# How to connect a load cell and INA125p correctly?

I have seen the attached floating online but i am unsure on the actual wiring of my load cell. I have also attached a picture of my load cell. The micro load cell of mine, has four wires and the data sheet says that Red = 5V, Green : + White : - and Black : Ground.

From the first picture showing the wiring schematic, is it safe to attach my Red wire to 5V coming the arduino 5V rail, and the other colours like the picture said(Green to 6, white to 7 and black to ground) ? I want to make sure because I dont have a spare INA125.

I was reading also about using regulated voltage inputs to the INA125p, but I am not sure how this regulation works.. Will the 5V from the arduino be sufficiently regulated ?

Also specifically i have a 0-780g load cell, what value should I make Rg be ? I am looking for a resolution of about 10g.

Kind Regards,
Iz

You need the datasheet for the load cell. Can you post a link?

Datasheet

Kind Regards,
I

According to the data sheet, the load cell wires are correctly identified. The output is 0.8mV/V at rated load, so that would be 4 mV output with 5V input. You would need a gain of about 1000 to bring this to 4V for the ADC, if using 5V for AREF. The resolution then would be better than 1 g.

Thanks @jremington.

I get 66mV at no load from the output of the amp. I just need to differentiate between three loading conditions : Full: 60g Medium : 20g Empty: 0g.

But when i put a 55-60g load i get 95mV, 20g changes to 68mV.

I used a 60 ohm Resistor for the gain resistor by paralleling two 100ohms and seriesing with a 10ohm.

Gain = 4 + 60kΩ/RG so 1000 = 4 + 60k/Rg, Rg = 60.24 ohms.

Can i get some insight as to where I went wrong please guys ?

I am thinking i should get another op-amp ? idk....

How are you measuring the output voltage of the amplifier?
What is the voltage applied to the load cell?
Are you using precision resistors for Rg? The gain could be higher.

The load cell evidently has an offset (imbalance) voltage which must be subtracted off. That can be done with a different circuit or by the Arduino.

I measure it using a voltmeter on pin 10 and grnd.

The voltage applied to the amp and load cell is 5.15V.

I am not using precision resistors for Rg. I should get some then ?

What ideas do you have to subtract that voltage circuit-wise ?

Update on the values : 5ohm Rg :- No Load:0.07V, 40g : 0.08V, 55g:0.1V

10ohm Rg :- No load:0.07V, 40g :0.08V, 55g:0.09-0.1V

60ohm Rg :- 0g : 0.07V, 40g:0.07-0.08V, 55g: 0.08V

I am thinking that I should use the 10V ref signal, so in that way these low values could scale up to higher voltages, and seeing that they are so small I doubt they will pass 5V in any case. As such they will give good 10bit ADC values which I could differentiate for three loading conditions. Because its only three conditions I need to differentiate.

Update on the values : 5ohm Rg :- No Load:0.07V, 40g : 0.08V, 55g:0.1V

10ohm Rg :- No load:0.07V, 40g :0.08V, 55g:0.09-0.1V

60ohm Rg :- 0g : 0.07V, 40g:0.07-0.08V, 55g: 0.08V

The circuit is not working, if Rg has such a small effect. I suspect that either you have a wiring error, an electrical noise problem, or the amplifier is oscillating. You should try adding bypass capacitors (100 nF in parallel with 10 uF) to ground at the Vin+ and Vcc terminals of the bridge and the amplifier. An oscilloscope would be extremely helpful in debugging the circuit.

The INA125P can't output rail-to-rail (0V to 5V). It should be around .5V to 4V. Since your results are outside of that range you must have something wired incorrectly.

Regarding Rg, a too-low resistor should show 4V; a too-high resistor should show .5V. You have to keep trying all the in-between resistances until you get an output around 2V. A low ohm pot (200ohm or less) is very valuable here when you're trying to find the right value because you want to get your Rg within a couple ohms for the best sensitivity and range.

Look at sheet #13 of the INA125 data sheet. This is Burr-Brown's recommended single supply circuit. If you just drive the bridge off the +5V rail, you will have issues at that high gain. They drive the bridge from a band-gap reference for a reason. It also provides the VCC/2 biasing to get in the middle of the Arduino's ADC range.

