0
\$\begingroup\$

What happens if I grab a 220 sine wave 60 Hz transformer with an output of 12v if i feed it with a 220v square wave? And what happens if I increase the frequency? Like 100 Hz? And if I grab the secondary and set it to 220v I know that it will burn out the transformer, but if I put a resistor high enough to limit the current in the secondary it will burn out too? Or will I get high voltage in the primary?

\$\endgroup\$
4
  • \$\begingroup\$ Do an FFT of the signal and keep in mind that the core losses increases about linearly with frequency. \$\endgroup\$
    – winny
    Jul 18, 2016 at 20:10
  • 1
    \$\begingroup\$ 220V on the secondary will saturate the core, drawing excessive current and hopefully blowing fuses.You'll get excessive voltage on teh primary, determined by the saturation, so probably only a fraction of the 4kV you'd expect. \$\endgroup\$
    – user16324
    Jul 18, 2016 at 20:57
  • \$\begingroup\$ Voltage time area is constant or you will saturate it if you go over. If you increase the frequency you can increase the voltage but the core losses will rise too. If you just need higher voltage, you can unwind the secondary and wind your own but you can easily get yourself killed in the process. \$\endgroup\$
    – winny
    Jul 18, 2016 at 21:09
  • \$\begingroup\$ I would think that you would need to look at where to add capacitance, series inductance and skin effect resistance to the equivalent circuit. \$\endgroup\$
    – user80875
    Jul 18, 2016 at 21:14

2 Answers 2

1
\$\begingroup\$

Depending on how accurate you need results to be, you can model a transformer as equivalent to a series resistor, inductor, and an ideal lossless transformer.

See here for the equivalent circuit of transformer. You can measure your transformer to find the parameters for the model. If you have a model you can study performance versus arbitrary inputs.

A 220:12 transformer has a ~18:1 turn ratio. If you apply 220V to the secondary it will develop 4 kV on the primary side, which could exceed the breakdown voltage of the insulation coating on the primary side wires, causing a short-circuit in the primary side or possibly destroying the insulation.

Mains is not to be trifled with. Do not exceed the voltage rating on a transformer, and if this is an experiment you should have an isolating transformer or residual current device/ground fault circuit interruptor protecting you.

\$\endgroup\$
1
\$\begingroup\$

A normal transformer is operating pretty close to core saturation - please take note that core saturation has nothing to do with loading currents taken by the secondary. All of the saturation story occurs unloaded.

It is the primary H field that is saturating the core (too much flux density) and that H field is determined by ampere-turns divided by length around the core. The current is determined by the primary winding inductance and of course "turns" are turns in the winding.

So, swap the drive voltage to the secondary (one eighteenth of the turns), and you have an inductance that is the square of one-eighteenth i.e. secondary inductance is \$\frac{1}{336}\$ of the primary. Then you apply 12V RMS (secondary nominal rated voltage) and you would note that the current it takes x number of turns (\$\frac{1}{18}\$ of primary turns) divided by length around the core (same as before) gives you exactly the same H field that was borderline causing flux density saturation.

Try doubling the voltage to 24V RMS and you begin frying the core rapidly.

If you increase the frequency by 2:1 the current halves (due to doubling inductive reactance) and saturation no longer is a bad problem. You can now drive 24V RMS on the secondary and get 440 V RMS out of the primary.

However, you will hit another problem; the laminations will progressively short out the windings as frequency increases. If you made a core from solid silicon steel you have, in effect, one massive shorted turn aka the core. So laminations are used and each lamination is insulated from each other thus causing eddy currents to be minimized but, as frequency increases, you will need a lamination width that reduce proportionately.

Do you see the problem in what you are asking?

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.