Determining current capability of transformer with no name plate

[Jester]

I'm adding CNC capability to a milling machine and need a suitable power supply for the stepper drivers. The stepper driver is rated for either 60Vac or 80Vdc (max). I dropped by our local surplus electronic store and found what is likely a good transformer for the application. It looks to be about 500VA  (size-wise), dimensions 5.375 x 4.5 x 3.5" and has AWG16 secondary leads, the only label was a piece of masking tape stating 108V CT. Price was $20 so I grabbed it.

I powered it up and the masking tape appears correct, about 55Vac centre tap to either side with no load. It would be nice to know the actual current or power capability.

Is there a way to estimate actual capability, perhaps by loading it and observing voltage drop?

[hlavac]

Transformer power is a square of the cross-section area of the inner leg of the EI core in cm^2. About 20% more for C core.

[Benta]

Transformer power is a square of the cross-section area of the inner leg of the EI core in cm^2. About 20% more for C core.

Huh? At a rough estimate, this one has a centre leg of 4 x 4 cm² = 16 cm² = 16 VA? I don't think so.

I'd estimate it at 200 VA, but without exact measurement it's impossible to say.

[hlavac]

Square the area. 256VA

[edpalmer42]

Two rules that I've used over the years are:

1.  Load the output until the voltage drops by 10% from the no-load value.  I think this rule came from a company that sold electronic surplus.

2.  Measure the cross-sectional area of the core.  Compare that to published data for similar transformers.  I think this rule came from the ARRL.

If there's a disagreement between the two, take the smaller value.

Remember that transformers are rugged beasts.  If you overload one slightly, it will just run a bit hotter.

Ed

[Benta]

Square the area. 256VA

Ah, missed the double square.
Is this rule for 50 Hz? 60 Hz? 400 Hz?
Don't think I'll add this one to my tips 'n tricks, but if it works for you...

[T3sl4co1l]

Measure the DCR of the primary, then double it.  This uses the assumption that the secondary uses proportionally sized wire.  This resistance is what causes the transformer to have worse regulation, and therefore suggests a practical current limit.  For a 10% regulation, use 10% of the mains voltage, i.e., 12V, then Ohm's law to get Imax = (12V) / (2*(pri DCR)).

It could also be 240V or more (typical for an industrial control transformer -- they're pretty common, and 108V possibly hints at three phase use?).  To check this, try connecting half the 108VCT winding to 120V, using a light bulb or load resistor to limit current.  Then measure the voltages.  If it's topping out not far above 59V (it may look like 70V or more because of the distorted waveform; a true RMS meter, or oscilloscope, would be ideal here), then it is indeed a 120:108VCT transformer, rated for the above current.

(Note that, because of the above regulation rule, using a transformer at half its rated voltage leaves you with one quarter the VA capacity -- a huge disadvantage.  On the upside, the core runs a whole lot cooler.)

Tim

[Benta]

Measure the DCR of the primary, then double it

Yes, and please short-circuit the secondary when doing this. Otherwise but you and your ohmmeter will get an unpleasant "whack" from the primary inductance. Just sayin'.

[Zero999]

Two rules that I've used over the years are:

1.  Load the output until the voltage drops by 10% from the no-load value.  I think this rule came from a company that sold electronic surplus.

2.  Measure the cross-sectional area of the core.  Compare that to published data for similar transformers.  I think this rule came from the ARRL.

If there's a disagreement between the two, take the smaller value.

Remember that transformers are rugged beasts.  If you overload one slightly, it will just run a bit hotter.

Ed

#1 isn't very good because larger transformers tend to have better regulation. Try that with a transformer rated to a few kVA and you will overload it!

#2 is much better because it's the core size which determines the power handling capability. Another thing you can do along the same lines is weigh it and compare it to similar sized transformers from various suppliers. Beware that the frequency makes a difference: 50Hz European transformers will be slightly heavier than 60Hz US units.

[edpalmer42]

Two rules that I've used over the years are:

1.  Load the output until the voltage drops by 10% from the no-load value.  I think this rule came from a company that sold electronic surplus.

2.  Measure the cross-sectional area of the core.  Compare that to published data for similar transformers.  I think this rule came from the ARRL.

If there's a disagreement between the two, take the smaller value.

Remember that transformers are rugged beasts.  If you overload one slightly, it will just run a bit hotter.

Ed

#1 isn't very good because larger transformers tend to have better regulation. Try that with a transformer rated to a few kVA and you will overload it!

