0
\$\begingroup\$

TL;DR My wheatstone bridge won't respond to small weight changes when using single supply.

I'm trying to wire up this circuit below using these 5kg load cells.

INA125 Wheatstone Bridge Load Cell

I am able to get the circuit working with a single supply to the INA125P but the output does not change (stays around 54mV) with small weights in the 10-100s of grams range and will then start working normally with larger weights.

I switched over to a dual supply which seems to work well and is extremely sensitive but I'd rather not need to add to the circuit for negative voltage nor have to offset the entire output voltage as to only output positive voltages to the ADC since it can't handle negative inputs.

Being as this is a prototype for a project I'd like to eventually make a small batch of, I'm amenable to using nearly any other configuration of electronics so long as cost stays fairly low. A different instrument amplifier that may be rail-to-rail, adding in an ICL7660 for negative voltage, or any other recommendation is welcome. That said, I'm extremely limited in physical footprint for the scope of this project and could not use a board like the HX711.

So my questions are:

1) Why am I not getting any response from the bridge with small loads when using single-supply and how can I fix it?

2) If it comes to needing to use a negative voltage, would an alternative rail-to-rail amplifier be a better selection bang-for-buck?

\$\endgroup\$

3 Answers 3

0
\$\begingroup\$

Why am I not getting any response from the bridge with small loads when using single-supply[?]

Your in-amp does not have rail to rail outputs. The minimum output voltage is (V-)+0.4 V:

enter image description here

With a single supply, if the weight measured is not enough to produce an output above 0.4 V, then the in-amp will not give a valid output.

how can I fix it?

See FIGURE 6 in the datasheet. By connecting IA_REF to the VREF2.5 pin, you can produce a differential output referred to 2.5 V, so that neither input needs to be near ground. You can then convert this signal with a differential ADC.

\$\endgroup\$
1
  • \$\begingroup\$ @scott, Great!. If you still have a problem with very small strains, you should look at the offset voltage issue mentioned in the other answers. \$\endgroup\$
    – The Photon
    Commented Nov 3, 2015 at 19:35
0
\$\begingroup\$

It's likely you're being hit by a couple of nonideal characteristics of opamps: offset voltage, and output swing.

Every opamp has an offset voltage, meaning that it will output slightly above or below the desired setpoint voltage. Since your output is ground-referenced, if the offset is negative, the INA125's output won't rise above ground until you've exceeded its output offset voltage. This is exacerbated by any inaccuracies in your resistor bridge, too. You can observe this if you graph the opamp's output in response to input and observe where the graph crosses the Y axis.

Additionally, no opamp can really output exactly ground. If we look at INA125's datasheet, we can see that typical negative output swing is V- + 0.15 volts - so any output that should be less than 0.15V will instead be about 0.15V.

The solution for both of these is to trim the output using the VREF terminal, as described in the "Offset trimming" section. Though the example schematic there uses a bipolar supply, you can adapt it to a single supply with a simple voltage divider as input to the buffer amp. If you're feeding the output of this into an ADC, provide it with the VREF signal as well, and do a differential conversion to get an accurate result.

\$\endgroup\$
0
\$\begingroup\$

Although I have not used that particular instrumentation amplifier, I'd hazard a guess that you have an input-offset-voltage problem.

Reading the data sheet for this device leads me to believe that if you introduce a small positive offset voltage at the IAref terminal, you may eliminate the problem.

Based on your comments, I suspect that raising the IAref terminal perhaps 50mV above ground may fix the problem. Do note that whatever voltage you raise that pin by will also offset the output voltage. That is: with no weight on the load cell, the output voltage will be at some voltage above ground instead of at zero volts.

In fact, that may be the best way to adjust the amount of offset that you introduce to the IAref pin. Make that voltage variable and increase it until you see the output start to go above ground.

Note that the impedance seen at the IAref pin must be low. I'd put a 10 Ohm resistor from that pin to Ground, then use a high-value resistor from your Vref pin to introduce the offset voltage.

Please let us know if this works for you.

\$\endgroup\$
1
  • \$\begingroup\$ Typical offset voltage is 50 uV. Typical minimum output voltage is 400 mV. Even with his gain of about 200, the output voltage is about 10x bigger as issue than the offset voltage. \$\endgroup\$
    – The Photon
    Commented Nov 3, 2015 at 18:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.