# Making a model of a transformer

According to this link, the equivelent circuit diagram of a transformer (seen from the secondary side) looks like:

Now, I want to find the values of the following constants:

$$R_p,X_p,R_c,X_m,R_s,X_s\tag1$$

So, my idea was to set $V_p=10\sin(2\pi ft)$ and $n=N_p/N_s=1$. Now I vary the input frequency and I measure a RMS current trough the load resistor that I put in series with $X_s$ (at the output). If I do that for six different input frequencies I can solve for the six constants in $(1)$.

Can I use that method of finding all the constants that I described in $(1)$?

• Just FYI, number of measurements != number of parameters that you can estimate. – loudnoises Apr 24 '18 at 13:01

## 1 Answer

The core losses Rc are quite strongly frequency dependent and level dependent, so varying the frequency will lead to mis-modelling of all the lossy R terms. I would expect the system of 6 equation to be quite ill-conditioned, so very sensitive to small reading errors.

Normally, we estimate Rp and Rs from DC measurements. We estimate Lp from an open circuit measurement and Lm from a short circuit measurement, as described in the rest of the paper you've linked.