Physical mechanism that makes opamp input track each other:
Opamp inputs track each other only in the case of negative feedback. In the other two configurations, positive feedback and open loop ,there is no virtual short between the two inputs of opamp. So one should first make this clear that this property of inputs tracking each other or virtual short between the two input terminals is a characteristic of the negative feedback configuration. This theory of negative feedback was originally developed for control systems and then after some time it was used in analog circuit design. To make you understand the physical mechanism I would first start with the negative feedback theory. Consider the figure below:
In negative feedback we are subtracting y
(output) from x
(input) and the result e
is then given as input to the amplifier. The amplifier amplifies e
and the result multiplied by the feedback element is then subtracted from input. When the input signal increases, error e
also increases, correspondingly the output of the amplifier also increases and therefore y
also increases. Now this increased value is then subtracted from x
. When the input signal decreases, error e
also decreases, correspondingly the output of the amplifier also decreases and therefore y
also decreases. Now this decreased value is subtracted from x
. Thus we can see that the negative feedback is working to minimise the error. This property of minimising the error between the inputs is a characteristic of the negative feedback. Now to understand we use the opamp in place of the amplifier A block and the feedback network is replaced by the R2 and R1 network. To understand the physical mechanism one should also know what are the blocks inside the opamp. An opamp basically consists of three stages: differential amplifier at the input, an intermediate amplifier for gain and an output amplifier as the output stage. The differential amplifier acts like a subtractor of the negative feedback configuration shown above , subtracting the output of the feedback network from the input, thus minimising the error between the two and this is how the two inputs of the opamp track each other. Although for calculations we do take that the two terminals are virtual short but there is a potential difference between the two inputs as the opamp we use is not an ideal voltage amplifier. As the specifications like voltage gain(infinite for ideal), input impedance(infinite for ideal) of the opamp we are using moves closer to that of the ideal voltage amplifier the potential difference between the two terminals decreases and we can say that the two terminals are virtually short.