# Inductive “kick” and possible dangerous experiment

So I have this unmarked, unknown transformer (probably from a microwave oven so it is pretty beefy) and I thought that a safe way to determine the primary from the secondary windings would be to put 1 volt of dc connected to one of the windings and interrupt it with a switch. Then I put my multimeter across the other windings and tapped the switch pretty quickly (say 5 times a second). My thinking was that I would either see an increase or decrease from 1 volt at the meter and know the ratio of turns in the two coils.

Two things seem to be wrong with this idea:

Firstly, the experiment seemed to step-down the voltage regardless of which side of the transformer I connected it to.

Secondly, at the time I didn't know about "snubber networks" so I did not include one. I've just now come to understand a little about "inductive kick" and realize that though it was probably too transient for my meter to register, it is possible I was generating voltages on the secondary upwards to the kilo-volt range and reverse voltages on the primary in the mega-volt range. I say this because I simulated this with Lt-spice and though I know I didn't account for capacitive coupling between windings, core losses, wire resistance, etc. the simulation shows some pretty nasty voltages. I'm sure there must have been some bad arcing going on in the switch which dissipated the bulk of the energy; however, could it have damaged the multimeter, and if I had hooked it to an oscilloscope capable of detecting such a signal, would that have damaged it?

Here is the Spice schematic:

And the output:

• Those megavolt values you are getting are probably a consequence of the ideal model (hint: V = L di/dt in an inductor) of the transformer. In a real transformer (with losses) those spikes will be much less noticeable due to transformer losses. – Renan Dec 1 '12 at 22:15
• Renan is right - you would need to include some series and resistance parallel capacitance in your inductor model (right click, try a few pF to start with). In reality the voltage will be high, but unlikely to be anywhere near 1MV. Also, limit the switch on to off time to a reasonable value like > 1us or so. – Oli Glaser Dec 2 '12 at 2:17
• You can easily measure the series resistance of the coils. Include them in your simulation. Also, you can most likely determine which is the primary side by its resistance alone. The primary side will have more windings. Also, any parasitic resitance on the primary side has little effect on the secondary side. The manufacturer would intentionally make the secondary side using thicker wire, since it will carry more current than primary. That's why the primary will have more resistance. – Jonny B Good Dec 2 '12 at 9:13
• Thanks. Any ideas as to why the experiment failed to show a step-up in one direction and a step-down in the other? – Salbando Jones Dec 2 '12 at 15:50
• I would try putting an audio signal in each side. You could have a laptop generated a low voltage, low power 60 Hz sine wave and compare the voltage gain at each. Your experiment would fail if the transformer is in fact an isolation transformer, or if the transformer is sufficiently "low-pass" in character compared to tapping the switch that it dissipates most of the energy from tapping the switch internally. – jbarlow Dec 6 '12 at 7:23

To partly answer your question: The transformer will be designed for and and optimized for a certain frequency of operation. Your switch opening/closing had what fundamental frequency? You don't know, and you have no way of knowing. It is entirely possible that the chatter was at such a high frequency that the ferrite core just swallowed it -> "step" down in both directions -> really just means that the transformer can't pass the frequencies.

What you care about is determining the transformer parameters so, here are some steps you can take:

1) measure primary and secondary DC resistances. While you're at it you might as well measure between winding too, just for additional data. Be aware that a High voltage winding will have a higher resistance - clues will start popping up right away.

2) Do you have an inductance meter? - measure primary and secondary inductance with the other winding both open circuit and shorted -> 4 measurements in all.

3) next perform a frequency sweep driving the primary and then the secondary with other winding open circuit and closed circuit. Ideally you'd measure I and V to see if you get any resonances. Keep the voltage low. NO need to measure the other winding yet, in fact you probably shouldn't and be careful if it really is a HV transformer.

4) analyze results and see if you can fit it to a transformer equation:

5) with the results of 4) hook up the signal source again and measure what you get on the secondary with various loads from Open circuit to some low resistance (guided by 4 above). Start out at low voltage, get comfortable.

warnings: - if it is a HV transformer slowly approaching its performance while quantifying it should keep you and your eqt safe. Of course once you start to narrow down its characteristics, start looking at suppliers catalogues for suggestions and testing techniques.

• I'm sure there are better suggestions than my simple list above.