[hemnath]

Are you sure this is the real drift, or maybe this is the max expectable drift?
Klaus

This is the drift specified in the datasheet. I assume it is the max. drift.

So, to measure the real drift I have to measure the output of the sensor at different temp. Is this the only possible way?

 

[KlausST]

Hi,

isn´t the pressure soensor a bridge configuration?

If so, then I don´t understand the drift specification.
--> Are you sure you understand the datasheet correctlly?
Maybe you misinterpreted the "total_resistance_drift" with "ouput_voltage_drift" or similar....

Simplest way (as so often) --> give us (a link to) the datasheet.

Klaus

 

[hemnath]

Attached the sensor datasheet.

I mean. When the sensor is subjected to different temperature, its output voltage changes. thus changing the display value.

Example:
lets say, at 15 deg C, Input Pressure = 10 bar, microcontroller processed and displayed: 10.00 bar
at 20 deg C, Input Pressure = 10 bar, microcontroller processed and displayed: 10.02 bar
at 25 deg C, Input Pressure = 10 bar, microcontroller processed and displayed: 10.04 bar
at 30 deg C, Input Pressure = 10 bar, microcontroller processed and displayed: 10.08 bar

Above example is just for reference only.

But I want to display as, Temperature 15 to 30 deg C, Input pressure = 10 bar, microcontroller processed and displayed: 10.00 bar

 

[KlausST]

Hi,

lets say, at 15 deg C, Input Pressure = 10 bar, microcontroller processed and displayed: 10.00 bar
at 20 deg C, Input Pressure = 10 bar, microcontroller processed and displayed: 10.02 bar
at 25 deg C, Input Pressure = 10 bar, microcontroller processed and displayed: 10.04 bar
at 30 deg C, Input Pressure = 10 bar, microcontroller processed and displayed: 10.08 bar

is this your assumtion, or do you really see these values?

The given values show a drift of: +0.04% x reading / °C

***
Please confirm:
* You use a 10bar (F.S.) sensor
* and you measure up to this limit (without leaving headroom)

Klaus

 

[hemnath]

is just a assumption.

Have I clearly explained you my problem. Please check thread #5

 

[KlausST]

Hi,

Sadly it´s not clear to me. I´m confused.

I don´t know why your assumption value differs from the datasheet value.
* either correct the drift according your assumption (+0.04% x reading / °C)
* or correct it according the datasheet. (Span and/or offset: 0.02%F.S./ ℃ )

But have in mind: there is adifference if you correct the F.S. span value or if you correct the actually measured pressure value.

It´s clear to me that you want to correct your pressure values, but are you sure they drift at all?


Klaus

 

[FvM]

Temperature coefficient of sensor which I am using is 0.02%F.S./ ℃

With this coefficient and measuring the temperature using temp. sensor. Is it possible to derive the equation to compensate for different temperature?

There's no equation to derive, just an empirical error function that can be recorded for each sensor exemplar. The specified drift is only a maximum value, it tells nothing about the actual behaviour.

You can buy high precision pressure sensors with built-in factory calibrated digital drift compensation. It's hard to achieve a similar compensation quality in a user calibration. In addition there non-correctable errors limiting the reproducibility like hysteresis.