[SOLVED] OpAmp noob - mV Amplifier - help needed !

OpAmp noob - mV Amplifier (pH Sensor) - help needed !

Hello folks! An op amp noob here.

I've to design an amplifier ---
Sensor output : -400mv to +400mv *DC* (buffered,hence impedance notta problem)
Amplifier output : +1v to +4v ( boundaries not strict ; have to feed it to ADC of a PIC running at 5v )

First the negative voltage part :I'm working with a 0-5v rail so I'm planning to raise the ground of sensor to 1V by a potential divider(though resistors will introduce errors). That'll bring the swing to +0.6v - +1.4 v.
Do you people have any suggestions on this ?

Second - The amplifier. In an ideal world,I would just feed it to a non-inverting opamp with a gain of around 3.

However, I've no idea what role would offset voltages would play ? Do you think they will be enough to drastically reduce the accuracy of the ADC (I need it to be about 1mv,if possible)
Can you guys help me on that please ?

I'm planning to use a LM324. I'm trying to avoid the expensive op amps as it's a low cost project and also ,if possible,potentiometers.

Any help/suggestion would be greatly appreciated !
Thank you ! :smile:

Hi

Using a LM324 in non inverting config with appropriate resistor gain (around 3 you said) is a good and cheap option. However offset of this amplifier is 2 mV typ. (7mV WC)... You will have to live with this otherwise you need to choose a lower offset amplifier (thus more expensive).

For the neg voltage part I don't fully grab what you are suggesting. Any schematic to help out?

Cheers

Thanx chris!!
7mv WC means "without compensation",right?
And with a again of 3,this would get amplified to 21 mv,right ?
That means,for any output,21mv will get added (or subtracted). Am I right ?

And as for sensor ground ,this is what I was talking about :


This 1V from pot divider shall form the ground of my sensor. So when the sensor tries to output -400mv, the net outputwill be 1+(-0.4) = +0.6 v. I need this as I don't have a negative voltage supply.
Is my thinking on the right track ?

"Raising" the sensor ground is only suitable, if no error currents or interferences are injected by the connected circuit. If the sensor is e.g. connected through long cables, it won't do it.

Regarding OP offset voltage, you also have the option to cancel it in your software. You should consider, that the zero uncertainty caused by regular resistor tolerances (assuming 1% parts) is much higher than the OP offset contribution.

The LM324 will not allow you to get 4V output with a 5v rail.
From the data sheet - Large output voltage swing 0V to V+ − 1.5V.
You would only get 3.5V.

@fVm : I do have the option to cancel out the offset as my sensor outputs 0mv at a calibration condition.And the circuit is connected by a 25cm long cable but I don't have any other option. So I'll try my luck. Thanx for the tip !

@RJM : Thank you buddy. You just saved me from a major embarrassment ! I don't know how I skipped such a major spec from the DS :shock:

The LM324 will not allow you to get 4V output with a 5v rail.

Click to expand...

It may allow it, depending on other circuit parameters, by placing a pull-up resistor.