That sounds really stupid, but yea I think I'm doing it wrong.
I have a 50 kg half bridge load sensor with .1mV/V gain.
(Amazon.com)
I put it in parallel with two 1000 ohm resistors.
I attached the circuit I made on ltspice here. I'm absolutely confused how to connect this.
I assumed I would finish the bridge, then amplify the signal, but nope.
The second op amp is a comparator which my arduino is replacing.
Google "instrumentation amp circuit".
Two fixed resistors in series between +5volt and ground, with the tap to one input of the instrumentation amp.
The load cell also between +5volt and ground, with the tap to the other input of the instrumentation amp.
Google "INA125 datasheet", and look at the big diagram.
When you understand this all, dump what you're doing.
Opamps and an Arduino 10-bit A/D = bad resolution, temp drift, hum, etc.
Use an HX711 breakout board (ebay).
Leo..
Definitely need an instrumentation amplifier configuration, which is 3 opamps internally - using
one of the purpose build chips for the purpose is much the easiest route. And you don't have
any offset trimming which you'll need I'm sure.
The two resistors forming the other half of the bridge must be precision low-tempco resistors,
something like 0.1% tolerance is a good start, and should be mounted close together to minimize
temperature difference. You'd normally trim one arm of the bridge with a high value variable
resistor in parallel.
This sort of circuit is inherently high precision and you cannot get away with ignoring that.
Couple other things that need mentioning:
10,000x gain on an op amp circuit is HARD. Dave Jones of EEVBLOG did a video about input offset voltage concerns when he was designing his uCurrent, and his amplifier only had 100x gain. The OP37 you have in the LTSpice simulation, with 60uV maximum input offset voltage, the 10,000x gain will change the output by up to 600 mV.
Also, on your schematic, you have the Variable-R8 junction shorted to the R7-R9 node. That is not how you create the other half of a Wheatstone bridge. The output voltage is taken between the two points, they aren't connected together.
okay thank you everyone! i already own the ad8226 instumentation amplifier, so would that be a good enough alternative to ina125? From what I can tell, ad8226 just has one reference voltage pin, while ina125 has a few. I'd just ground it anyways.
meaghanp:
okay thank you everyone! i already own the ad8226 instumentation amplifier, so would that be a good enough alternative to ina125? From what I can tell, ad8226 just has one reference voltage pin, while ina125 has a few. I'd just ground it anyways.
The INA125 also has an excitation voltage regulator. An unstable load cell voltage is unstable results.
With an instrumentation amp, you're still stuck with Arduino's low resolution 10-bit A/D (~1000 A/D values).
As said, use a HX711 breakout board.
The HX711 is instrumentation amp, PGA, load cell voltage regulator, mains hum filter, and 24-bit A/D all in one.
Leo..
Wawa:
The INA125 also has an excitation voltage regulator. An unstable load cell voltage is unstable results.
With an instrumentation amp, you're still stuck with Arduino's low resolution 10-bit A/D (~1000 A/D values).
As said, use a HX711 breakout board.
The HX711 is instrumentation amp, PGA, load cell voltage regulator, mains hum filter, and 24-bit A/D all in one.
Leo..
In case OP doesn't understand the lingo, excitation refers to the power applied to the bridge circuit. This voltage needs to be as noise-free as possible, and should not just be connected to a digital power rail (meaning Arduino's 5V pin it out), and especially not a switching regulator output.
INA125 has an internal reference, but if you want to use a different amplifier you should be looking at voltage reference chips to excite the bridge, not voltage regulators.
