Maximum resolution from 100g load cell + quiic scale

Other
1

Hello,

I’m using a 100g mini load cell with a quiic scale board, connected to a feather m0 adalogger, writing data to an SD card. The load cell is working, but the noise floor is ~40k and small changes in weight are tough to read accurately. Is there any way to improve the resolution of the sensor?

Thanks,
Jason

2

You can try controlling environmental stuff (temperature, grounding, shielding) and whatnot more tightly to help a bit, but for the most part they are a bit noisy

You can try using a kalman filter or other statistical means of reducing noise (moving average, calibration curve) as well…after doing a series of calibration measurements

3

Measuring very small forces with a gauged load cell is always going to be a problem. The aluminum, in this case, has to deflect for the electrical gauge to bend and there just isn’t much there. Mechanical isolation becomes an overwhelming problem and the inherent noise never goes away but it does change unpredictably. Nonlinear performance creeps in at some point, too. Ultimately, you’ll need a different technology altogether for high resolution, low value force measurements. Commercial scales and balances usually use magnetic force restoration techniques beyond 0.01 grams and/or about 100k graduations. I’ve heard of intriguing parallel plate capacitive and acoustic measuring methods, too. But it won’t be easy with an inexpensive little chunk of metal like a load cell.

I’ve been tempted on and off to get one of those super cheap portable pocket scales, palm sized, battery powers, gas station/dispensary grade scales (example below) to goof around under the hood and really put it to the test. It’s possible that they’re using some neat, thought provoking tricks and maybe even a better load cell.

4

I am currently using a strain gauge based sensor for this project and it’s actually working quite well! But the problem is the other one that I’m using is about 100X more expensive and I wanted to go cheaper. This is a link to that one.

The main difference I can see in the specs sheet is it has a 2 mV/V rated output, while the Sparkfun Mini load cell has a 0.6 mV/V. This is a 3X difference in sensitivity, but I have beautifully clear data using that load cell, and don’t see anything with this one. I’m a bit surprised at the difference.