Hello everyone, I am working on a project that uses an Arduino and a s-beam load cell. This is my first time ever using a load cell and I am having trouble understanding how it works.

Sources:

I followed a very good tutorial found here: http://edg.uchicago.edu/tutorials/load_cell/

I am using the SEN7110 S beam load cell, datasheet link: http://www.scalemarket.com/files/master/pdf/SENTRONIK/7110.pdf

And here is the link to the INA125P the instrumentation amplifier I am using: http://www.ti.com/lit/ds/symlink/ina125.pdf

Experiments:

I connected the circuit just like the tutorial advised, the only thing I changed was the connection of VrefOut to 5V as seen in red in the picture below
Yellow = Green in my load cell

* The problem is that the values being read from the Arduino do not change even though I apply force to the load cell.

After a lot of trials and reading from forums, I tried to exchange the Green and White cables, Green to Vin- and White to Vin+, and measure the output.
The outputs starting changing as I hoped but another problem came up.
Here is a graph of the values I acquired in one of the tests

The Y axis is voltage and the X axis is the number of the sample, ie sample value #1 was 1.1V

** As you can see the values jump in unpredictable/undesirable ways

Below is a test done right after the initial test:

As you can see it jumps in even more undesirable ways

*** And the "zero" force or initial voltage did not return to 1.1V, instead the voltage has an offset of about .5V

I searched around on how to check if my load cell was working correctly, I found this video

And tried to measure the resistances, but I don't know which resistance should I be comparing to the one in the datasheet

Here are the resistances that I measured

Cables Resistance

Green to White 357 Ohm
Red to Black 354 Ohm
Red to Green 281 Ohm
White to Black 289 Ohm


To recap the problems I have and don't understand:

* Values do not change even though I apply force to the load cell.

** Values jump in unpredictable/undesirable ways.

*** Load cell does not return to the initial, "zero force", voltage.

Any help on understanding the load cell is appreciated!

Let me know if there is other information that I could provide to clarify what I am trying to do, I fear that my load cell might be broken, but if it is not the how could I fix my circuit so that I get stable readings from the load cell.

 

hi,
As you are using a single supply, connect the IN129P as shown in this image from the datasheet.
Disconnect the Arduino and measure the voltage across the load resistor on pins #5 and 11, apply force to the Load cell, tell us what you measure, with force and no force to the L/C
ie: use a 2.5Vref as shown
E

 

hi,
As you are using a single supply, connect the IN129P as shown in this image from the datasheet.
Disconnect the Arduino and measure the voltage across the load resistor on pins #5 and 11, apply force to the Load cell, tell us what you measure, with force and no force to the L/C
ie: use a 2.5Vref as shown
E

Thank you!

I will try this on Tuesday and let you know how it goes since the equipment is in my school.

Something that I just now realized, on the image from the datasheet, the load cell is not directly to Vcc or 5V, could that be a mistake I am making?

 

I will try this on Tuesday and let you know how it goes since the equipment is in my school.

hi,
As you are not familiar with L/C use, its a good idea to build a basic analog set up using the IN129A and check the voltage output range, before connecting a MCU, ADC.
Note also the sensor datasheet states 10Vdc/ac, so with a 5Vdc excitation, the mVolt output of the L/C will be half of that at 10Vdc ext.
ie: 10Vext > 3mV/V , 5Vext > 1.5mV/V , 2.5Vext > 0.75mV

E

 

hi kenlide,
This LTSpice plot image will give you an idea what the Vout range should be from zero to full load when using a 2.5Vexciation voltage.
Checking your resistance values in post #1, suggest that the load cell could be faulty.?
Is it a new L/C or is recovered from a previous project.?

Ignore the LTSpice 'asc' file, it is posted for any other LTS user member, who may find the INA129P circuit useful.
The components enclosed by the dashed rectangle are the load cell and the INA129P circuit.

E

 

Hey, sorry I haven't been able to measure the values you mentioned since I haven't had the chance to go to my school and get the equipment.

Yes, the L/C should be new since I bought it only for this project.

That's how I would like my outputs to look! If anything with a better 'gain' or 'range'

 

That's how I would like my outputs to look! If anything with a better 'gain' or 'range'

hi,
I would suggest using a Vext of 5Vdc, this will increase the L/C output to 1.5mV/Vext.
Also the Rgain resistor on the INA129A could be reduced in value to increase the gain of the INA.

What is the maximum voltage of Vout you require for the Arduino ADC input.??

E

 

hi!

I apologize, I couldn't get back to you yesterday because I ran into some trouble and was not able to get the circuit together.

I believe the absolute maximum rating for the Arduino ADC input is 5V.

I tested the circuit you advised, this is what I mesasured:

The Rg that I used was 10kOhms

1) To make sure the L/C was connected I measured

Red to GND 2.49V
Green to GND 1.22V
White to GND 1.28V

2) When measuring pins #5 and #11

With no force, I measured 0.61V

As soon as I applied force 2.20V
* And this value did not change as I applied more force, it seemed to have saturated?

Another thing to note is that when there was force the Green to GND measured 1.46V

What are your thoughts? Let me know if I can be more clear in the explanation!

