FP-AI-100 "all" channel reads const. 0's, single ch. do not!?

[Kjell-E]

Hey!

I'm in the middle of a project here and I have a weird problem with my Fieldpoint (hereafter FP) analog in module.

It could be a hardware malfunction (I sure hope not) or a software configuration issue, I do not know.

The problem>> Choosing my FP address as FP\FP @ 192_168_0_99\FP-AI-100@3\All

ALWAYS returns 0's on all channels. If I read channels one by one,
I see bitnoise like I would expect.

This is what I've done to try and track down the error:
To elliminate my own code as part of the problem, I open up MAX and:

My System - Data Neighborhood - FieldPoint Items (FieldPoint Control
Area.iak) - FP @ 192_168_0_99 - FP-AI-100@3
Then click (select) the "All" item in the I/O Data tab
Then click "Start", the result is a constant 0 reading. Adding channels
one by one by clicking the corresponding item name will show bit noise
variations.

The same result happens when using unmodified, official FP examples
(Analog Input.vi for FP AI100 and Analog Input Express.vi for FP AI100).
The comments in the example clearly indicate that I should see 8 traces
updating (all channels).

See attached MAX screen shots to see what I'm talking about. I can not see anything out of the
ordinary in the AI100 settings. The DeadBand is set to 0 for all channels.

Also, when I was playing around in MAX I tried to define my own channel item.
I added all channels and called it Kjell but with the same result as for the default "all" item.
Furthermore, I tried some random deselection of channels, but for any item with more than one channel defined, it reads all as 0.


Any suggestions as to how to proceed with troubleshooting, or even better, how to fix this?

Thanks a bunch!
Kjell

[NathanT]

Hello Kjell,
      Have you tried wiring a know voltage source to the AI-100?  Also try resetting your FP bank to factory default settings.  Do you have any other AI modules to test?  Does the behavior persist with just a single FP Read on your block diagram?  Let us know if these steps help.

Cheers,

NathanT

[Kjell-E]

EDIT: Also added the Fieldpoint AIK file so people can review my settings.
/end of edit

Tried with known input voltages hardwired to terminals,

Tried resetting to default values (also tried fiddling with any and all settings there is)

Read once, read in a loop, same result.

------Updates--------
I opened up the subvi's responsible for reading the fp device all the way down to where it opens a callback to the dll file. I changed the default value of the array at the point prior to the callback and watched with probes and breakpoints how it was a non-zero (set by me) value into the callback function, and it came out from the callback function as a 0. (If the item is a single-channel item, it comes back as the actual reference voltage I put in on the AI module).

This tells me several important things>> The error is not in any of the vi's it must be in the callback file, which is given as:
\LabVIEW 8.0\vi.lib\FieldPoint\subVIs\FPLVMgr.dll


At this point I have also excluded the error to be caused by a hardware error in the module, based on the fact that a LabWINDOW CVI code opens an FP reference to read all channels from that device just fine. In the CVI code it gets an array of proper values for all channels.

This again seems to isolate the dll file I mentioned above as a likely source of the error since the CVI does not use the same dll files for callback functions.

I have included two zip files, one contains a bitmap image of the CVI code snippet that reads and displays data from the AI input buffer, the other is the full CVI project with source code and all, if anyone wants to peak around in it.

Again, it seems to me that the problem must be in FPLVMgr.dll or in one of its dependant files. Since I am not an NI employee or engineer, I do not have access to pursue this issue any deeper into the code.

Message Edited by Kjell-Edmund on 02-08-2007 06:27 PM

Message Edited by Kjell-Edmund on 02-08-2007 06:28 PM

[NathanT]

Kjell,
      Thank you for the in-depth troubleshooting steps.

        Please do the following:
1.  Put the controller in safemode
2.  FTP to the controller and copy the entire contents to your Windows machine and attach it to this thread
3.  Generate a MAX report file and attach it to this thread.  Directions can be found here. 
4.  Install or Repair the most recent version of the FieldPoint drivers. 
5.  Update the controllers firmware
6.  Format the controller and reinstall software onto it

Thanks,

NathanT

[Kjell-E]

Thank you!

I will try that, however, I am hesitant to make any software changes to either controller or the windows drivers at this time. The reason is that we are 1 week away from completing an extended period of testing, and any downtime as a result of a potential complication during this operation is considered to be a completely un-acceptable risk.

Once the current testing phase is over, I will proceed to reinstall the FP controller software, drivers and firmware.

I hope to have the MAX report generated and posted here before the end of the weekend.

I likewise appreciate the suggestions and especially the links provided.

Thank you!

Kjell

[NathanT]

No problem Kjell,
        I can appreciate that you don't want to do anything too drastic with your deadline closing in on you.  A workaround for this issue would be to iterate through each channel of the AI-100 and just put those values into an array.  I look forward to seeing your MAX report and files from your controller. 

Good luck with the rest of your testing,

NathanT

[Kjell-E]

Attached is the MAX report (custom) in a zip archive.

I've got an observation on the rest of your instructions though. Since I have an FP1600 ethernet "controller", as opposed to an RT controller, is that not a "dumb" controller? That is, I can not ftp onto it..

Also, I do not know how to put it in safemode, perhaps again this is something only available in RT controllers?

As far as Firmware and driver updates, I'm still holding off on those until next week.

I appreciate the help so far, and hope that you will keep an eye on this thread for updates.

Kjell

[NathanT]

Hello Kjell,
      Thanks for posting the MAX report for me to look at.  I did some more searching and found another case of this behavior which results from upgrading LabVIEW and the FieldPoint Drivers and not upgrading the firmware on the 1600.  I am very certain that upgrading the firmware will fix the problem behavior that you have been experiencing.

To upgrade the firmware:

"After installing LabVIEW 8, and consequently the FieldPoint 5.0 driver, you will need to upgrade the firmware on the controller to version 502. This version is shipped with FieldPoint 5.0, and if the default installation path was used, it can be found in c:\Program Files\National Instruments\FieldPoint\Firmware\Ethernet\FP-1600.

To upgrade your firmware, perform the following steps:

1. Open Measurement and Automation Explorer (MAX)
2. Select your device under remote systems
3. Select the firmware tab
4. Browse to the file in the above mentioned directory named FP1600Eth0502.fpf
5. Select upload
6. Reset the device

Once the firmware has been installed, and the device has been reset, the FieldPoint read should return the expected values."

Cheers,

NathanT

[Kjell-E]

Updating the Firmware of the Fieldpoint FP-1600 ethernet module fixed the problems I had with the AI-100 modules not returning all the channels properly for the "all" channel item.

Thank you for finding and pointing out the solution to me!

Updating the firmware did not break compatibility with our existing DAQ software, so in hindsight, I could have done it a week and a half ago.

Cheers,

Kjell

[NathanT]

Kjell,
      I am glad to hear that updating the firmware took care of the issue.  I hope the rest of your application goes well.

Have a great week,

NathanT