Load Cell

[jcw0220]

I have a load cell (lcl-113g from Omega) that I want to hook up to my LabJack U3-HV. Being compleatly new to this I don't know how to handle the five wires. There is a white black green red and an uninsilated ground wire. Aside from the ground wire what do I do with the others?
thanks
jcw

[LabJack Support]

That sensor simply provides a raw bridge circuit, without any signal conditioning, so the quick answer is that you need an LJTick-InAmp. See Appendix D of the LJTIA datasheet:

http://www.labjack.com/downloads.php


Here is the wikipedia page about a wheatstone bridge:

http://en.wikipedia.org/wiki/Wheatstone_bridge

[jcw0220]

Correct me if I'm wrong but the gauge is providing a Wheatstone bridge. So there is a Source Voltage and an Output Voltage. My problem is figuring out which wires correspond to which Voltage and after figuring that out which terminals on the LabJack those correspond to. And I have the amplifier on the way.

[LabJack Support]

See if the following helps:

http://www.labjack.com/forums/index.php?showtopic=3665

[jcw0220]

I dont know how I missed that. Thanks

[calvin.anderson]

I am also very new to the the world of Labjack. I have replied to this thread since my first issue pertains to a load cell. The manufacturer is Anyload, and the model is 101NH and its capacity is 3000 lbs. I am attempting to use a U3-LV LabJack. Currently, I have attached the four wires to the LabJack in the following order:

Red (+Exitation) ==> VS
Black (-Exitation) ==> GND
Green (+Signal) ==> FIO0
White (-Signal) ==> FIO1

I have created a channel in DAQFactory Express called LOADCELL. I have set the channel # to 0, and since I am attempting to use a differential analog input I have marked channel 1 in the quicknotes section.

I am attempting to look at the graph section and the numbers that I am receiving are negative which are definitely lower than I would have expected. This would tend to tell me that I have something hooked up backwards or incorrectly.

I am aware of the conversion ability of DAQFactory Express and would like to use this in order to convert the signal the Load Cell is producing into a legible force in lbs. I have read the available tech sheet and I believe the graduation is 3.0 mV/V which is not what I would have expected since I thought most graduations for units of this type were 2.0mV/V.

I have used a DMM in order to measure the expected excitation voltage and I am receiving nearly exactly 5.0V. This seems to be perfectly fine.

After all of this bable, I need help in two places. First identifying if there is a problem with the load cell. Secondly, understanding if the load cell is hooked up to the LabJack correctly and if it is how I can get better numbers i.e. decide what the offset should be and what the gain should be in order to convert signal mV to force in lbs.

[LabJack Support]

First identifying if there is a problem with the load cell.

I would use the test panel in LJControlPanel for initial troubleshooting, rather than DAQFactory. It is a simpler program and eliminates more unknowns.

Use your DMM to measure the voltage from FIO0 to FIO1 at the same time as you monitor it in LJCP. If you change the weight on the load cell by 3000 pounds, you should see the voltage output change by 15 mV.

Secondly, understanding if the load cell is hooked up to the LabJack correctly and if it is how I can get better numbers i.e. decide what the offset should be and what the gain should be in order to convert signal mV to force in lbs.

You need an instrumentation amplifier such as the LJTick-InAmp. See above. Your full expected signal swing is only 0-15 mV, and the resolution of a differential input on the U3 is about 1.2 mV, so you will only have about 12 steps across your entire range.

[calvin.anderson]

Thank you for your quick reply. I have a DMM and I am measuring the voltage across the FIO0 and FIO1 terminals. The voltage that I measure across them is 0 VDC. I have taken the time to make certain that I have good connections between the terminals and the wires of the load cell. Between GND and FIO0 I have a value of 2.503 VDC and between GND and FIO1 I also have a value of 2.503 VDC. This makes me believe that I do not have the LabJack hooked up correctly, the Load Cell is broken, or I do not have the LabJack set up properly.

When I opened LJCP the FIO Channels are set up as follows. Their associated LJ readings are also included.

=========================================
FIO | AIN | DI | DO | Voltage/ | Neg.
Chan | | | | State | Channel
=========================================
AIN0 X 2.453 SE
AIN1 X -0.006 AIN1
AIN2 X 0.378 SE
AIN3 X 0.294 SE
FIO4 X
FIO5 X
FIO6 X
FIO7 X
=========================================

It is very possible that I do not have something set up correctly, since I am so new to this.

