Get SCXI-1100 offset voltage and using multiple voltage ranges

[eer]

Hi all, 

I have a SCXI-1100 that is read by a PCI-6036E card, which is used to read a stain gage balance that has external excitation. Six strain gage channels are used using a range of -100mV to 100mV and an excitation channel with a range of -10V to 10V on the same SCXI-1100 module. Currently the strain gage channels are not accurate or reliable with high zero readings, noise and different readings on different channels.  Does this configuration lead to issues with accuracy on the millivolt channels since the gains may not be changed between each channel, even though you can still set individual ranges in DAQ Assistant?

Also, the original VI was created using LabView 7.1, which is no longer supported, used a command "calgnd" in order to get an offset voltage to correct any following readings. The current method, DAQmx, does not have an ability to request "calgnd" but does have a zero once function. So does the zero once do the same as the previous "calgnd"?

Thanks, 

Ken

[Chris-G]

Hi Ken,

To answer your second question, the calgnd function no longer exists in DAQmx. However the Auto Zero Mode property that you mention does perform this same function. You can access and set this attribute using a DAQmx Channel property node. The property can be found at Analog Input >> General Properties >> Advanced >> High Accuracy Settings >> Auto Zero Mode. See the following articles for more information:

Removing Large Initial Offset

Setting the Auto Zero Mode

To address your first question, it's difficult to say why you are seeing odd readings. Using bias resistors may help eliminate the noise. Check that the jumpers on the module are configured correctly based on your setup. See Page 2-6 of the SCXI 1100 User Manual for more information about this.

Hopefully this helps,

Chris G

[eer]

Thanks for the reply. Was hopping that there was some obvious reason for what I am getting. I am going to explain the setup and the values I am getting. 

The first issue I was getting was that the first channel read was giving unrealistic numbers 22millivolts instead of -0.7mV I read on a NI 9211, which we are trying to switch away from (sampling rate too low). I then decided to ground the first channel, ai0, and read it so that I don't get any bazaar values. This helped and not helped the situation, since it does give better readings but shows that the grounded channel has a voltage of 0.5 millivolts. So next I guess that this is an offset of some sort, but that's where things get extra bazaar. 

So with the channels set with a range of 50mV to -50mV, I was getting a value for voltage0 of about 1.5mV (show in figure), I then reversed the leads and read a new value of about 1.8mV. This behavior would suggest a offset about 1.65mV with is not what the grounded channel (voltage6) would suggest. After reviewing the other channels they also conclude that the true offset is at 1.65mV. 

So I guess my new question is; how would you correctly measure this? All I can think of quickly is to read both the positive and negative voltage and average. But there should be a better way. I have yet to look into the "DAQmx Perform Bridge Offset Nulling Calibration" and will do so tomorrow. 

Aside, the offset seemed to vary with the voltage range selected, since the voltages did move significantly between a 50mV range and 10mV range. 

Thanks,

Ken

[Chris-G]

Hi Ken,

In general, the system should be configured as a differential acquisition. Neither of the leads coming from the strain gauge should be wired to ground. They should be connected to the CH0- and CH0+ (pins 32 B and C). When configured as differential, the DAQ card simply subtracts the voltage of the negative channel from that of the positive channel. So you shouldn't have to measure the positive and negative lines independently. Just configure the acquisition mode to be differential.

If you are seeing an odd offset or high noise try to connect bias resistors to the two lines. Connecting a 10k-100kOhm resistor between the signal wire and a ground pin will sink current flowing through the line that may be causing the offset/noise. More information about connecting bias resistors to a floating signal source can be found in the Field Wiring Guide.

Chris G