How do I connect a futek LRF325 load cell to an scc-68 terminal block

[slashc]

Hello,

I am trying to connect a futek LRF325 load cell to an scc-68 terminal block without any SCC modules. The terminal block is then connected to a PCI-6052E Daq card and I am trying to read the voltage output of the load cell using an analog input driver in matlab's XPC target environment. I am exciting the load cell using a 10V power supply, which is plugged into the building's power and the DAQ card is also being powered by the building's power through a computer (5V). I thus assumed that my signal was a ground-referenced signal source. I have tried connecting the load cell under the ground reference settings differential (pins 28+ and 61-) and as reference single-ended (pins 28+ and 64-) and have had no success in reading in the output voltages from the load cell using XPC target. My driver configuration settings for the differential ground are as follows:

channel vector: 5 (corresponds to AI4 on DAQ card)

range vector: -0.05V (corresponds to +- 0.05V output range of load cell)

input coupling vector: 2 (corresponds to setting for differential ground reference)

sample time: 1

PCI slot: [0,10] (corresponds to 6052 board location)

I can read the voltage output off of the load cell using a multimeter and it is in the expected range. I have tried running a test program through the XPC target environment with the analog output driver (similar settings as above) for the 6052E board and it has worked just fine. I also have tried changing around the channel vector for all 8 AI differential channels. In the manual for the scc-68 terminal block they only give instructions for connecting AI signals when the SCC modules are installed and they refer to using the E series help manual for information about connecting AI signals in other situations. I checked the E series help manual and connected the load cell using the differential settings as above (pin 28+ and 61-). Have I connected the hardware correctly?

Thanks,

Michael Dawson

Research Assistant

Advanced Robotics and Controls Lab

University of Alberta

[S_Hong]

Message Edited by S_Hong on 07-02-2008 06:02 PM

[S_Hong]

Hi slashc, 

All our PCI cards are compatible with Windows and LabVIEW RT  and we do supply drivers for our DAQ cards for the listed Operating Systems. However, we as far as supporting third party OS, you may either need to write your own driver for the card or find someone who has written a driver for this, but we do not support it. However, it is sometimes possible to get our hardware working on these OS.

As for the connections, it looks like you have your load cell connected properly. Are you using an external excitation source for your load cells? You may also need to configure your task as far as gage resistance and excitation voltage.

[slashc]

Hi S_Hong,

Thanks for the reply.

I reviewed the spec sheet for the load cell and it lists the shield as "floating" so I think the signal may actually be floating? I measured the resistance of the load cell to be about 350 ohms. If I understand the e series manual correctly then I need to connect a 35kohm resistor between the negative source signal and the AI GND (i.e. connect the resistor between pins 61 and 64).

Load Cell Spec Sheet:

http://www.futek.com/files/pdf/Product%20Drawings/lrf325.pdf

Do I also need to connect a 35kohm resistor from the AI+ (pin 28) to the AI GND (pin 64)?

The load cell is currently being powered by an external power supply, but this is not a good option because the voltage shifts slightly whenever I turn it on or off. To measure the excitation voltage directly into my measurement program I was thinking of exciting the load cell with a 10V analog output signal directly out of the daq card. Is this a good idea or are the AO gnd lines not meant to handle any significant current? The other option I am considering is using the +5V power source (pin5) to excite the load cell at a slightly lower voltage than it was calibrated at.

Thanks,

Michael

[S_Hong]

Hi slashc,

Unfortunately the 6052E cannot provide external excitation and it is probably not a good idea to use the AO since it might not be able to provide enough current to your load cell. I believe that with your current hardware, you will have to use an external excitation source.

As for the connections to your device, the load cell in itself should already be using four strain gages in a Wheatstone bridge configuration and therefore, you must specify in your task that you are using a full-bridge configuration. You must also specify that you are using an external excitation and specify its voltage. I have attached an image of the connections you should be using in your setup if your are to use AI4.

