Quarter Bridge Strain Gauge with HX711 Amplifier

Hi there, I am trying to use a HX711 Amplifier with an arduino to measure strain according to this tutorial: Strain gauge theory Arduino scale tutorial HX711

I have the same wiring as seen at 6:25 of this video: How STRAIN GAUGE Works | Precision SCALE With Arduino - YouTube , except that I use only a quarter bridge instead of half bridge, i.e. just one strain gauge in the wheatstone bridge. The strain gauge and resitors are all 350 Ohms. The code is as in the tutorial:

#include <Q2HX711.h>

const byte hx711_data_pin = 3;
const byte hx711_clock_pin = 2;
Q2HX711 hx711(hx711_data_pin, hx711_clock_pin);
void setup() {
  Serial.begin(9600);
}
void loop() {
  Serial.println(hx711.read()/100.0);
  delay(500);
}

The problem is that I am currently only getting constant values of 8300 or so at the moment and they are not changing when I bend the strain gauge. The resistance would change with bending when I measured with a multimeter so the gauge definitely works.

Any help is greatly appreciated. Cheers.

Using the HX711 apply your bridge excitation (Red + and Black -). Measure your excitation voltage to your bridge. Next with excitation applied measure the bridge out while applying strain. The bridge output voltage should change. Does it?

Ron

There's no such thing as a quarter bridge! That's simply a resistor!

You need to complete the bridge with three (0.1% if possible) low-tempco metal-film resistors of the same value
if possible (otherwise you'll have to pick values that keep the two arms of the bridge at the same ratio).

Standard quality resistors won't do for this, strain guages are precision sensors where microvolts make
a difference. Its vital the resistors are low temperature coefficient or the output will drift around due to
thermal differences (bridge resistors self-heat, choosing 1W rather than 1/4W could help).

[ Actually it might be worse than that - some strain guages rely on each arm of the bridge being
closely matched in temperature(*), so the element doesn't have to be such low-tempco ]

(*) because they are interdigitated on the same substrate.

Ron_Blain:
Using the HX711 apply your bridge excitation (Red + and Black -). Measure your excitation voltage to your bridge. Next with excitation applied measure the bridge out while applying strain. The bridge output voltage should change. Does it?

Ron

Hi Ron,
thanks for the reply, the excitation voltage when measured seems to be zero across the red and black on DC and only a couple of mV in the AC. This is indeed strange. Same when measuring the output voltage. It does not change when strain is applied. Any ideas as to what's wrong? I had the arduino on when measuring.
Cheers.

MarkT:
There's no such thing as a quarter bridge! That's simply a resistor!

You need to complete the bridge with three (0.1% if possible) low-tempco metal-film resistors of the same value
if possible (otherwise you'll have to pick values that keep the two arms of the bridge at the same ratio).

Standard quality resistors won't do for this, strain guages are precision sensors where microvolts make
a difference. Its vital the resistors are low temperature coefficient or the output will drift around due to
thermal differences (bridge resistors self-heat, choosing 1W rather than 1/4W could help).

[ Actually it might be worse than that - some strain guages rely on each arm of the bridge being
closely matched in temperature(*), so the element doesn't have to be such low-tempco ]

(*) because they are interdigitated on the same substrate.

Hi Mark, thanks for the reply. What do you mean by 0.1%? 0.1% of what? Also do you have ones you recommend that you could post a link to? What will happen if I just use normal resistors?
Cheers.

Multiplier7:
Hi Ron,
thanks for the reply, the excitation voltage when measured seems to be zero across the red and black on DC and only a couple of mV in the AC. This is indeed strange. Same when measuring the output voltage. It does not change when strain is applied. Any ideas as to what's wrong? I had the arduino on when measuring.
Cheers.

