 Current-transformer excitation current

A current-transformer excitation current is the part of the secondary-referred input current needed for the magnetic core's response and losses rather than delivered to the external secondary circuit. In an ideal current transformer this contribution is neglected. In a real equivalent circuit it helps explain both current-transformer ratio error and current-transformer phase displacement.

Choose terminal polarities so ideal current reproduction is in phase. Let $K_n$ be the nominal primary-to-secondary ratio, $I_0=I_p/K_n$ the secondary-referred input phasor, and $I_s$ the measured secondary phasor. The current balance isRepresent the excitation branch locally by current-transformer excitation admittance $Y_e$ and induced secondary phasor $E_s$. With total secondary winding-plus-burden impedance $Z_t$,The NIST treatment gives the corresponding magnetizing, leakage and burden equivalent circuit and shows how those impedances enter the complex ratio.^[ref]

For an original local model, write $Y_e=G-iB$ with $G,B>0$, and take $Z_t=R_t>0$. When $|Y_e{R}_t|\ll 1$,The amplitude deficit is approximately $G{R}_t$, while the secondary lead is approximately $B{R}_t$ radians. For stipulated $G=0.001\mathrm{S}$, $B=0.005\mathrm{S}$ and $R_t=1\Omega$, these are about $0.1\%$ and 5 milliradians. At an induced $1\mathrm{V}$, the excitation components are 1 milliampere in phase and 5 milliamperes in quadrature; the real branch power is $G|E_s{|}^2=1\mathrm{mW}$.

An increased current-transformer burden raises the induced potential required for a given secondary current. The actual excitation response need not remain linear: core behavior, frequency, waveform and prior magnetic state can change the local parameters. The NIST equivalent-circuit discussion explicitly notes the nonlinear core and current/burden dependence.^[ref]

A proposed characterization varies a known connected burden and compares measured current ratio and phase, keeping the primary-current and frequency conditions recorded. A model that fits one operating point is not thereby validated through saturation or for arbitrary harmonic components. Separating excitation from load delivery explains why a turns ratio alone cannot specify practical measurement accuracy.