Does the cRIO Scan Engine Support Offset and Shunt Calibration on the 9237 module

[viScience]

Thanks.

[Bassett_Hound]

Hi Sachsm,

The Scan Engine does support Shung Calibartion for the 9237 and you can find the VI to support it here:

C:\Program Files\National Instruments\LabVIEW 8.6\examples\CompactRIO\Shared\Set Shunt Calibration (Scan Interface) .vi

Currently it does not support the Offset feature but it is a feature we would like to support in the future.

Let me know if you have questions.

Thanks,

Basset Hound

[viScience]

Well, that is really a kicker.  So far in my testing I have observed nearly a 1% delta in zero offset across channels on a 9237 module. 

[viScience]

Could you supply some operating instructions for the Shunt Cal. vi?  I had imagined that such a vi would allow you to enable and disable

the application of the internal shunt resistance.  Also another question regarding the zero offset...  I am trying to understand why I am seeing a 1%

ch-ch delta in ambient and shunt cal measurements using a full bridge pressure transducer on a 9237.  As I move from ch1..ch4 I see a 1% error

across the channels.  I am reading the data out as fixed point so I presume it should have incorporated the internal offset calibration values from the

factory.  Could you explain what difference, if any, is there between the factory calibrated offset and the value you would get if you enabled the

zero offset programmatically on the 9237?

Thanks very much for your help.

[Robbob]

Sachsm,

The shunt calibration VI is used as any other subvi, put it down, input the modules and slots, and run.  It can be 'deactivated' simply by making the calibration zero.

Taking a look at the 9237 user manual, I found this link: http://digital.ni.com/public.nsf/websearch/57BEE29D9B9331BE482573530078D386?OpenDocument

Your offsets are affected by what sensors are plugged in.  Unplugging your sensors and putting them in different channels has a small effect on the other channels, so you should calibrate the channels only after you have your setup how you want it.

Message Edited by Rob_K on 09-15-2008 04:43 PM

[viScience]

Really now, what is the point in trying to answer a question that you did not bother to read carefully?

The application note you mentioned pertains to half bridge configurations, in my case

I am using a full bridge pressure transducer.  The Shunt Calibration vi that I was refering to is the

Set Shunt Calibration (Scan Interface).vi for the cRIO when using the scan engine.  Since there

was no documentation it was unclear what value 'shunt value' I am suppose to be inputing and where this information is being stored.

My problem as I originally stated was that I am clearly seeing a internal zero offset in the 9237 module, not

the transducer.  I was asking if this type of amplifier zero offset is normally calibrated out at the

factory.  The scan engine, unfortunately, does not support offset calibration.  Since it appears to

be a significant source of error (>1.5%) then I will have to write a utility that performs this function

and records the zero offset in a data file which will then have to been ftp'd to the target.  I was hoping

for a simpler solution.

[Bassett_Hound]

Sachsm,

The documentation for the Set Shunt Calibration (Scan Interface).vi should have been in the VI help but it looks like it didn't make it so I'll make sure it gets updated.  I'll double check my understanding tomorrow morning but the documentation should be the following for the inputs of this VI:

slot -  Specifies the chassis slot of the strain module (9237, 9235, 9236 ). Valid values are 1 through N, where N is the number of slots in the chassis.

channel - Specifies the channel to be affected.  Valid values are 0-N where N is the channel for shunt calibration.

value - Turns Shunt calibration on or off. Valid values are 0 or 1 where a value of 1 will turn shunt calibration on for the given slot and channel.

Now, for your orginal question about offsets on the AI inputs, here are my addtional thoughts:

Unlike our SCXI products, the NI 9237 does not have an internal hardware nulling circuitry because its input range is sufficiently wide so that the inputs will not saturate even with a very large initial bridge offset.  Since the 9237 does not have any hardware nulling circuitry, you have to perform offset nulling with software compenstation, and save the value to use later in the application as you mentioned.  For example, when an offset null is performed in DAQmx, the entire offset is stored in the Initial Bridge Voltage property and is stored in software.  You should see the same offset if you used the module from DAQmx, FPGA Interface, or in the Scan Interface.  With that being said, I believe what your seeing is really just the resting measurement of the module and transducer together and you should be able to zero out your measurement by performing Software Compensation.

Regarding, Offset Calibration, I think its important to make clear exactly what it is providing as the name is a little misleading.  The Offset Cal feature of the 9237 is really more like an Auto Zero used for eliminating offsets generated by an amplifier stage and doesn't provide the same behavior as offset nulling. In Auto Zero Mode, the 9237 shorts the input channel to ground and subtracts the obtained measurement from all subsequent samples. Performing an auto zero is a quick calibration technique that takes place at the beginning of a scan.  A typical use of the Offset calibration feature would be to set it true at the begining of acquiring data and then leaving it on indefinitely.

Personally, I do not bother enabling offset cal in my FPGA applications because the measurement of the sensor at rest takes into account all sensor and module offsets.

And finally, if your looking for some information on how to perform shunt calibration, I like to refer to the example VI located here:

C:\Program Files\National Instruments\LabVIEW 8.6\examples\CompactRIO\Module Specific\NI 9235\NI 9235 Getting Started

As it has walks through the programming steps of setting up Shunt cal and applying the Gain Factors.  Its written for LabVIEW FPGA but it should be easy to use with Scan Interface.

Hope that helps a little bit,

Let me know if you have questions and I'll help where I can or bug the right engineer.

Basset Hound

[viScience]

Steve - Thanks for the detailed response, I think I will take your suggestion and simply offer a maintanence utility that will zero any analog input channel (end to end)

in my system.   I have already created a component that provides inline filtering, mx+b scaling and now zero offset for any scan engine analog I/O.  I was originally

thinking I would also include the Offset Calibration but I will wait until it comes out in the next version of the scan engine 🙂

P.S.  You might want to review the Getting Started Examples you refered to...it seems there is a significant bug in the code, see attachment.

[Bassett_Hound]

Glad the post helped out.  If you have other feedback on features need for the Scan Engine please let me know and I'll see that they get prioritized for our development.

I guess that VI really should say take Lots of samples and use just one of them.  Thanks for the catch and I'll make sure it gets updated to average values.

Steve

[viScience]

Hi Steve,

Here is my list so far...

Programmable Scaling

Import/Export (Autoloading config)

Interop with some sort of CVT and the DSC (It would also be nice if tag names could appear as block diagram rings.)

Waveform Mode (Streaming or Block)

LP Filtering

Zero Offset (Both OffsetC and End to End)

Most of all, if possible without a NDA, to know what feature are most likely for the the 8.6.1 release.