[SOLVED] Need help using LM358N as instrumentation amp

I am trying to hook up a pressure sensor to a National Instruments DAQ analog input to read the pressure reading.

However as the sensor is unamplified it is too noisy to read. I built an instrumentation amplifier to amplify the circuit with some LM358N that I had lying around. I followed the example circuit in the data sheet.


Using an excitation voltage of 12V for the pressure sensor, and the same 12V for the LM358N single supply.
It seems to be working, with a caveat, I do not get the calculated amplification. I am using R1 =100K and R2 = 10K, which should give me a gain of 21.

However using a 500mV square wave, i could only obtain a 1V square wave on the output.

I am not sure what is going on here though I my guess is that the opamp output does not have enough voltage to swing about.

Could some one help me out here?


Thanks!

Perhaps , it is better to use INA series Instrumentation amplifiers.

cmotif said:

I am not sure what is going on here though I my guess is that the opamp output does not have enough voltage to swing about.
Could some one help me out here?
Thanks!

Click to expand...

You must use split supply, unless you provide a specific biasing circuitry for single supply operation.

LvW said:

You must use split supply, unless you provide a specific biasing circuitry for single supply operation.

Click to expand...

hi do you mean the non inverting output of the final opamp should be tied to 6 volt via the 100k resistor? which is half of the supply.

otherwise I do not quite understand what you mean by a specific biasing circuit.

- - - Updated - - -

mvs sarma said:

Perhaps , it is better to use INA series Instrumentation amplifiers.

Click to expand...

hi, thanks for the recommendation. I am trying to use the opamps I have off hand hence am not purchasing new ones.

cmotif said:

hi do you mean the non inverting output of the final opamp should be tied to 6 volt via the 100k resistor? which is half of the supply.

Click to expand...

The output will look after itself. It is the inputs that should be biased to about half the supply voltage. When the inputs are connected to pins 2 and 4 of the sensor, the sensor should provide this bias.

The circuit looks like it should work fine. Maybe there was just some problem with the way you tested with the square waves. Have you tried it with the sensor?

godfreyl said:

The output will look after itself. It is the inputs that should be biased to about half the supply voltage. When the inputs are connected to pins 2 and 4 of the sensor, the sensor should provide this bias.

The circuit looks like it should work fine. Maybe there was just some problem with the way you tested with the square waves. Have you tried it with the sensor?

Click to expand...

I have tested it out with the sensor with a split supply for the amp and the gain is correct. Hence the problem is how I hooked up the amp.

I am now trying to see if it works just as well with a virtual ground at 6Vdc.