@Chagrin I wired everything to the image I have in post #1. I verified my connections more than twice....

@rmetzner49 Thanks, but I am having a little difficulty understanding due to my small knowledge in electronics. Are you saying then I should connect my pin 4 to pin 13 instead ?

@jremington

jremington:
According to the data sheet, the load cell wires are correctly identified. The output is 0.8mV/V at rated load, so that would be 4 mV output with 5V input. You would need a gain of about 1000 to bring this to 4V for the ADC, if using 5V for AREF. The resolution then would be better than 1 g.

Looking back now I don't really follow how you calculated the gain to be 1000.
Because, I may be wrong but I thought it was 0.8mV at 780g and according to @Chagrin the output for 0-5V rail to rail is apprx 0.5~4.0

Thus at 780g i should want 4.0V instead of 0.8mV hence 4/0.8mV = 5000

Shouldn't the gain be 5000 ?

jremington:
According to the data sheet, the load cell wires are correctly identified. The output is 0.8mV/V at rated load, so that would be 4 mV output with 5V input. You would need a gain of about 1000 to bring this to 4V for the ADC, if using 5V for AREF. The resolution then would be better than 1 g.

Yes 4mV by 1000 is 4V but how did you arrive at 4V in the first place ?

I need help understanding how to calculate the gain....

Thanks

At peak load it outputs .0008V (.08mv) for each volt of excitation voltage. Your excitation voltage is 5V, so .0008V * 5V = .004V.

The output swing of the INA125P is approximately 4V (4V - .5V is approximately 4V). The multiplier ("gain") required to bring .004V to 4V is 1000.

An ideal InAmp or OpAmp would be able to output voltages from GND to Vin (whatever voltage you're powering them with). But there's no such thing as an ideal InAmp or OpAmp, and instead they only get pretty close to GND and Vin. In the case of the INA125P this is stated on page 2 of the INA125P datasheet; in the Output section it states the typical positive is V+ - .9V and the typical negative is V- + .4V. And yeah, again that's approximately 1V and .5V. Those "min" values show a worst-case chip and the swing could be as bad as 1V to 3.3V (assuming a 0V to 5V system).

Apologies for using so many approximations, but in the end (again per the INA125P datasheet) it all tells you to start with a resistor around 60ohms. You really need to fiddle with that resistance to get things tuned right. You can try calculating it precisely but it's a bit of a moot point; you just need to get it in the ballpark and go from there.

Hi,
How it is fixed and where you are applying load, thanks .

Tom.....

Look at sheet #13 of the INA125 data sheet. This is Burr-Brown's recommended single supply circuit. If you just drive the bridge off the +5V rail, you will have issues at that high gain. They drive the bridge from a band-gap reference for a reason. It also provides the VCC/2 biasing to get in the middle of the Arduino's ADC range.

I have to agree with Metzner49. The INA125 is designed for a bi-polar supply . You can't run it off a single ended 5V supply. You would also probably need a differential op amp circuit to convert the +/- output to single ended 0 - 5V signal.
op amp differential to single ended

Look at the INPUT COMMON-MODE VOLTAGE vs OUTPUT VOLTAGE, VS = ±5V graph on page 5 and the Output voltage graph on page 7 of the datasheet. That says it all. You see the negative output voltage ? Also, the chip's resolution is a function of the differential voltage across the supply. +15V - (-15v) = 30V rail to rail. Running if off a lower +/- supply compresses the output ,reducing the resolution (I would think) .

If you want better resolution you should be putting the INA125P back in its package then go on eBay (etc.) and buy an HX711 module. You could run it on a wider +/- supply, etc., but you'll never get better than the Arduino's 10bit ADC accuracy.

but you'll never get better than the Arduino's 10bit ADC accuracy.

Unless you paid \$15 for this.

raschemmel:
Unless you paid \$15 for this.

I'd think the \$3, 24bit HX711 would be a better choice than a \$15, 16 bit ADC.

And the HX711 is made specifically for tasks like this. It is a dual input differential amplifier with integrated 24 bit ADC.

I'm not endorsing this specific eBay seller:
http://www.ebay.com/itm/HX711-Weighing-Sensor-Dual-Channel-24-Bit-Precision-A-D-Module-Pressure-Sensor-/161264280835