Question regarding reading two strain gauge input modules using one analog input

[chrwit]

Hello everyone,

I am constructing a weighing system using eight Transducer Techniques MLC load cells connecting to two Advantech ADAM-3016 strain gauge input modules (four load cells per module). I have been having problems connecting the module outputs in a way that adds the voltages together. I tried connecting the module outputs in parallel, which averaged the signal and unfortunately does not work for my application. I also tried connecting them in series by connecting the +Output on Module 1 to the -Output on Module 2, and then reading the +Output on Module 2 and the -Output on Module 1. This made the signal extremely noisy and unusable, even after adding a 1 uF cap. The signals are being read by a NI-PCI-6224 card in a LabVIEW Real Time PC.

Any insight on how to add the two output voltages without introducing noise would be greatly appreciated. Please let me know what other information you need.

http://downloadt.advantech.com/ProductFile/Downloadfile3/1-FUXAMA/ADAM-3016.pdf

<- Link to the Advantech 3016 module user's manual.

http://www.transducertechniques.com/mlc-load-cell.aspx

<- Link to the Transducer Techniques MLC load cells.

Thanks in advance.

 

[VEBill]

Quick answer: bridge circuits are often used in this sort of application.

 

[VEBill]

Yep, look at the Wiring/Color Code tab on the load cell webpage.

Due to the (I presume 'common to both') excitation voltage, you're not going to be able to wire the outputs in simple series.

 

[VEBill]

The excitation voltage is specified as 10 VDC. So the output will be a small voltage signal centred at 5 VDC absolute relative to the excitation voltage.

You can't just wire the outputs in series.

If there's any available trick to putting bridge circuits in series, it's not springing to mind.

One could employ two independent 10 VDC power supplies with floating outputs. Or batteries. Then the outputs could be wired in series to add.

 

[chrwit]

VE1BLL,

Thanks for your input. Do you mean set up the strain gauge input modules in a bridge circuit? My problem is with the signal once it leaves the ADAM-3016 signal amplifier module. It's my understanding that the input from the load cells is working correctly. I am using a 5V excitation voltage.

I have a PDF of the schematic in question at this link:

http://www.filedropper.com/cs598081-

The issues I believe are to the right side of the amplifier modules. This is the way I currently have them set up.

Thanks again for your help.

 

[IRstuff]

I don't understand why you want to add these measurements together, which can be easily done with a summing amp, but analog addition is a noisy and less than accurate affair.

TTFN
faq731-376

 

[LiteYear]

One could employ two independent 10 VDC power supplies with floating outputs. Or batteries. Then the outputs could be wired in series to add.

I think you're right, but even then I think the two sensitive circuits would be susceptible to corruption from each other. I suspect you'd have to go with an op-amp stage for each load cell signal that removes the bias, and then add the outputs together in series.

 

[VEBill]

(The filedropper link didn't work for me.)

My previous posts were aimed at the "using eight ... load cells connecting to two ... input modules (four load cells per module)." I think I misread your original post - I thought you were trying to add the voltages from the load cells themselves. Sorry if I was confused by the eight into two.

The Adam3016 claims complete 3-way isolation, so you should be able to add output voltages as you're attempting. At least conceptually.

If the signals are small, then avoiding noise often just goes back to basic principles (shielded wiring, twisted pair, avoid ground loops, etc.) Can you rerun your integration with the inputs set to simulated 100% load (max scale)?

It can be useful to see if the noise is actually a completely separate issue from the summing by series wiring concept.

 

[BrianE22]

How do you get 4 load cells/module for the amplifier? I see 4 gauges (1 load cell) per module.

 

[chrwit]

Thanks for all your input everyone.

Brian, I meant to say I have four load cells connecting to each amplifier. I think VE1BLL is right though, and I am working on some noise mitigation right now. We will see how it goes.

Again, I appreciate everyone's input on this.

 

[BrianE22]

I still don't get it. How do you attach 4 load cells to an amplifier designed to accept one load cell?

 

[VEBill]

Has anyone else been able to hack their way into the OP's 'filedropper' link to view the "PDF of the schematic in question"? I've tried on multiple devices and the link is not working.