Strain gauge / Wheatstone bridge

oscarBravo

So I have this vague project in mind that involves weighing heavy things and reading the weight value into a microcontroller.

I have some strain gauges I got from China, sans documentation. They are apparently designed to handle up to 50kg each, and have three wires: red, black and blue.

After some limited experimentation with a cheap DMM (I'll check some more tomorrow with my Fluke and see if I can confirm), it seems that each one is configured as a potentiometer: black-to-blue resistance is 2kΩ, and red-to-black and -blue are both 1kΩ. Pushing as hard as I can on the cantilever, I can get the resistance to increase by 2-3Ω on red-black, and decrease by 2-3Ω on red-blue.

This suggests to me that it's designed to form half of a Wheatstone bridge. I could complete the bridge with two 1kΩ resistors, but I also thought about using two of the gauges together to both double the sensitivity of the bridge, and double the weighing capacity to 100kg. As best I can figure it, this would involve connecting the blue on each gauge to the black on the other, connecting an excitation voltage across the blue/black junctions, and measuring the voltage (via an in-amp) across the two reds. Does that make sense?

I'm also wondering how I'd go about wiring four of them up, for 200kg capacity and quadruple the sensitivity. Any thoughts?

DublinDilbert

It looks like you have load cells, which are stain gauges mounted onto a structure.

If you want to use both of them you might need to have one in compression and the other in tension.

You might need to worry about temperature effects, dummy stain gauges can be used to compensate for this.

oscarBravo

DublinDilbert wrote: »

It looks like you have load cells, which are stain gauges mounted onto a structure.

Yup, sounds right.

If you want to use both of them you might need to have one in compression and the other in tension.

I suspect they're already set up that way internally, hence the potentiometer effect. I'm not quite sure how, from looking at them, but what else could explain one increasing and one decreasing resistance from the same applied force?

You might need to worry about temperature effects, dummy stain gauges can be used to compensate for this.

Again, if it consists internally of two strain gauges, wouldn't this effectively self-cancel any thermal effects?

This is the sort of circuit I had in mind:

oscarBravo

Talking to myself, but...

oscarBravo wrote: »

After some limited experimentation with a cheap DMM (I'll check some more tomorrow with my Fluke and see if I can confirm), it seems that each one is configured as a potentiometer: black-to-blue resistance is 2kΩ, and red-to-black and -blue are both 1kΩ. Pushing as hard as I can on the cantilever, I can get the resistance to increase by 2-3Ω on red-black, and decrease by 2-3Ω on red-blue.

It seems they're definitely half-bridges. The total resistance across blue/black doesn't vary at all with force, but the red moves by a couple of ohms between the two.

This suggests to me that it's designed to form half of a Wheatstone bridge. I could complete the bridge with two 1kΩ resistors, but I also thought about using two of the gauges together to both double the sensitivity of the bridge, and double the weighing capacity to 100kg. As best I can figure it, this would involve connecting the blue on each gauge to the black on the other, connecting an excitation voltage across the blue/black junctions, and measuring the voltage (via an in-amp) across the two reds.

So I did this, and I can measure up to 5mV across the reds when I push down on either or both of the load cells. I guess I have my answer.

I'm also wondering how I'd go about wiring four of them up, for 200kg capacity and quadruple the sensitivity. Any thoughts?

As for this, I guess I just use two in-amps to two MCU inputs and add the combined readings?

oscarBravo

OK, one last update and I'll shut up about this until I've actually built something

This is what I've figured out for wiring four of them up through a single in-amp:



The load cells are wired in a ring, black-to-black and blue-to-blue. The excitation voltage is wired across two of the opposing red wires, and the other two red wires are the inputs to the in-amp.

Simulations indicate that increasing the force on any one or any combination of the load cells will give a linear increase in voltage on the amp input.

whizbang

If you are paralleling the load cells (mechanical connection) you have added the capacity to 200, but for the strain gauges, now you have a non-linear system. Much like parallel batteries, you can't control the balance between cells.

Series connecting the load cells, reduces the total capacity to less than 50, but now you can add the voltages for more sensitivity.

oscarBravo

Interesting - how would you go about wiring that? Also, given that it consists entirely of linear components (resistors), why would it be non-linear?