I have been trying to use the DAQmx SCXI Calibration VI's to perform a full external calibration of one of our SCXI-1520 modules. I have read through the archived tutorial here, and I have read through the SCXI-1520 Calibration procedure here. I have been able to perform the verification procedure for each gain and channel of the module, but I get error -200546 (shown in the first file below) when I commit to the DAQmx Close External Calibration VI.
I have attached my code, along with a comma-delimited text file that is used to pass the gains to the Setup Calibration VI. This sub-VI requires as inputs the verification arrays that are generated by the verification procedure (these are 3D arrays of size: 147 rows, 3 columns, 8 depth).
Also, does anyone know of a way to adjust the excitation calibration constants with modern DAQmx? There is no mention of this procedure in the archived tutorial, but these constants certainly exist and are referenced by the 1520 calibration manual.
In general, calibration problems tend to be complex. If you have an active SSP, you may consider contacting us at (866) 275-6964 to open a service request.
For assistance through this thread however, it may be helpful to get more details. The error message you're getting is related to the signal being provided from your calibration hardware. The reading is out of the range expected by the calibration VI, or is excessively noisy. Can you give some further details about the hardware you are using the provide the calibration signal, and how it is wired to the SCXI-1520?
After talking with a coworker today we realized that I had been passing the incorrect reading for the "Measured Voltage" input of the Calibration VI. I had been passing the scaled voltage reading from the SCXI-1520 rather than the raw voltage reading. I have changed the hardware setup to collect the raw voltage reading from the backplane of the module with the DMM. I will provide an update on whether this worked once I have a chance to run through the procedure again.
For the calibrator we are using an Agilent 3245A, and the DMM is an Agilent 3458A. Each of these instruments have an accuracy that is greater than the recommended instruments by the 1520 calibration procedure. The output of the calibrator is connected by a shielded BNC cable with a pigtail to the screw terminals of a TBX-96, which is connected to the module by an SH 96-96 cable.
I have had the opportunity to run the calibration procedure a number of times since my last post. I am now able to complete the procedure without any errors from Labview. My problem now is that the calibration does not appear to be stable. By this I mean that the post-calibration verification does not pass for all channels and gain settings, and does not appear to have a consistent pattern to the channels that fail.
With my current setup I have been using 3 points for the external calibration. Will increasing the number of data points in this section of the procedure increase the accuracy of the calibration?
I am also still at a loss for how to adjust the excitation voltage constants, other than trying to get Legacy DAQ to work on our machine. Does anyone have any knowledge about this?
Unfortunately, due to the age of the SCXI 1520, there isn't a lot of material out there, especially for edge cases like trying to use DAQmx for calibration. The only supported process for performing your own calibration is documented in the manual you linked, using the legacy NI-DAQ driver.
The whole calibration process should only require a maximum of three different measurement points for each gain and channel. Three measurements points are required for verification, but only two for adjustment of the gain constants. They do request that you take 5000 data points for each measurement and average them to compare with the specifications table.
If you continue to get unexpected results after following the calibration process, I would recommend that you try using legacy NI-DAQ first. If you still get unexpected results, the device may be faulty. We can certainly process an RMA for diagnosis and repair if that is the case.