# Designing a transformer with low usage time but high power output for small interval of time

I want to design a transformer ideally for arc welding purposes. The thing is that I want it to be designed in a way that the windings allo overloading of the transformer itself.

For example I had a transformer from old radio, the area of core of the transformer was no more than that of a normal 100 watt transformer but when I shorted the secondary leads of transformer at around 40 volts the calculated power output was 650 watts! The varnish inside the transformer started boiling after a minute or so. But I used that transformer for two years, at least at 200 watt (not full time).

Now the thing is that this property of being able to be overloaded is related with number of turns per volt of the transformer. To be more specific, in normal transformers we divide 7.5 by core area in inches to get number of turns per volt. I know that to make "overloadable" transformer we have to do with this number (7.5) but I want to know figures to be in safe range without tripping my breaker at no load or frying the transformer before I even put any load on it since these transformers have high no load current, and have a working transformer at the end.

• ”The varnish inside the transformer started boiling after a minute or so” Don’t! You will not be able to design any overload potential by having too few turns in the primary as the core would saturate immediately. Urms=4.44fNAB always applies! Jan 28, 2019 at 7:14

There are design features that can be used for welding transformers. A fixed or adjustable magnetic shunt or a series reactor are two concepts that you should investigate. A magnetic shunt diverts magnetic flux to a path that reduces the magnetic coupling between the primary and secondary. That increases the stray reactance similarly to putting reactance in series with the transformer. That does nor really allow the transformer to be overloaded for a longer time, but it uses the transformer more efficiently by reducing the voltage once an arc has been struck. The same thin can be done electronically. How to design a welder is too broad a question to be answered here, but more detailed information can be found in resources that can be found on the internet.

Well, you would have to know how your breaker is set up... there is something known as inrush current and for about 3 line cycles you will go over the rated current, first cycle is around 25 times, then half that(12.5 times rated current) and finally down to a number close to rated current. This is approximate and not perfect. This only happens if you turn on the transformer and it was off.

So I would look up the curve of your breaker and the current you plan to run through your transofmer and see if it would trip it in the given time(3 cycles is around 50 milliseconds, depends on the frequency of your network) knowing the current and the time should be enough to find out if it will trip or not.

You want to avoid saturating your core as someone else already mentioned, if you have a BH curve or Vt curve available you can estimate how long it will take at a certain voltage for it to saturate and you should operate on a timeframe shorter than that to avoid overheating.

Each regular transformer does exactly what you want. If it is overloaded for long period of time, then it melts. Designing a transformer with higher flux density is just a trick, where you get a lighter transformer claiming the higher power, but it's a junk.

If you want a quality transormer, then you should buy a regular transformer with nominal flux density that is impregnated with resin. Adding a thermistor (or PTC, PT1000, KTY) and suitable thermal protection relay you may use it with high overload within small on time and then let it cool for certain time.

The problem with high inrush current at switch on is typical symptom of a transformer being wound for higher flux density than it should be. Also the output voltage of this kind transformer will sag when heavily loaded.

So my answer is: there is no overloadable transformer /or/ all transformers are overloadable. The best one has nominal flux density and all the window area filled with copper wires, and impregnated.

• Well i think all transformers are not overloadable talking in terms of power rating.I came across a transformer rated at 24V,6A.It delivered exactly what it was rated for.I think thhe manufacturer made it using least amount of copper possible as the transformer delivered rated power only when shorted.This means that the rated power was peak power.Maybe you are right.I figured it out while writing this comment. Jan 29, 2019 at 14:53

The transformers Short circuit current is determined by the Open circuit voltage divided by the total impedance .Leakage inductance can be manipulated to some extent by transformer design .You want lots of leakage inductance so techniques like magnetic shunts and split bobbins could solve your problem .If you have a standard transformer with low leakage inductance you can place inductance in the primary .The added primary inducter must be rated for mains and the expected current .Lamp ballasts work well here .