Pressure measurement using MPX2102

[lowpowermcu]

Hi all,

I am interfacing a pic microcontroller controller to MPX2102 pressure sensor and as you know the atmospheric pressure is equal to
Atmospheric pressure = Po + measuresured pressure by sensor
with Po is the pressure at sea level and 25 °C.
I want to have the best accuracy so I an thinking to add a calibration phase:
1. first phase: in which user provides the Po (he can refere to some websites).
1. second phase: in which user provides a known pressure. This is to compensate the gain error

Am I doing things corectly, what do you think?

PS: ofset can be compensated in first phase

Thanks & Regards,
MCU Lüfter

 

[FvM]

It's not exactly clear what you are asking for. To perform an atmospheric pressure (barometric) measurement, you'll need an absolute pressure sensor (MPX2102Ax type). P0 of an absolute pressure sensor is vacuum (0 bar absolute). It most cases, a single point (span) calibration will be sufficient.

 

[lowpowermcu]

Hi FvM,

My sensor is MPX2102ASX and according to research this sensor provides the relative pressure and not the absolute one.
As an example the sensor provides about 272.72mB so I need to add the atmospheric pressure at sea level which is equal to ~1013mB
Consequently the pressure where I am is 1013 + 272.72mB
I hope it is clearer now and thanks for your help

 

[FvM]

I don't have the part, thus I can't verify or disprove your information. The datasheet states clearly that MPX2102ASX is an absolute sensor.

Generally absolute and relative can't take each others place.

 

[lowpowermcu]

Hi FvM,

You are right the MPX3202 is an absolute pressure sensor.
I want to use this sensor to make an altimeter but I didn't succed to have the required accuracy.
I have read 210 samples at ambiant pressure in Input variable(I am using scialb; see attachement)
then I computed the pressure using the formula pressure = input/13.2e-6; // MPX2102 sensitivity when powered by 3.3V is 3.3 * 40mV / 10V = 13.2µV/mB
and I found that the minimum pressure is 1118.1724 mB
and the max pressure is 1119.8996 mB
which givres a spread of max pressure - minimum pressure = 1.7271742 mB
This spread of 1.7271742 mB is not good to have a 1-meter precise altimeter if the air pressure decreases by about 1mb each 8m because I need to have no more than 1/8 mB precision.
Do you know if there is a solution to use the MPX2102 as an altimeter and how to filter out samples in order to get 1/8 (0.125) mB precision?
I hope I am clear.

Thanks

 

[FvM]

According to my own experience with silicon pressure sensors and data sheet information, I don't see why the sensor should have mB orde of magnitude short term uncertainty. The sensor will expose temperature drift over full temperature range in several 10 mB order, also a mB range pressure hysteresis according to the datasheet. But under constant conditions (thermal equilibrium) pressure reasings should be constant as well.

You didn't tell about the measurement conditions, but if they have been fairly constant, I expect amplifier noise and limited ADC resolution as dominant source of uncertainty.

 

[lowpowermcu]

Hi FvM,

The measurement conditions is as following:
1. I powered the sensor by 3.3V generated by E3648A (Agilent)
2. I connected sensour output to multimeter 34411A (Agilent)
3. I let the sensor in a constant (ambiant) pressure: left at table - no special environement
4. I sampled the sensor output 210 times
5. I compute the pressure using the forumla pressure = sensoroutput x sensitivity
No amplifier is used.
Do you know how much precision can I get with MPX2102A sensor? I didn't find the info in datasheet.

Thanks & Regards,

 

[FvM]

I would expect less "noise". May be some filtering (parallel capacitor) is required.
There's no "precision" specification. In most cases, precision will be defined by offset and scale drift and hysteresis. They can be translated into measurement accuracy.