4
\$\begingroup\$

I am working with building a digital scale (of sorts) and have a single load cell working connected to my microcontroller.

If I wish to use multiple load cells, can I simply connect their signal outputs together in series, then amplify that signal? I do I need to collect the separate signals from each cell, amplify each one, then process each one, and sum them?

Thanks

\$\endgroup\$

3 Answers 3

Reset to default
6
\$\begingroup\$

According to Vishay Revere, they should be connected in parallel like this: -

enter image description here

\$\endgroup\$
5
  • \$\begingroup\$ Thanks Andy. Given that my microcontroller is detecting voltage - surely connecting them in parallel would not result in a summing of voltages. Because the force is distributed across the load cells - I would then only be detecting a fraction of the load? (say, 50% in a 2-cell scenario where force is evenly distributed) \$\endgroup\$
    – Adam
    May 26, 2014 at 20:33
  • \$\begingroup\$ @Adam I would imagine that sensitivity is reduced by n where n is the number of parallel loadcells but you ought to google this. I've used plenty of L-cells and strain gauges but never in parallel. \$\endgroup\$
    – Andy aka
    May 26, 2014 at 20:52
  • \$\begingroup\$ Thanks again Andy - really appreciate the pointers. I am currently using a single beam cell, but I need the mechanical stability of a 4-point system - however I don't want to have to use 4 amplifiers and consume more data pins! Thanks for the help. \$\endgroup\$
    – Adam
    May 26, 2014 at 20:57
  • \$\begingroup\$ Your uP doesn't have to do any summing. Just putting them in parallel is correct. If you imagine your two cells on each end of a long board, when you parallel them (assuming they are the same rating), you can stand on either end or the center of the board and read the same output. Industrially we have up to 8 in parallel on some of our scales here, feeding a common amplifier. \$\endgroup\$
    – R Drast
    Jun 4, 2015 at 11:04
  • \$\begingroup\$ @ R-Drast but the values you get from the distributed loadcell system will be a fraction of the actual weight right? \$\endgroup\$
    – McLosys Creative
    Nov 25, 2019 at 7:26
2
\$\begingroup\$

Connecting them in parallel is what is normally done. This averages (not adds) the signals from the individual cells, as indicated by R Drast above. Stand on either end or in the middle of the board and you will get the same answer in each case.

You cannot connect them in series (to add the outputs for higher sensitivity) without having isolated power supplies for each individual cell - in other words, too much trouble.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ The signals are averaged, but 'summing' is achieved by multiplying the mV/V to load force scaling factor by the number of load cells (assuming load cells are the same). ie, if it was 2mv/V with a full scale of 2000lbf, then 4 loads cells would output 2mv/V with a full scale of 8000lbf \$\endgroup\$
    – VoteCoffee
    Sep 20, 2018 at 19:29
0
\$\begingroup\$

Just a note, you are summing the resistance changes by connecting in parallel That is what you want to do; Connect in parallel, now both Resistive sensors are in parallel and their effects are added.

i.e Draw the complete circuit to make it clear.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ Connecting is parallel averages. It does not sum. Summation has to be acheived by factoring the scaling from mV/V to load force by the number of load cells (ie multipling the average by the load cell count). \$\endgroup\$
    – VoteCoffee
    Sep 20, 2018 at 19:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.