DAQPad 6016 displaying incorrect data

[gd420mj]

I use a Daqpad 6016 to acquire a variety of analog inputs. It has tasks setup in Measurement and Automation Explorer. It uses NI-DAQmx tasks. We use an exe program that was compiled in Lab Windows CVI 8.0. It is a stand alone program that runs a test stand. The compiled program has not been changed in years.

Recently, the daqpad stopped acquiring data all together. It wasn't even seen by MAX. (measurement and automation explorer). This is actually the second instance of this occurring.

In 2006, a year after the system was setup, the 6016 failed to be seen in MAX. I re-installed the 6016 device through windows hardware, and re-installed MAX 4.0, DAQmx 8.0, Traditional NI-DAQ 7.4.1 from the disks supplied with device. This solved the problem. I also had to re-import my DAQmx tasks. (saved as configData.nce)

Recently, in early September 2011, the 6016 stopped working again and was not seen in MAX. We re-installed the device through windows hardware manager.

Then we attempted to re-install the daq softwares listed above. We had various problems with the install. But eventually, after a few attempts and installing each part one at a time, we had success.

THe 6016 was now seen in MAX. We re-imported the DAQmx tasks and everything seemed to be ok.

We ran our exe program and noticed some errors. some of the data acquired was incorrect. One analog input seemed to be double what normally should be. Another analog input, from a temperature transducer signal conditioner (analog voltage) that is normally stable and increases gradually, was constantly changing (flickering between small range i.e. Instead of displaying 90 deg. it would flicker between 70 and 110).

This behavior of these channels is not normal. Nothing else in the system has changed. Non of the calibrations of any of the sensors, and non of the wiring.

What could be causing the 6016 to display incorrect data, or re-scale it or be unstable?

The DAQconfigData.nce file as far as I know, has been unchanged since 2006. The file we imported should be the original. However, when viewing each task in MAX, there are is no specific scaling defined. Only differences between tasks is the Input range. Some are 0-10 some are 1-5. This is not the problem. These settings are all constant and have not been changed since 2006.

Computer is running Windows XP. There is a 6508 data acquisition card installed also, running two relay boards. No problems with that. There is a USB printer, mouse, and keyboard installed as well. All basic Dell.

The test system is installed in England and we are providing support over the phone from the USA.

Attached are two photos. One is of a test panel we can run with the test stand program. It shows raw values from the 6016 and the calculated value from the exe file.

THe other photo is the basic items we installed for the daq.

[gd420mj]

Example of error.

One analog input channel of DAQPad 6016:

Physical voltage input as measured with DMM is 1.1 VDC. Out test program displays 2.79 VDC. No value for any analog inputs in max

[gd420mj]

Test panels on MAX read correct voltages for Analog inputs. Values in CVI compiled program are incorrect.

[Kyle_K]

If you are getting the correct values from MAX test panels, then the problem is with either CVI or the code. We can test the CVI possability by running an example that uses some of your channels. If the example gives you accurate data then the problem must reside within the code. 

Also, what kind of difficulties did you have with the reinstallation? Is it possible that these difficulties are what led to this issue?

Kyle K.

[gd420mj]

Hello Kyle K,

Thanks for the reply. I had a lengthy phone discussion with NI Tech support, Michael M. last week about this. (I posted here before I called tech support, thinking I would not be able to get help, but its god to keep this going for others to see).

Michael also suggested to create a sample program to collect data and test CVI. I won't be able to perform this until I travel to the equipment which is in the UK and I am here in the US.

What type of problems could occur within CVI that would cause this? and how can they be corrected?

Also, how could the code develop a problem? It was last compiled in CVI in 2006. The exe file they use dates back to 2006. There have been no changes since then. LabWindows has been open and there was a test program compiled in 2009, but again, no problems until September 2011.

Here is a list of NI software installed on a Dell Optiplex GX520, P4, 3Ghz, 512 M Ram. Windows XP

NI Instrument I/O Assistant

NI LabView Runtime Engine 7.0

NI LabView Runtime Engine 7.1.1

NI LabWindows/CVI 8.0 FDS Package

NI LabWindows/CVI 8.0 Run-Time Engine

NI MAX 4.0

NI Meas. Studio VS2003 7.1.0.308

NI PXI platform services for Labview realtime 1.5.1

NI PXI platform services for Windows 1.5.1

NI Session Manager 3.5

NI VI Logger 2.0.1

NI 488.2 2.4.3

NI DAQmx 8.0

NI VISA 3.4

NI Visa Runtime 3.4

NI Visa Server 3.4

Traditional NI-Daq 7.4.1 (Legacy)

We Re-installed NI-DAQmx and whatever else was required to go with it. I don't remember all the components.(Probably MAX, Legacy, 488.2) I believe there were some errors like a file not found, but everything was successfully installed. We had to do a few software components, one at a time.

.

Ahy additional suggestions would be greatly appreciated. Thank You

[Kyle_K]

I don't know of any reason why either your code or CVI would change. All we can do is continue to check values at different points in the data and narrow down where it changes from what we expect. Make sure that everything you installed is the same version you used before. 

After further discussion with some of my colleagues, we realize that this could also be an issue with a custom scale. If your program was written to utilize a custom scale and your new setup does not contain that same scale, that would explain why the values appear incorrectly. To correct this, you would have to manually recreate that scale in MAX. 

Kyle K.

[gd420mj]

No scaling was performed in MAX. THe c-code read all raw values and performed all calculations. The photo I attached, test-panel-A, shows that we read the raw value from the DAQpad-6016 and then our calcualted value that would be displayed in the normal test interface window.

Two things worth mentioning:

1) ,When reading one particular Ananlog Input, the channel read stable in MAX. When the cvi program was run, the channel would fluctuate and even occasionally display a negative value in the raw data entry of our test panel. The calculated values appear correct.

2) For a different Analog Input for example, we would measure 1.1 VDC at the 6016 terminal block, and test panel would display 2.80 VDC in the raw column. Only problem with this channel, we could not measure it in MAX, since the voltage was only present during cvi program operation. I suppose there could still be a problem with the DAQ 6016.

THere appears to be about 3 or 4 channels giving us problems and displaying incorrect data. At the time on troubleshooting, those channels could not be tested without the cvi program running. With this in mind, and the fact that the cvi program displays the raw value read from the 6016, could there be a problem with the 6016? Keeping in mind that some analog inputs are working correctly are are stable at all times.

Thanks,

Steve W

Breeze-Eastern

[gd420mj]

NI,

I need information on the following:

If choose to re-install LabWIndows CVI 8.0, what problems will I ecounter?

Will I have to call NI for activation? (We activation code that came with software) This computer is the original computer it was installed on.

Will I encounter any problems with running my compiled cvi code program? WIll I have set up any addtional items within CVI? or should it be all contained?

Please reply as soon as you can. Thank You.

[Kyle_K]

You should not encounter any problems from reinstalling LabWindows CVI. The activation code you have and the already compiled code will both work as long as you install the same version of LabWindows that was previously on the machine. 

Kyle K.