About voltage measured from a high input resistance sensor with PCI 6251

[木虫]

Hi,

     I  have a problem about collecting Voltage through a high input resistance sensor with PCI 6251.

     According to "Field Wiring and Noise Considerations for Analog Signals", I put two 20K resistors as the bias resistor (in DIFF mode). But I found the data is totally different from the voltage read from a hand-held DMM (I think the input resistance of the DMM is about 1M), the data is also not stable. For example, the voltage is about 0.2V on the DMM (this voltage is correct), but the it is about 0.03V (not stable)  measured by the 6251. 

      I would like to know that whether the input resistance of the 6251 decreased after two bias resistors were connected between DIFF channel and the AI GND? Whether I can remove the two bias resistor?

      Thanks!

[johnsold]

The bias resistors are effectively in parallel with the high resistance sensor.  Note that the diagram below does not include the input resistance of the AI device (or DMM).

Since adding the bias resistors reduces the signal by almost a factor of ten, the source resistance must be on the order of 400 kohms. So, even a 1 Mohm DMM will have a significant error.

The bias resistors should be larger than the source resistance by a factor greater than the reciprocal of the acceptable error percentage.  They also must be small enough so that the bias currents needed by the inputs can flow without introducing significant errors.  I suspect that you may be close to the region where no values can satisfy both criteria.

A better solution may be to put an instrumentation amplifier near the sensor and then send a low source impedance signal to the measuring devices.

Lynn

[木虫]

Hi, Lynn, thanks for your help. I will think about advice and try it.