Labview 2023 can't see Data Translation DT9829

[amilford]

Labview 2023 can't see Data Translation DT9829. I need an updated driver I think. Your company doesn't seem to be on their device list.

I need to run the VIs with your device, but Labview can' see it.

[DAQman]

To use the DT9829 in LabVIEW, you must first install the OMNI CD - required software option. When you have completed this step, install the LV-Link3 software. The programming examples are installed into \National Instruments\LabView 2023\user.lib\LV-Link3\Examples\. The tool pallet can be found on the User Libraries pallet.

https://mcc.download.ni.com/#downloads/DTSoftware/Omni-CD/

https://mcc.download.ni.com/#downloads/DTSoftware/LVLink/

[amilford]

I did all that before I posted here. I used your VIs that also couldn't find you device. QuickDAQ is the only program that sees it and can use it. The program doesn't calibrate fine enough to uses for load and torque cells. I will be using this for a cycle tester that runs parts for 10K or more cycles

[DAQman]

I've installed LV-Link3 into LabVIEW 2023 without issue. To test the DT9829, use the example DtAinContinuous.vi. Because it is configured for a single-ended input, you must modify the code to set it to be differential. Next, on the front panel, reduce the sample rate to 100 and samples per channel to 10. Click the LabVIEW Run button and after a few seconds click the Device control and the DT9829 should appear. Press the front panel Start button to run the example.

[amilford]

That kind of works. I can't really get any substantial reading. I press on a load cell and the line goes flat. Why can't Labview see it as a device? I need to be able manage it in the Measurement and Automation to adjust the scaling for adjustments for different loads and torques. I tried QuickDAQ and it doesn't work. I have calibrated weights. Do I need to have power running to it besides the USB?

[DAQman]

Our devices are not integrated into Measurement & Automation Explorer. I think only NI devices are allowed. Because of this, they are configured programmatically. A load cell is a bridge-based sensor with a different example, DtAinBridgeContinuous. I've modified this example to read the first four channels and attached it below. You'll notice that the DtCreateAinChan uses an input string of "0,1,2,3" to define the channels, and DtOLSetChanBridgeOptions is repeated, one for each channel.

 

DtAinBridgeContinuousFourChannels.vi

[amilford]

Thank you, but the .vi didn't work. I tried playing around with, but to no avail. No signal shows. It's one straight red line that I can't get to move.

We'll be running our tester with the DT9829 for a day or 2 and the then have to recalibrate it again for a new force. I don't see how to calibrate your device properly. I tried using gage factor, but it isn't accurate unless you calibrate it for a constant static load. I have too many errors trying to setup the DT9829 with the calibration .vi. 

The computer that will be running this will not have Labview on it, but be running executable(s) in a run-time environment. I would prefer it to have the calibration included so I only have one executable that is flexible, but it will require a difficult program. If I can better understand how to interact with your device, I can could understand how to program with it. It's too bad your company isn't on the NI device driver list or that you didn't use the "develop driver" function in Labview.

If QuickDAQ could calibrate properly, that would be all that was needed to log the load and torque data to compile. We test the closing and opening of latches. We watch the torque for both directions and see if the closing forces changes over time. 

[DAQman]

Please provide the factory calibration sheet for the load cell and torque sensor and a schematic showing how they are connected to the DT9829. 

[amilford]

4 load cells and 4 torque cells. I wired them next to each other. L1 and T1, L2 and T2, L3 and T3, L4 and T4. See attachment:

All red wires to terminal 35 (EXC+). All black wires to terminal 34 (EXC-). 

DT9829 Future Wiring.pdf

[DAQman]

Who manufactures the load cells and torque sensors and what are their full part number? 

[amilford]

Load Cells

Manufacture: Omega Engineering Inc. 

Part Number: LC202-500

 

Torque Cells (20 years old)

Manufacture: Omega Engineering Inc.

Part Number: TQ502-125

[DAQman]

The Omega datasheet for the load cell shows a typical sensitivity rating of 2 mV/V. The actual value is on the calibration sheet it shipped with and will be slightly different. With 3 Vdc of excitation, it outputs 6 mV (2 mV/V * 3vdc) with a 500 lbs load applied. The four-channel example I provided returns volts, and to scale the mV reading to lbs, multiply the reading by 83,333.3 (500 / 6 mV). For example, 3 mV times 83,333.3 is 250 lbs. You'll need to add the code to perform this calculation. 

Omega says the torque sensor sensitivity is 1-3mV, and the actual value is provided on the sensor's calibration sheet. Assuming it's 2 mV/V, the same calculation applies. With 125 ft-lbs applied, it will output 6 mV. So the scale factor is 125 ft-lbs / 6 mV or 20,833.3. 

