Not enough information, are you using a commercial sensor which comes without the bridge, and putting the bridge around it. In this case the device should have been characterized by the manufacturer. Or are you using a custom in house sensor and putting the bridge around it, in that case you are probably best to do a temperature run of your system with four precision fixed resistors making up a bridge, and checking the performance of the electronics. Then substitute the sensor for one of the bridge arms and see what happens. This should tell you what the characteristics of the sensor are. You then have to compensate, typically on the diagonally opposite arm of the bridge. If the characteristic is linear or quadratic, you should be able to get a reasonable fit, otherwise it has to be a mean squared fit.
In doing temperature measurements, are you using a proper temperature controlled cabinet with stirred air, and giving the unit time to reach equilibrium after each temperature step. Have you done an error budget for the whole system, excluding the pressure sensor. To do a temperature run takes up to half a day, air pressure might change significantly over that period. Why are you changing the gain, it introduces another variable. Pots also have a terrible temperature coefficient unless you use foil types. It is common practice to apply the reference voltage to the output stage of an instrumentation amp to compensate for the temperature coefficient of the reference, your error budget should determine if that is needed.
Probably not much help to you, you have probably been through these issues already, but best I can do with the limited info supplied.
There is quite a bit of info on the web about this topic, have you checked any web sources out?
cheers,
Richard
In doing temperature measurements, are you using a proper temperature controlled cabinet with stirred air, and giving the unit time to reach equilibrium after each temperature step. Have you done an error budget for the whole system, excluding the pressure sensor. To do a temperature run takes up to half a day, air pressure might change significantly over that period. Why are you changing the gain, it introduces another variable. Pots also have a terrible temperature coefficient unless you use foil types. It is common practice to apply the reference voltage to the output stage of an instrumentation amp to compensate for the temperature coefficient of the reference, your error budget should determine if that is needed.
Probably not much help to you, you have probably been through these issues already, but best I can do with the limited info supplied.
There is quite a bit of info on the web about this topic, have you checked any web sources out?
cheers,
Richard
