MPX5100DP: strange reading

[andikun]

Dear NXP tech.
I just bought an mpx5100dp sensor. I tried to read the datasheet but I don't really understand. Especially in the sensitivity which shows 45mV/kPa.

I tried to do a test to read analog arduino for each of the 900 data that has been read by five(5) MPX5100DP sensor. in free air. The following is a snippet of the Arduino reading for each type of sensors,

dp1	dp2	dp3	dp4	dp5
30	32	34	29	30
30	32	34	29	31
29	31	33	30	32
30	32	34	29	29
29	31	34	29	30
30	32	33	30	30
29.88338	31.45304	33.57585	29.56553	30.66047

*The last row is the average
Question:
1. Why can each sensor read differently?
2. Am I justified in subtracting each sensor output by its integer average? even if I subtract it then some will not produce a zero value. Does the company have a solution for this problem?

I need these sensors to do readings in a sample in my research so I try to make the resulting data logical. But I think this difference is illogical even though the sensors are placed in the same place at a very close distance.

Thanks.

[diazmarin09 · Accepted solution]

Hello Andhy,

I hope all is great with you. Thank you for using the NXP communities.

In this case, the sensitivity refers to the voltage output for every kPa.

Do you mean that you are reading the atmospheric pressure? What is your local atmospheric pressure?

In this case, I do recommend performing an auto-zero.

I believe that this is caused for the offset error. Sources of offset errors are due to device to device offset variation (trim errors), mechanical stresses (mounting stresses), shifts due to temperature and aging. Performing auto-zero calibration technique will greatly reduce these errors. This technique can easily be implemented when an integrated pressure sensor is interfaced to an A/D and a microcontroller.

I hope this information helps.

Regards,

David