Im trying to read the resistance of a sensor. The sensor resistance varies from 0 to 500k Ohms. I'm supplying the sensor with a voltage of 1V and im reading current that flows through the resistor with a sens resistor and an opamp to amplify the signal. The circuit is in the attachment. I've simulated it and it works perfect. Lets say for example my Sensor is 430k Ohms and im using a 10k ohm sens resistor. The opamp amplifies the voltage across the sens resistor by 20 times. The voltage across the sens resistor is around 30mV and the output signal of the opamp is 600mV.

The used opamp is AD8418, it amplifies the signal by 20 times and it has an offset voltage of 100uV. The supply voltage range is between 2.7V and 5.5V.

But after testing it on a breadboard, the voltage across the sens resistor acts weird when i plug it into an opamp. So the voltage across the sens resistor is perfectly 30mV without the opamp. But when i connect it to the input of the opamp, the voltage becomes 45mV instead of 30mV. And the output voltage is 900mV.

The question/problem is: Why did the voltage change across the sens resistor? If the sensor resistance decreases then the error decreases a bit as well.

 

I'm not going to read through the entire AD8418 data sheet, but what jumped out at me on a quick scan of it is that the input bias currents are spec'd at 130 μA typical. Given the extremely high value of your sense resistor, this could result in as much as 1.3 volts of input offset error. That may be where the problem lies.

 

I'm not going to read through the entire AD8418 data sheet, but what jumped out at me on a quick scan of it is that the input bias currents are spec'd at 130 μA typical. Given the extremely high value of your sense resistor, this could result in as much as 1.3 volts of input offset error. That may be where the problem lies.

Is there a way to calculate it?

 

Is there a way to calculate it?

Yes. Voltage = current times resistance: 130 μA * 10000 ohms = 1.3 volts. At least, that's what it would be if your unit had the "typical" input bias current AND if input bias current was the cause of your problem. It might or might not be. I suggest you read through the data sheet THOROUGHLY to see if something else might be going on.

 

Yes. Voltage = current times resistance: 130 μA * 10000 ohms = 1.3 volts. At least, that's what it would be if your unit had the "typical" input bias current AND if input bias current was the cause of your problem. It might or might not be. I suggest you read through the data sheet THOROUGHLY to see if something else might be going on.

Thanks sir, it's indeed the bias current thats causing the problem. I've tried it with 200 ohms sensor and a 10 ohms sens resistor and it almost didnt have any error in it

 

Rather than directly grounding the (-) terminal, connect it to ground through a resistor of the same value as your current sense resistor. That should give you some relief.

 

Thanks sir, it's indeed the bias current thats causing the problem. I've tried it with 200 ohms sensor and a 10 ohms sens resistor and it almost didnt have any error in it

Good.

But given how you're using this chip, I have to ask: why don't you simply use an ordinary op amp configured as a non-inverting X20 amplifier instead of this very specialized (and rather expensive) part? You aren't using any of its special features (such as its extended input common-mode range) as far as I can tell.

Even a lowly LM358 would suffice; if its input offset voltage is too much, you could use an LMC6482. Either one is a lot cheaper than the AD8418. To completely eliminate input offset voltage error, you could use a MAX44241.