# Easy way to estimate the power handling capability of an unknown transformer

I have quite a lot of old scrap mains step-down transformers (220V-240V 50Hz), salvaged from old equipment during 30+ years.

Although I religiously marked the primary terminals of all of them as a "best practice", most of the times I lost track of (or I could never get) data about secondary maximum current and, especially, VA rating.

Most of them are classic E-type cored ones, but a few are toroidal.

Is there a way by simple measurements to discover a conservative estimate of their max VA rating? If I could (under)estimate that value within a 15-20% tolerance I would be happy.

IIRC I heard something about weighting the transformer and apply some proportionality factor (k VA/kg). Is that a reliable way? And which is this factor? Is it the same for any size/weight (I think the heavier I've got weighs about 10kg)? Does this method work also for toroids? Are there better methods that are more or less that easy?

## EDIT

(to clarify some points)

• I asked for simple measurement techniques. Ideally they should be feasible with a True RMS multimeter and an oscilloscope, but this latter possibly only to monitor the secondary (I don't have an isolation transformer nor an high voltage probe). And I don't even have a variac, but I don't consider having a variac something in the realm of "simple" measurements.

• I already know, as I've stated before, the primary voltage. Knowing the secondary voltage is just a matter of measurements. The only vital parameter to reuse those transformer is knowing an estimate of their VA rating.

• Please, provide a clear procedure. Saying that by measuring the secondary wire gauge will tell me its ampacity is not enough. Which formula do I have to use? I never designed a transformer, so I don't know how secondary winding geometric parameters are chosen.

If you have an oscilloscope, an ammeter, a variac, and several loads, you can crank up the voltage on the primary side and measure the waveform of the secondary under increasing loads.

When the waveform starts to distort, the core is saturating. Watching the current on the primary to make sure you don't exceed the capability of the input wires is also a good idea.

Once you have a working input voltage, run the transformer at about 80% of the saturation load for awhile to see if it gets too hot. (Do this on a non-flammable surface, with a fire extinguisher nearby.)

• Sorry, I asked for a simple way to estimate that. I don't have a variac and I don't consider having one a common thing. BTW, I don't need to know the primary voltage, I already know that, as I said in my question. – Lorenzo Donati -- Codidact.com Apr 9 '18 at 16:45

IIRC I heard something about weighting the transformer and apply some proportionality factor (k VA/kg). Is that a reliable way?

Yes

And which is this factor? Is it the same for any size/weight (I think the heavier I've got weighs about 10kg)? Does this method work also for toroids? Are there better methods that are more or less that easy?

Go to an online catalogue, like RS for instance, and look at their transformer pages. The VA/kg factor will vary between E-core and toroidal, and will vary slowly with weight, so make sure you compare the same type and similar weight of transformer.

Once you've established a rough total VA for the transformer, how do you establish the VA for each secondary?

As you have already identified the primary, plug it in, and measure the voltage of each secondary. Unplug it, and measure the resistance of each secondary. The rated VA of each secondary will be roughly proportional to its $V^2/R$. This R and V measurement neatly combines effective wire length and area measurements.

How reliable is this method?

It's an estimate, to get you into the right ballpark. A transformer's rated power is just that, a power under certain specified conditions. By default, the typical power transformer in the online catalogue will be rated for temperature rise under continuous use with a resistive load, and if that's how you want your transformers rated too, then the estimate will be reasonable.

Unfortunately, we can't look into the transformers you have to determine the rated maximum operating temperature of the insulating enamel on the actual copper wire used. It might be 75C, it might be 105C, it might be more. In terms of temperature rise above a 25C ambient, the difference between a 50C rise and an 80C rise corresponds to sqrt(8/5) = 26% current.

Have any of the transformers you have, or can see in the lists, got 'spare weight' in them? Brackets, resin encapsulation, weight that isn't VA-contributing copper and iron? That will cause a systematic error in the estimate.

Once you have an estimate for the transformer VA, run the transformer into that load, and feel it after a few minutes, if it's not too hot, run it some more. Unplug it and measure the winding resistances again. Copper has a 0.4%/C temperature coefficient of resistance, which means they will increase in resistance by about 10% for every 25C temperature rise. This means you can accurately assess the temperature rise of each individual winding. Personally, I tend to be conservative and stick at a 50C maximum rise, assuming the cheapest enamel insulation.

• (+1) That's nice, and really what I was looking for: a simple and reliable procedure that didn't require complex setups. Thanks! I'll let the canonical 24 hours pass before accepting it. – Lorenzo Donati -- Codidact.com Apr 9 '18 at 17:03
• BTW, you confirmed it is a reliable methods. Do I need some safety margin after estimating the VA ratings? Could I stick with the computed values without risking overloading the transformer, or is it better to derate them, say, 10% or more. What derating factor do you suggest? Or I'm just being too paranoid? – Lorenzo Donati -- Codidact.com Apr 9 '18 at 17:07

You can use a variac to (safely) identify primary and secondary, and you can caliper the wire to determine gauge and therefore ampacity. From this you can get a VA estimate.