Whether load cell output in mV will linearly increment according to weight added to it?

 

Ideally, yes, but . . . .

A load cell is ratiometric, that is, its output is proportional to the power supply voltage.

So the output is linear taking into account the stability of the power supply and all the expected uncertainty factors: linearity, hysteresis, temp variations, and where the measurement is along the accuracy curve, and aging factors like how the bonding has degraded over time.

 

In case of zero load, the load cell is showing some mV output with supply voltage - whether this load cell is defective? Otherwise how this zero error can be adjusted? What is the proper method for calibration of load cell?

 

To have a considerable mV output from a load cell without any applied load may indicate an overload. This could be from direct capacity overload, off axis loading, or moments, such as from installation over torque. It can also be observed from fatigue from usage over time and high cycle counts.

If the zero is to0 high the sensor should be replaced. If the zero is within spec, usually found on the sensor spec sheet or product description page, then the sensor can be tested for performance by accounting for the zero through a Tare function, or subtraction of the starting zero amount from all sequential readings.