 

[amilford]

I tried your program. It can't see any channel. I tried the unmodified .VI that yours was based on. It can't see the channels either.

Both .VIs require initialization or identification of the board. The only .VIs that work are "..getboard.vi" and "...getboardname.vi". 

QuickDAQ almost works for what we need. 

It seems with your board that we need a program that does the following:

IdentifyBoardandandDefineChannels.vi  +  CalibrateLoadCells.vi  +  ReadandLogLoadandTorque.vi

I need your device to work before the 30 days are up. If this this can't work, I will have to return it. The NI option requires I spend $8000.00 to do the same thing, but it's with their devices.

[DAQman]

Using QuickDAQ's Configuration dialog, I set channel 0 Input Type to Bridge. This enabled the Configure and Calibrate button in column 10. Next, I ran the calibration and entered 2 for Transducer Rated Output and 500 for Transducer Capacity to mimic your load cell. Then, I ran the offset calibration with the input at zero volts. I did not run the shunt calibration. Afterward, I could apply 0 to 6mV and see the display go from 0 to 500 lbs. 

I also ran the LabVIEW sample and although it doesn't scale the voltage to LBS, I did get the correct reading when I applied 0 - 6mV. 

When you say you can't see the channel, do you mean no trace is shown on the graph? Set the y-axis to -0.01 and +0.01 volts. You should be able to squeeze the load cell and see the voltage change. It may go from 0 to 0.006 or 0 to -0.006 volts depending on how it's wired. 

If nothing works, reduce the setup to a single load cell channel, and disconnect the other sensors. Then the wiring is simply two wires to the excitation and the input channel. 

Lastly, we don't supply macro VI's that could calibrate, measure and log the data for you. That's what QuickDAQ is for.

[amilford]

QuickDAQ is ok. I have calibrated weights.. It's a lot of guessing just to get kind of close.

 

Same results as before. Labview shows no data at all. See attachment.  

No reaction no matter what. Same signal using one channel or all 4. The program doesn't seem to see the device.

I wasn't asking for a .vi to be made or given, but if the DT9829 can't do what I need it to do, then it is not the option for us.

So far Labview is useless with this and QuickDAQ is ok, but I need to send signals from the digital output on the board too. I don't see a way of using the output channels.

[DAQman]

If the channel returns 0.200 volts that usually indicates a wiring issue like a loose wire. Send me your version of the four-channel VI, and I'll give it a go here. 

When I initially tried QuickDAQ, I got 900 Lbs when I was supposed to be reading 500 Lbs.  Then the strangest thing happened. I used the DT9829 Calibration Utility and calibrated the x50 voltage range, but only for the first channel. When I returned to QuickDAQ and tried the sensor calibration again, it worked. It was spot on 500 lbs. I even tried another channel and it worked. I can't say for sure that the DT9829 Calibration Utility fixed anything. Maybe it was a coincidence. 

[amilford]

I checked the wiring before because I thought it was that, but it works in QuicDAQ. I attached the one you sent me with a few more display options. No program change.

Are you adding outside voltage beyond the USB?

My numbers are close in QuickDAQ, but they are off by .25lbs or 1lb as I larger weights. Next week I will try repeatability with our Instron. 

My other question about QuicDAQ, can I send an output signal from the digital output on the board? I see a tab, but there isn't anything on it. Is there something there with licensed versions?

DtAinBridgeContinuousFourChannels.vi

Edited April 28, 2023 by amilford

[DAQman]

There is no QuickDAQ output option for the DT9829. The output option is for devices that have waveform output. 

Attached is a new version of your LabVIEW vi. I thought it best if we worked with bigger numbers, so I added scaling and offset correction. The first channel has a multiplier of 8333 or 500 lbs / (2mV/V * 3vdc). The second channel uses 20833 or 125ft-lbs / (2mV/V * 3). The data sheet needed to be more specific, so I assumed the torque sensor sensitivity is 2mV/V. 

To test the vi, I used a voltage source with a 1k ohm resistor connected from the minus input to excitation minus. I ran both QuickDAQ and the vi, and both returned the same numbers, so I feel confident it's working. 

 

DtAinBridgeContinuousFourChannels2.vi

[amilford]

I'll try it on Monday. I've been using 250lbs load cells for testing. I am currently setting up a test in the test lab for the fixture I will use the DT9829 to upgrade so I can't tare down the tester because it is being used.

How exactly did you wire the voltage source the DT? I feel like I might be missing something in your description.

[DAQman]

I have a voltage calibrator that has an isolated differential output. I connected the (+) & (-) to the DT9829 pins 1 & 3. I didn't have a 350 resistor so I used a 1000 ohm and connected it from pin 3 to pin 34 to simulate a load cell circuit more closely.

[amilford]

Ok, so you made your own wheatsone. Making sure I understood so I could replicate it. Thank you