Trouble measuring strain accurately using HX711

Other Hardware Sensors
1

Hello all, I am trying to do some strain measurements using strain gauge.

I am using a quarter bridge circuit as shown below. R3 is a potentiometer set at 120 ohm.
image
I can't use half bridge or full bridge configuration for my application.

I am using BX120-3AA strain gauge with sparkfun HX711 amplifier.

  1. From the datasheet the full scale differential input is +/-0.5(AVDD/GAIN).
    Gain =128
    AVDD= 4.2 V when measured using multimeter.
    Full scale differential input = +/- 16 mv.
    The maximum output reading from HX711 amplifier is 2^23 for positive range. So, can I assume 1 reading from HX711 corresponds to +16 mV/2^23 = 2 nV (approx)? Can I use this as calibration factor to convert the Hx711 readings to Vo?

  2. The value obtained from the HX711 is drifting continuously.

    image
    image777×246 5.8 KB

    Though this value is drifting continuously, I am getting 16 stable bits for 2 minutes.

But after 2-3 minutes the values start to fluctuate largely. I am only getting 9 stable bits.

image
image776×145 4.05 KB

The readings shown here are average of 10 readings. I am using shielded cable (CAT 7) from strain gauge to HX711 in two wire configuration.
Connections
I'm not sure what could be causing this drift? What is the best ways to resolve this?

  1. Can this drift be due to self heating of strain gauge? The SG details are:
    Resistance=120 ohm | Grid dimension= 3 mm X 2 mm | Excitation voltage= 4.2 V | Power =36 mW.
    If I had to reduce the excitation voltage I can provide 2.5 V by connecting a ~75 ohm resistor in series to bridge circuit such that bridge Vo is 1.25 V (Min common mode input of HX711 is AGND + 1.2 V). Even at 2.5 V excitation voltage the output readings are too noisy.

  2. I want to find Noise Free Resolution of my system. I saw a formula here in post #5.
    NFR= (full scale load) / ((2^bits)0.128 S)
    The strain limit of strain gauge is 2%. Does it mean the the SG can measure 20 milli strain in both compression and and tensile. But the HX711 can measure only up to +/- 16 mV roughly +/- 7 millistrain . So how to measure sensitivity of the system? And what should I consider as the full scale load to find NFR?

  3. As the SG is sensitive to both strain and temperature. If I don't apply strain to SG then can I measure temperature change from the readings. If yes what would be relation for change in resistance / °C ?

2

Yes, of course. Is the gauge just flapping around in the air, or is it properly and firmly attached to a nice thick piece of metal?

3

Its bonded to a 4mm thick aluminum plate.

4

These may be of interest:

1 Like
5

Thank you for the references. Very informative.

As I bought this BX-120-3AA from local shops I do not have proper datasheet of the SG. If any one has used this SG before could anyone share its datasheet?

I found in a generic datasheet that thermal characteristics are represented in ppm/ deg C . What does it mean? How to convert that to change in resistance/ deg C?

6

Check out the HBM website. Tons of good information about strain gauges there.

Your datasheet's value of ?? ppm/degC is probably its self-compensation value (about 23 for strain gauges designed to be applied to aluminum) as explained in the following page from that site.

7

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.