MPXxx6115A Accuracy Specification Clarification

[david_dobrowols]

I am looking for some clarification on the accuracy specification for the MPXxx6115A pressure transducer series. The datasheet gives a transfer function with a +/- error term and also a plot of typical, min and max operating curves. My question is if I am trying to consider error at any given point, do I need to assume that the transducer is non-linear such that you could have an operating curve that is actually alternating between min and max error? See red line in example below illustrating what I mean. 

Or does the transducer behave linearly somewhere in that error band as the example below? 

The difference in the two assumptions drastically changes my analysis of our products theoretical error after we calibrate it, so any clarification I could get on this specification would be appreciated.

Thanks!

[reyes]

Hi David,

You first assumption is correct, the transducer is non-linear such that you could have an operating curve that is actually alternating between min and max error.

Regards,

Jose

[david_dobrowols]

Hi Jose, 

Thanks for the reply. I am thinking that while the first case is possible based on the specs, it would be unlikely to occur (please correct me if that is incorrect). Does NXP have any data indicating how likely the first condition is to occur? Or any data that indicates typical expected linearity of these sensors? 

Thanks, 

David

[reyes]

Hi David,

You are correct, first case is possible based on the specs, but it would be unlikely to occur. We do not have any data indicating how likely the first condition is to occur, or any data that indicates typical expected linearity of these sensors other than the mentioned in the datasheet, however, in order to occur this condition, the error factors need to be very variant, these factors includes: Pressure Error (not likely to change during time), Temperature Factor (would be a problem if varying below 0C or above 85C), and Voltage supply (output of the sensor is ratiometric with the input voltage, so changes in the input, would cause changes in the output).

Regards,

Jose

[david_dobrowols]

Hi Jose, 

Thanks, this is all great information to assist us here. In our application, our operating range is 16-30C, so temperature error will always be 1 per the specs. Our device also monitors input voltage to the transducer and calculates the pressure by referencing the output from the transducer to the measured input voltage, so I don't think input voltage will affect our error (at least that is what my calculations are showing). 

You mentioned that pressure error is not likely to change over time, but I think the key question for us is can it vary across the transducer range? What I am still unclear of for pressure error is if that is a typical characteristic (i.e. a specific transducer that reads low will read low across the range or have some linear gain/loss on the error as pressure increases) or if the pressure error term could be both alternating high and low resulting in the operating curve that I am describing. 

Thanks, 

David

[reyes]

Hi David,

It is not typical that the sensor’s pressure error vary across the transducer range.

Pressure error term could be both alternating high and low resulting in the operating curve as you described, this pressure error is in the range of ± 1.725 kPa and cannot be eliminated. It could be a linear or curved response, however, it is minimum and should not impact your application.

Regards,

Jose