MAP Sensor Error Tolerance

I've just installed a 4 Bar Omnipower MAP sensor to my Toyota 86 in preparation for a tubro install.

It reads about 2kPA lower than the OEM one under atmospheric conditions (engine off, OEM reads 100.6kPA, Omnipower with supplied calibrations says 98.5kPa measured minutes apart).

- is this error within tolerance for MAP sensor?

- will it have negative effects on the system when used for turbocharged applications?

Just trying to get a feel for if this is normal or do I have a faulty sensor.

That's about .5% error which is considered normal. As you increase the range of the sensor, the absolute value of the error will increase even if they have the same percentage error. 0.5% is normal for solid-state sensors, laboratory grade sensors can often be found with 0.1% error but are not typically used in motorsport or automotive applications.

So error is measured on full scale value (2kPA over 400kPa = 0.5%). Got it.

Given I know this is reading 0.5% low should I tweak the calibration table up by 0.5% at both ends of the calibration table? This assumes the error is linear.

(I have a Bureau of Meteorology station near me - I could park next to it, log onto their website to get a live pressure reading and then scale the sensor to get the most accurate reading outside a lab bench).

Yes, you could try to calibrate your sensor (or the OEM one -- I wonder which is the most accurate?)

I say take any error in your sensor and just accept the fact that your calibration (tuning) will account for any small inaccuracies. If it reads 390 KpA, and your engine needs enough fuel for 400 KpA, how will you know if the sensor or your calibration is incorrect? In the end your engine will be calibrated for the sensors installed.

I have the data from 36 Toyota 86's that was collected on the same day, in the same conditions, and the variance in the factory MAP sensor is in the same range as you have seen.

As long as you do not change the MAP sensor after the engine has been tuned, then the variance will not be an issue, as long as the variance remains consistent. The issues occur when the sensor is changed, and the tune is not checked after doing so, so a reading of 200kPa on one sensor may be read as 180kPa on the next, leading to the tune being out. The other problem is if the sensor is not consistent in it's reading, where on one run it may read 200kPa and the next, 190kPa, even though the actual boost level doesn't change.