OK, back in your previous thread I mentioned using 350 Ohm precision resistors for your bridge circuit. That would be for example 0.1% precision metal film resistors. The idea is the bridge needs to be balanced and balance is achieved using precision resistors. On the bright side since bridges commonly use 120 and 350 ohm legs finding the resistors should be easy.
Your bridge has an excitation input typically red (+) and black (-). With a balanced bridge the output should be zero volts. Increasing and or decreasing the strain gauge force applied will imbalance the bridge.
I am guessing you are using a HX711 module board?
Ron

Hi Ron,

ok thanks, I must have missed that in the previous thread. I guess I missed that part about precision resistors in the previous thread. I will order some. But I'm wondering, isn't the bridge already balanced since I am getting a zero volt reading for the output voltage? And shouldn't I still get something for the excitation voltage with the standard resistance? Could there be something else wrong as well?

Cheers.

The Wheatstone Bridge design shows up in many circuits. In this case with a 350 ohm strain gauge and 3 precision 350 ohm resistors you get a balanced bridge. If you draw the bridge and use 4 each 350 ohm resistors and do the math the bridge output voltage will be zero volts. Note I am not taking things like temperature coefficients into consideration.

One leg of the bridge is a 350 ohm strain gauge with no strain applied. Now if we apply a strain the gauge resistance will slightly change. That change causes imbalance in the bridge resulting in a voltage at the output. Draw a bridge, do the math and note the output. Now change your strain gauge to 349 ohms and 351 ohms. See what happens? The change will be proportional to the strain applied.

I am guessing you are using a small HX711 module and not just a plain HX711 chip? Since I don't know how you have things configured I have no idea what your excitation voltage is? Some HX711 modules provide 4.3 volt excitation (I think). I assume you are powering your Arduino for 5 volt operation and the HX711 off the Arduino 5 volts?

One more time I will tell you when building a bridge for use with a strain gauge things like temperature compensation and a very stable excitation voltage are very important.

Ron

Multiplier7:
Hi Mark, thanks for the reply. What do you mean by 0.1%? 0.1% of what? Also do you have ones you recommend that you could post a link to? What will happen if I just use normal resistors?
Cheers.

Try actually searching for "0.1% resistor", answers all those questions. I did explain why normal resistors won't do.

Ron_Blain:
The Wheatstone Bridge design shows up in many circuits. In this case with a 350 ohm strain gauge and 3 precision 350 ohm resistors you get a balanced bridge. If you draw the bridge and use 4 each 350 ohm resistors and do the math the bridge output voltage will be zero volts. Note I am not taking things like temperature coefficients into consideration.

One leg of the bridge is a 350 ohm strain gauge with no strain applied. Now if we apply a strain the gauge resistance will slightly change. That change causes imbalance in the bridge resulting in a voltage at the output. Draw a bridge, do the math and note the output. Now change your strain gauge to 349 ohms and 351 ohms. See what happens? The change will be proportional to the strain applied.

I am guessing you are using a small HX711 module and not just a plain HX711 chip? Since I don't know how you have things configured I have no idea what your excitation voltage is? Some HX711 modules provide 4.3 volt excitation (I think). I assume you are powering your Arduino for 5 volt operation and the HX711 off the Arduino 5 volts?

One more time I will tell you when building a bridge for use with a strain gauge things like temperature compensation and a very stable excitation voltage are very important.

Ron

Hi Ron,
thanks again for the reply. I have now ordered some of those high precision sensors and I understand that they are necessary.
Yes, it is as you described, I am using the red HX711 chip as can be seen below:
hx711
I have wired it to the 5V of the arduino. The only thing I find weird is that there is a 3.7 V excitation voltage across the black and red but still absolutely nothing happens to the raw readings, and they remain completely constant. Shouldn't they be at least somehow changing, even if the bridge is not completely balanced (as it is not now due to me only using regular resistors), since I am actually measuring an output voltage of a few 10s of millivolts now across the bridge?

Cheers.

Yes, if you apply excitation and flex the strain gauge you should see a small change in the output. A very small change but noticiable on a good DMM or bench meter.

The module gives a data out but you should see an analog voltage change at your bridge out.

Ron