Pressure and Load Measurements: How-To Guide

[slashc]

Hi S_Hong,

I have narrowed down the problem to either the 3rd party xpc target driver or the analog input functions of the board itself (since the analog output functions had already been tested and found to be working ok). I installed the NI-6052e board on a windows computer and ran a analog input test panel in the NI measurement and automation program. According to page 10 of the DAQ Getting Started Guide the graph should have shown a sinusoidal wave between 1 and -1 V. Instead the following results were obtained: http://www.ualberta.ca/~mrd1/6052e.JPG. Does this mean the board is damaged? If it is damaged, is it possible to send in for repair or do you think it will need to be replaced?

- Michael

[S_Hong]

Hi slashc,

It appears as if nothing is connected to that particular channel in your Test Panel. If you are feeding a voltage into AI2, make sure that it is well connected and that you select the appropriate Input Configuration. One other test I would suggest to perform is a loop-back test, that is, connect an analog output channel to an analog input channel on your board to see if you can measure anything. To do this, you can wire pin 22 (AO0) to pin 68 (AI0) and set a DC voltage in the Analog Output tab in Test Panel (Make sure to update the value). Once your output is set, go to your Analog Input tab and set your Input Configuration to RSE. You can then set your channel and measure the voltage. Hopefully you will be able to see something. Please keep us posted.

[slashc]

Hi S_Hong,

For some reason I thought the NI 6052e board would generate it's own test signal in the test panel. Anyways, I attached a signal generator to AI4 and was able to read in the sinusoidal signal as seen here http://www.ualberta.ca/~mrd1/6052e03.JPG. I then tried reading in the voltage from the load cell excited externally by a 10v PSU in labview signal express and the results can be seen here: http://www.ualberta.ca/~mrd1/6052e02.JPG. I measured the output from the load cell beforehand on a multimeter and was expecting an output between -0.06 to 0.5 mV. At this time the signal was still responsive to my adding and removing masses from the load cell. In signalview I tried reading in the load cell output on several different DAQ channels and I also tried playing around with the labview step settings, but I would always obtain the same noisy signal. After sometime the signal started floating around and finally settled looking like this (1.2V problem): http://www.ualberta.ca/~mrd1/6052e04.JPG I also tried connecting the load cell to an oscilloscope and found the signal to still be noisy, but did not run into the 1.2V problem. I'm going to try a different power supply and then a different daq card to try and troubleshoot the problem. I'm also going to remake the cable connecting the load cell to the terminal block. Do you have any ideas about what might be causing the 1.2V problem?

Thanks,

Michael

[S_Hong]

Hi slashc,

I'm glad that you got your output to work. It seems that the reason why you may be getting unexpected results is because there may be an error in your setup. One thing to note for custom voltages with excitation is that the voltage read in MAX is going to be your scaled voltage and not your actual voltage. Your scaled voltage should be the actual voltage coming in your DAQ device divided by your excitation voltage. So for example, if I am applying a 5V input and am using a 10V excitation source, MAX should show 5V/10V = 2.5V. Therefore, it is normal to read different values on your multimeter and MAX. Another thing to note is that the max & min range you specify in your task represents the range of your scaled data. Since this range will determine the gain of your programmable amplifier to maximize your ADC, it is very important to set this correctly. I'm a bit surprised that MAX hasn't generated any errors with your current configuration as that would mean that your input voltage should be 50V.

Also, you should applying a larger load to your load cell since at full-scale, your absolute accuracy is rated at 4.747mV which might not be good enough to measure 0.5mV. The absolute accuracy will depend on your full-scale and is documented in the specifications sheet.

As for the 1.2V you are observing, I believe that this is simply due to the fact that you are railing in your measurements since this is showing you the scaled voltage.

[slashc]

Hi S_Hong,

I do not think I was actually scaling before as I left the box unchecked. The full scale output of the load cell is actually about 20mV (2mV/V x 10V excitation voltage), but I was testing with a 500g mass that gives an output of 0.25mV, so that was my expected result. I tried adding scaling and I am still receiving a similar noisy signal (http://www.ualberta.ca/~mrd1/6052e5.JPG). After a while the signal jumps up to about 1.2V and unplugging and plugging the power back in brings the signal down to about 0V similar to the image above.

I've tried using a different PSU and also redoing the connectors of the load cell with proper cabling. I'm going to try reading in the load cell on a different DAQ card and see if I get the same results.

- Michael