My rules are intended for transformers like the one in Jester's photo.  Most of us aren't interested in transformers rated for a few kVA.

Ed

[Jester]

Well this transformer is similar in size, and is rated for 5A, 500VA

https://www.digikey.com/products/en?keywords=167P100%20

I load tested my bargain transformer (ac):
 
   0 A = 109 V
   3 A = 107 V
   5 A = 105 V
   6 A = 104 V
   7 A = 102 V
   8 A =  99 V

Load test (dc) [two diode, full-wave with 4700uF]:
 
      0 A = 78 V
   1.5 A = 74 V
   5.0 A = 70 V
   7.0 A = 66 V
   8.0 A = 65 V

Primary resistance is 0.232
Secondary resistance is 0.130


For a 10% regulation, use 10% of the mains voltage, i.e., 12V, then Ohm's law to get Imax = (12V) / (2*(0.232)). That would be 25.9A!

(Note that, because of the above regulation rule, using a transformer at half its rated voltage leaves you with one quarter the VA capacity -- a huge disadvantage.  On the upside, the core runs a whole lot cooler.)


I will use two diodes and perhaps 2000uF, to create a center tapped full wave rectifier (70-77Vdc)

[retrolefty]

Transformer power is a square of the cross-section area of the inner leg of the EI core in cm^2. About 20% more for C core.

 The US ham radio ARRL manual from decades past use to publish a graph of max power handling in watts Vs inner cross-sectional area. Hams use to rewind secondaries to save $$$.

[hlavac]

Is this rule for 50 Hz? 60 Hz? 400 Hz?
Don't think I'll add this one to my tips 'n tricks, but if it works for you...

It is based on a rule for minimal cross section area for given power based on typical peak magnetic flux density in a EI type core of 1T. I don't think frequency is a factor in this one

S = sqrt( P / B )   S is minimum area [cm^2], P is input power [VA], B is maximum flux density [T]

[Benta]

I don't think frequency is a factor in this one

So people building high frequency DC/DC converters to get the size of the magnetics down are on a wild goose chase? I don't think so.

[Tom45]

So people building high frequency DC/DC converters to get the size of the magnetics down are on a wild goose chase? I don't think so.

Also, aircraft using 400 Hz to reduce weight.

[Zero999]

Two rules that I've used over the years are:

1.  Load the output until the voltage drops by 10% from the no-load value.  I think this rule came from a company that sold electronic surplus.

2.  Measure the cross-sectional area of the core.  Compare that to published data for similar transformers.  I think this rule came from the ARRL.

If there's a disagreement between the two, take the smaller value.

Remember that transformers are rugged beasts.  If you overload one slightly, it will just run a bit hotter.

Ed

#1 isn't very good because larger transformers tend to have better regulation. Try that with a transformer rated to a few kVA and you will overload it!

My rules are intended for transformers like the one in Jester's photo.  Most of us aren't interested in transformers rated for a few kVA.

Ed

The first rule is only valid for a very specific sized transformer and is therefore useless, if not dangerous advice. Small transformers have a much poorer regulation factor than 10%, so you'll grossly underestimate the power rating and large transformers have a much better regulation factor than 10%. It also depends on the construction: i.e. the amount of copper, the core and whether it's toroidal or E-core.

See, data sheets below.

For that particular manufactures range of toroidal transformer, only the 120VA units have a regulation factor of 10%, larger units are better and smaller poorer.
https://www.rapidonline.com/pdf/82719.pdf

Here's another data sheet from a different manufacture, again showing 10% regulation for small transformers, much less for larger units.
https://www.rapidonline.com/pdf/525323.pdf

Other then voltage, the only electrical test which are valid for estimating the power rating of transformer, is the winding resistance, as it will give you an idea of the amount of power dissipated and the expected regulation factor, which is also likened to the power rating.

[alm]

Would monitoring the temperature and voltage while applying an increasing load be a useful test? In an overload, I would either expect the windings to overheat, or the voltage to drop due to either IR drop or leakage reactance. So monitoring both should give you a decent estimate of maximum power.

[T3sl4co1l]

Yes, temp rise at ratings should be something like 40C with most designs.

That, too, has exceptions: large, air-cooled transformers run hotter, on account of their power density.  The higher operating temperature draws more convection air.

It'll take some hours to come to thermal equilibrium, too, so it's not as convenient a test...

Tim

[alm]

True, it would be a very slow process, but it would seem like a test that is less reliant on the design choices of the transformer than estimating based on core size or weight.