 

Ok, you guys seem to have the hardware end covered. However, we cannot tell what the Arduino is reading without the code. Kenlide, can you post the Arduino sketch here (with CODE tags) so that we can look at it?

 

Sure!

I am using the ReadAnalogVoltage example with an added for loop and a one second delay.

The setup code stays the same.

Here is the loop code:

void loop() {

  for(int sample = 0; sample < 200; sample++){
    int sensorValue = analogRead(A0);
  
    float voltage = sensorValue * (5.0 / 1023.0);
 
    Serial.println(voltage);
  
    delay(1000);
    }
}

 

The Rg that I used was 10kOhms

hi k,
That will give a gain of only 5.!

Green to GND 1.22V
White to GND 1.28V
From these values the LC is outputting a 1.28v-1.22v = 60mV.

Look at this sim plot, using a 60mV signal , 10k [ gain 5]

As a test, with the L/C connected, short between Gn and Wh and measure the Vout for the INA, post what you see.

E

 

Sorry, I haven't been able to get back to you.

I will try the same circuit with a Rg of 1kOhm to see the output.

When I short Green and White, do I short them connected to the IC? What should I expect my output to be?

I have taken the equipment from my school to my house to be able to do more test.

I am missing one more component, once I picked it up I'll get back to you. I should have it by today.

 

When I short Green and White, do I short them connected to the IC? What should I expect my output to be?

hi k,
Initially keep Gn & Wh connected to the to the IC [INA129], short Gn & Wh together then measure Vout from the INA120, the Vout should be close to zero volts, let us know what you measure.

If you get a high voltage reading for Vout and if you have a DVM [meter] that can display DC mVolts.
Disconnect the Gn & Wh from the OPA inputs and measure directly across the Gn & Wh wires, let us know the mVolt reading.
[the Rd and Bk must be connected as normal to the INA and the project must be powered On for these tests]

I have taken the equipment from my school to my house to be able to do more test.
This made me chuckle, we've all been there.!

E

 

Hey,

Hahaha, I have the equipment for a whole weekend to fix

* I tried the same circuit with a 1kOhm resistor for Rg but the output seemed to saturate at 2.35V.

** With the same setup previously (using the 10kOhm for Rg), I shorted Gn and Wh and I measured 1.5mV.

*** I decided to measure across Gn and Wh just to see what I would get and I got 200mV.

What do you think? The 1.5mV reading seems to be ok

 

hi k,
I would say your L/C is well off calibration.
You could try to inject a offset null into the Gn or Wh terminal signal wires, to cancel the 200mV offset.
Do you want to try that.?

The 1.5mV test looks OK.
E

EDIT:
Please confirm that you have the wiring as this d/s clip.

 

I am not sure how to inject a offset null into the terminals. How would I do that?

I will check again, I will rewire as well to make sure, I will post if anything changed.

Btw I appreciate the help a lot!

 

Here is a picture of the circuit, this might help

 

hi k,
Will check your layout, let you know if any problems also a offset nulling method.

With regard to the L/C, going back to your readings of:
Cables Resistance
Green to White 357 Ohm
Red to Black 354 Ohm
Red to Green 281 Ohm
White to Black 289 Ohm
&
1) To make sure the L/C was connected I measured
Red to GND 2.49V
Green to GND 1.22V
White to GND 1.28V

and the fact you have measured a static offset voltage of 200mV, I would say the L/C is damaged.

It will be possible to inject a 200mV offset null voltage back into the L/C bridge, BUT it highly likely that the calibration of the L/C has been degraded and will never give accurate signal readings.

E

 

hi k,
Ref your B/B layout.
Why is pin #5 linked to #4 and #14 and not to 0V, which it should be.?
[keep the link on #4 and #14]

4 core screened L/C cable:
Black to 0V, OK
Green linked to a Yellow wire then to pin #6, OK
White linked to a Red wire then to pin #7, OK
Red linked to a Blue wire then to pin #4, OK
Connect the screen to 0V

You have a 10K Rgain it should be 100R, pins #8 & #9

Add a 47nF or 100nF across the +5V and 0V rails, also at least a 10uF cap.

Redo the tests in post #13, after you have made these changes.

E

EDIT:
Please check your board against this attached image A004, I had a pin error on the previous circuit.!
Pins 8 & 9

 

Hello!

"and the fact you have measured a static offset voltage of 200mV, I would say the L/C is damaged."

I think this might be the case :/ I ordered another load cell to test and see if that is the case.

"Why is pin #5 linked to #4 and #14 and not to 0V, which it should be.?"

I just noticed this, that connection is from post #2, I will connect it to ground and see the outcome.
If the Vout is not zero( 200mV when there is no force in a L/C) then would this mean the cell is broken?

"Connect the screen to 0V"

Is that the "shield"? I though that was optional or not? It looks very weird so I didn't connect it

"Please check your board against this attached image A004, I had a pin error on the previous circuit.!
Pins 8 & 9"

Don't worry, it's correct, the mistake I had was the pin #5 and pin #4 connection.

I will retry the test with 100 Ohm for Rg.