[calvin.anderson]

I am sorry to post again so quickly. I did have another question though about the amplifier. Could I, using and OpAmp circuit amplify the return signal in order to get more than 12 steps. I think that you suggested a LJ sponsored amplifier, but I am not sure where I should look for that.

[LabJack Support]

I have a DMM and I am measuring the voltage across the FIO0 and FIO1 terminals. The voltage that I measure across them is 0 VDC. I have taken the time to make certain that I have good connections between the terminals and the wires of the load cell. Between GND and FIO0 I have a value of 2.503 VDC and between GND and FIO1 I also have a value of 2.503 VDC. This makes me believe that I do not have the LabJack hooked up correctly, the Load Cell is broken, or I do not have the LabJack set up properly.

What load do you have during this measurement? If you have 0 load, then all these measurements sound correct. If you have 3000 pounds, you should see that one line changes to about 2.5105 compared to GND and the other changes to about 2.4955 compared to GND, and thus the difference between them is about 0.015 volts.

I did have another question though about the amplifier. Could I, using and OpAmp circuit amplify the return signal in order to get more than 12 steps. I think that you suggested a LJ sponsored amplifier, but I am not sure where I should look for that.

You need an instrumentation amp. You can buy an in-amp chip (or 2-3 op-amps) and build your own, but I recommend the LJTick-InAmp from LabJack:

http://labjack.com/catalog/ljtick-inamp

The other option is the U6, which has a very nice instrumentation amplifier built-in.

[e-choi]

I have an exact same units: U3-HV and LC304-7.5K Omega load cell.
connected as suggested using excitation voltage supplied by VREF.

conversion I'm applying is ((Value-0.375)/201)*(7500/(0.002*AIN0)) where AIN0 is reading VREF.

with no load I'm reading -0.1214 lbs. every 6th reading or so, I get -10.125 or so; is this implying that my accuracy is within 10 lbs? is this an electrical noise(blimp) that can be removed? It's not critical and within error rate of the sensor unit but wondering if electrical noise or the load sensor unit.

I did not connect to FIO4 ports as suggested.

[LabJack Support]

The typical noise level of the U3 is +/-1 count, which is very good. On low voltage channels that equates to +/-600uV, while on AIN channels that equates to +/-5mV. See if those numbers match up with what you are seeing.

[e-choi]

For a week or so my conversion had 0.37 as the offset which was the average measurement - after conversion it read 0lbs. Over the weekend, it had drifted to 0.40 as the offset - after conversion it read 290lbs. (no extreme temperatures)

What would explain the drift of this offset gain?

[LabJack Support]

There are many possible things. I would try to isolate which piece has changed. Are you using a low-voltage or high-voltage analog input?

[e-choi]

Nothing what so ever changed. No one touched it; only thing if changed at all would be that it is cooler by less than 10 degrees Fahrenheit.

I am using the U3-HV with LC304-7.5K; where I have the LJTick-InAmp connected to AIN2/AIN3, the sensor connected to INA and using VRef from InAmp. I wiggled the wires to see if any affect there may have, but same. additional info: I am also applying same excitation source to a potentiometer.

I've corrected my gain offset to 0.4 and know what to look for later but would like some ideas to explore and understand.
Again, nothing changed at all - powered off and resumed after a weekend.

[LabJack Support]

Nothing changed on purpose, but something did change. If nothing else, your measured offset changed by 30 mV.

Any particular reason you have the LJTIA plugged into AIN2/AIN3? Low-voltage channels such as FIO4/FIO5 would be better.

Also, using a low-voltage channel such as FIO6 would be better to monitor VREF also. You will have to set FIO6 to the special "0-3.6V" range by specifying a negative channel of 32.

What are you using for software? If you did not before, I suggest you now log raw values in addition to your scaled readings. It would be useful to see the raw voltages from AIN2 and AIN0, to see if both changed or just AIN2.

Use channel B of the LJTIA to monitor offset of the LJTIA/U3. Jumper INB+ and INB- both to GND. Set channel B to have the same gain as channel A. Log the voltage from that channel with your other data to see if the offset changes.