Amplifier offset

hi.. i've some problem in op-amp circuit and wish to have some help here.

when i shorted inverting and non-inverting of the opamp (tl084/tl082 - supplied with +-5V) to ground, the output should be 0V but came out around 5V. as i know this is because of the offset and there're some way to eliminate the offset but i'm not very sure of it.

i tried applying some voltage (hard time adjusting using variable resistor) in series at either input but could not get result of 0V. but when i apply some voltage at the output and trim it, i could get the output of 0V (i'd tried). is this the correct way to eliminate offset voltage or is/are there any other more practical way to do so? i'm confused as from what i read, it supposed to be applying some voltage in series with the input instead of output.

secondly, i tried eliminating the offset (apply voltage in series at output), i manage to amplify the input to my desired gain last week (in breadboard). somehow, i made the same connection (exactly the same) again this week (in PCB board) but the output seems to be increasing (even though i did not increase the voltage in series with output anymore).

is there any reason for the increasing voltage? i'm still figuring out how come i could not get the same output (maybe connection?!)

anyway, ideal op-amp without offset ever exist?

thanks..

A TL081C has a max input offset voltage of only 10mV and a typical voltage gain of 200,000. So of course its output will be offset if there isn't any negative feedback to reduce the gain.
Opamps are usually used with plenty of negative feedback to set the amount of gain to a usable number, extend the bandwidth and to reduce its amplification of its input offset voltage.
The single opamp TL081 has pins for connecting an offset adjustment pot.
You can connect a pot to an input to feed in your own offset adjutment.

More expensive opamps are available with their input offset voltage laser trimmed to be very low.

What is your signal that is so low in level that the input offset voltage matters?

audioguru said:

A TL081C has a max input offset voltage of only 10mV and a typical voltage gain of 200,000. So of course its output will be offset if there isn't any negative feedback to reduce the gain.
Opamps are usually used with plenty of negative feedback to set the amount of gain to a usable number, extend the bandwidth and to reduce its amplification of its input offset voltage.
The single opamp TL081 has pins for connecting an offset adjustment pot.
You can connect a pot to an input to feed in your own offset adjutment.

More expensive opamps are available with their input offset voltage laser trimmed to be very low.

What is your signal that is so low in level that the input offset voltage matters?

Click to expand...

thanks audioguru. if u remember my previous project using load cell, the signal is lower than 1mv when a load of 70kg is applied.

i need one opamp to amplify the signal to fit my voltage reference (+5v)and another for the peak detector circuit. hence i use tl802. i tried measuring the circuit i'd done and the output is around 5V even there's no input at all.

thanks!

So you are still trying to amplify the tiny output signal from a load cell?
Doctors use ECG monitoring circuits for displaying the tiny signal from a heart beating by using an instrumentational amplifier IC. Try one.

With signals level less than 1mV, the opamp offset should be lower than 10nV for 1% accuracy. How about drift due to temperature changes?

Much better to use chop-mode stabilised opamp in the front end in my opinion.

audioguru said:

So you are still trying to amplify the tiny output signal from a load cell?
Doctors use ECG monitoring circuits for displaying the tiny signal from a heart beating by using an instrumentational amplifier IC. Try one.

Click to expand...

yup! i was doing other part all this long as i never thought that i would have this offset problem. do you have any recommendation for common instrumentation amplifier?

actually i'm running out of time and afraid i could not get to replace it with another amplifier (my fault, sigh!). hence, i'm now struggling thinking ways to trim off the offset. if i'm able to get a multiturn adj pot, wonder if it could compensate the offset easily but connecting the pin1 and pin8 to it.

eblc1388 said:

With signals level less than 1mV, the opamp offset should be lower than 10nV for 1% accuracy. How about drift due to temperature changes?

Much better to use chop-mode stabilised opamp in the front end in my opinion.

Click to expand...

hi Chung,

i'm not really sure if its the temperature which cause the problem. may i know what actually chop mode stabilised opamp? thanks

You don't short the opamp's output to change its voltage because the opamp will use all its gain and output current to fight it.
You simply add a very small current to the inverting input that is the opposite to the offset.

Zane83 said:

hi Chung,

i'm not really sure if its the temperature which cause the problem. may i know what actually chop mode stabilised opamp? thanks

Click to expand...

For ordinary opamps, even when you carefully trimmed the offset to zero, difference in ambient temperature might affect the trimming and caused an output voltage which you don't want. This may not be a problem if the signal input level is a few volts, but yours is only 1mV so it becomes a big problem.

A chopper stabilized opamp would dramatically reduce or balance out the offset voltage of the opamp itself via chopping the input signal On and OFF so that you will always get zero voltage output for zero volt input.

You can get lots of information about them by entering "chopper stabilized amplifier" in Google.