The build-in voltage, Vbi, has a specific value for each pn junction at thermal equilibrium: around 0.7V ... 0.9V. (Vbi ~ kT/q * ln (NA * ND / ni2))
Using a voltmeter between anode and cathode, it injects current thru terminals, so that the equilibrium don't still occur.
For zero currents, a zero voltage is measured.
How can we simultaneously capture zero currents (for equilibrium conditions) and real Vbi that is non-null.