Measuring Force in Compression: S-Beam Load Cell Voltage Output INA114

Other Hardware Sensors
1

Hi Folks,

I'm using an Anyload 101NH-2k b 2000lb capacity s-beam load cell to attempt to get accurate readings for 0-1000lb direct compressive loading application.

The load cell is designed for compression and tension. The output sensitivity is 2.99mV/V, and the excitation voltage I have been experimenting with is an arduino's 5V rail. Attached is a picture of the load cell's calibration document with other ratings (many of which I don't understand fully).

I am using a TI INA114 amplifier (http://www.ti.com/lit/ds/symlink/ina114.pdf) to amplify the signal voltage to a readable range for arduino. I've experimented with a number of gage resistance (Rg) values to modify the gain on the amplifier, and have found that there is a drastic trade-off between range and resolution with different gain values.

I was hoping to have +/-1lb accuracy for a range of 0 - 1000lb+ compression loads, but I can't figure out how to achieve it.

My testing indicates:
-For a gain of ~186 (Rg = 270ohm), I can achieve 0-870lb range with +/-2.3lb resolution. (No load = 2.5V, max output 4.28V)

-For a gain of ~377 (Rg = 133ohm), I can achieve 0-430lb range at +/-1.19lb resolution. (No load = 2.49V, max output of 4.28V).

I'm thinking of putting a DPDT switch into a permanent perfboard circuit so I can switch between these two "modes" (low res./high range mode 1 and high res./low range mode 2)--this will suit my purposes to a lesser extent, but ideally I would just have +/-1lb resolution from 0-1000lb!

I feel like this range and resolution is achievable, but I just don't know how to do it.

MY QUESTIONS:

  • What can I do to achieve this range and resolution?

  • Why does my load cell output half the excitation voltage at no load? I assume it's because the load cell is designed for tension and compression, not one or the other (the voltage varies in opposite directions in tension vs. compression). Am I wrong about this? Is there a way to re calibrate so I get the full 0-5V range with a 5V excitation voltage while maintaining good resolution?

-Why is there a limit of 4.28V output no matter what the gain is? I would like it to boost back close to 5V if possible, but it seems like the output voltage is limited to a certain percentage of the excitation voltage for some reason.

Thanks,

Leo

2

Seems to be a job for the HX711.
Intrumentation amp, regulated exitation voltage, digital selectable gain, and 24-bit A/D, all in one.
Breakout boards are just a few dollars on ebay.
Leo..

3

Wawa:
Seems to be a job for the HX711.
Intrumentation amp, regulated exitation voltage, digital selectable gain, and 24-bit A/D, all in one.
Breakout boards are just a few dollars on ebay.
Leo..

Hi Wawa,

So after looking at the HX711 data sheet, it seems like the board has options for 32, 64, or 128x gain--less than what I'm using right now to get the required range.

But it also seems like the 24-bit A/D built in could get me a much larger possible range of digital data compared to Arduino's 10-bit ADC.

Not sure if it will do exactly what I want, but it might, and at the price it's definitely worth a shot!

Thanks for your reply,

Leo

4

-Why is there a limit of 4.28V output no matter what the gain is?

That is a limitation of the INA114 amplifier. It depends on the supply voltage.

5

jremington:
That is a limitation of the INA114 amplifier. It depends on the supply voltage.

Thank you for confirming that! I suspected as much.