Hi, i designed an amplier i.e. signal condition circuit for strain guage sensor to get +/-5V amplifier output. here i used LT1014 quad op-amp as lnstrumentation amplifier. When strain gauge at balanced condition i was getting amplifier output 10 mV as offset . Therefore my question is could anyone suggest me that how can I implement auto-zero amplifier with LT1014(Quad op-amp) IC to reduce 10mV offset??

 

Hi, i designed an amplier i.e. signal condition circuit for strain guage sensor to get +/-5V amplifier output. here i used LT1014 quad op-amp as lnstrumentation amplifier. When strain gauge at balanced condition i was getting amplifier output 10 mV as offset . Therefore my question is could anyone suggest me that how can I implement auto-zero amplifier with LT1014(Quad op-amp) IC to reduce 10mV offset??

I believe that is not possible. How would the stain gauge know the difference between real signal and no signal?
If you are using a strain gauge bridge it is normal practice the balance the bridge manually.

The next option would be to zero the amplifier output with offset trim or DC offset input.

You might also use AC coupling depending on your range of input frequency.

 

Doing auto-zero externally is rather complex.
Much easier is to use an op amp with a built-in auto-zero circuit such as one of these, which generally reduces the offset to the low microvolt region.

 

It could be as simple as a very slow R-C Filter on the Output.
It depends on exactly how this device is being used.
.
.
.

 

It could be as simple as a very slow R-C Filter on the Output.

How will that help a DC offset in the amp?

 

I doesn't help with the Amplifier Offset,
it would simply counter-act it over a very long period of time.

I'm basically referring to using an automatic, extremely slow changing, "Sample and Hold" Circuit
to provide an Output correction to "zero" over an averaged time period of, lets say, an hour,
but the actual measurement use time for the device is less than one second,
and maybe only once per minute.
You would get a maximum of 1/3600 of the current error that is
currently being displayed with the real-time, uncompensated, measurement.
This would require over an Hour to zero the reading automatically,
but would zero immediately with a single button push, or at power-up.

Like I said, it depends on the use case,
what, and how, and how often, is something being measured .

The use of a "Strain-Gauge" usually indicates weight measurement.
I don't understand how anyone could expect perfect, absolute, repeatable readings from
any device that experiences physical movement, or dimensional changes with temperature.
But maybe certain types of measurement devices are good enough that people
start to treat them as absolute measurement standards, like a Machinists "1-2-3" Block .

Without knowing any of the usage details .........
I would use 2- identical "Strain-Gauges" and zero them against each other.
Then mount them side-by-side, with only one Strain-Gauge experiencing the measured "Strain",
this would compensate for Drifting Temperature/Humidity/Resistance and External Noise,
in both of the Sensors, simultaneously, and the Amplifier .

The Differential is what is wanted, and nothing else,
so ALL components, from beginning to end,
must be identical,
but with reversed electrical polarity.

Now I'm certainly not a professional Instrumentation Design Engineer,
but this is how I would do it if I was looking for the most accuracy that I could achieve ...........

 

Can't get much easier than this:


Chopper stabilized opamp.

 

I think we may have lost the TS.

 

I hope he found an answer.