# Can a step-down transformer be used as a step-up transformer?

#### MJHanagan

Apr 19, 2012
7
Hi everyone - I need to produce a high voltage to test the dielectic breakdown of a material sample. My power supply only goes up to 300V but I need to apply a voltage of at least 1000V. I would like to know if I can use a common step-down transformer as a step-up transformer by simply swapping the line in and out wires. I have a low-power 120V in/12V out transformer that has a top voltage rating on the insulation of 2500V - would this be suitable to achieve 1000V?

#### shrtrnd

Jan 15, 2010
3,826
Your theory is correct, but if you're saying you intend to apply 300VAC to the output of a 12VAC transformer, my suggestion is: don't even think about it.
Whatever insulation you're referring to, is not going to withstand what's going to happen to your 12V wiring or the windings on your transformer.

#### MJHanagan

Apr 19, 2012
7
My plan is to only apply enough voltage on the secondary windings to generate 1000V on the primary side (i.e. about 100V). I don't want to violate the transformer's insulation rating of 2500V - I think that's the limitation trying this whacky method.

#### (*steve*)

##### ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,505
The 2.5kV rating of the transformer is between windings (presumably)

If you apply a high voltage to the secondary and get an even higher voltage across the primary you may cause discharge within the primary because the voltage difference between adjacent layers of windings exceeds the capacity of the wire's insulation.

Remember that you're contemplating a ten-fold increase in voltage which may well exceed the limits the original designers engineered in.

It could also be that the manufacturing method and materials exceed the requirements and that they will withstand the 1000V. That would be fortunate. Just don't bet on it.

#### duke37

Jan 9, 2011
5,364
Do you need DC or AC?

You will need negligible current to test for insulation breakdown so you could use a voltage multiplier, this is two stacks of capacitors with diodes connected between them.

You could make a high frequency inverter with insulation between each layer in the transformer.

You could use an electric fence energiser which gives out pulses of about 5kV.

You could buy a Megger which is made to do this job.

Apr 21, 2011
129
Surely putting 100V across the 12V winding will cause so much current to flow, that the 12V winding will burn out very quickly.

Or if the current is limited to what the 12V winding can withstand, there will be a lot less than 100V across the 12V winding.

Unless you plan to use PWM to limit the current.

The idea sounds flawed to me.

#### (*steve*)

##### ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,505
Surely putting 100V across the 12V winding will cause so much current to flow, that the 12V winding will burn out very quickly.

Not necessarily.

Apr 21, 2011
129
Not necessarily.
I am confused. Assuming that the OP was planning to put 120V mains (or is 100V?) across the 12V winding, AC at 50/60 Hz, and assuming that the transformer is a typical power-supply type transformer, I would have thought that burn-out is pretty much inevitable after a matter of a few seconds, or less.

The mains winding is designed to handle mains voltage without drawing excessive no-load current, the 12V side is not, as it will have fewer turns and probably thicker wire.

Are my assumptions wrong? Or is my understanding of a typical 12V transformer lacking?

EDIT: Ah, I get it.... maybe the fuse will blow first, so the winding might not burn out. Still no use to the OP though

Last edited:

#### jackorocko

Apr 4, 2010
1,284
unless the insulation on the wire breaks down there shouldn't be a problem unless you exceed the current rating. That surely depends on the VA rating of the transformer and the load attached to it.

Apr 21, 2011
129
I am clearly not going to convince anyone here.

All I ask is that when you try it, you have a fire extinguisher to hand, and have somebody standing next to the mains breaker.

#### BobK

Jan 5, 2010
7,682

With 1/10'th the turns, the secondary should draw about 10 times the current that the primary would when connected to line voltage. However, it also should be able to handle 10 times the current. No?

Edit: No i am wrong, the inductance is proportional to the square of the number of turns! Which means it would draw 100 times the current that the primary would.

Bob

Last edited:

Apr 21, 2011
129
My understanding of theory is almost nil, but I am a very practical person, and have messed around with electronics a bit.

My simplistic half-understanding is this:

Feed mains into a lot of turns with a lot of inductance, and both the wire resistance and the inductance will resist current flow well.

Feed mains into a few turns with less inductance and there will be far less resistance to current flow. You now have a big voltage and a big current in a short length of wire. This will create heat, especially if you draw little or no load from the other winding.

I have little clue how the maths works, and I could be completely wrong, but I sense that the 12V transformer is going to get very hot.

Imagine one of those instant-heat solder guns. It has many turns on the primary, and only a turn or two of really thick wire as a secondary. Imagine connecting mains to just one or two turns of thick wire - the fuse won't stand a chance.

I suspect a similar kind of process will operate for the OP's 12v transformer, but less dramatically.

I am open to being proved wrong, but until I am, I will not be experimenting with the mains in this way.

Last edited:

#### duke37

Jan 9, 2011
5,364
A mains transformer will not be greatly over designed and so 12V on its output will be taking the core close to saturation. Putting a larger voltage in will give core saturation and current will be limited by the wire resistance - Bang.

More voltage can be used if the frequency is raised but you are into a design project here. There will be severe insulation problems if the voltage is raised significantly.

Apr 21, 2011
129
A mains transformer will not be greatly over designed and so 12V on its output will be taking the core close to saturation. Putting a larger voltage in will give core saturation and current will be limited by the wire resistance - Bang.

More voltage can be used if the frequency is raised but you are into a design project here. There will be severe insulation problems if the voltage is raised significantly.

@duke37

Thank you

#### (*steve*)

##### ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,505
A mains transformer will not be greatly over designed and so 12V on its output will be taking the core close to saturation. Putting a larger voltage in will give core saturation and current will be limited by the wire resistance - Bang.

More voltage can be used if the frequency is raised but you are into a design project here. There will be severe insulation problems if the voltage is raised significantly.

The key is that you may need to increase the frequency.

The inductance of the secondary may indeed be sufficient to prevent peak currents from reaching a dangerous level, however it may not leave much margin to allow an increase as you draw power from the other winding.

The problems are also that the core may not be as useful at a higher frequency, and (as you say) the insulation becomes an even larger issue

It's not necessarily going to work, but it's not as straightforward as hexreader thinks.

#### MJHanagan

Apr 19, 2012
7
My intention is to apply this nominal 1000V to and insulator material. Thus where will be no net power drawn from the primary windings (more or less just an open circuit). My understanding is there will be some net current in the secondary windings but this curent will setup an opposing magnetic field in the core which will prevent it from "short circuiting".

I will be powering the secondary side using a voltage supply wherin I can limit the current output to something like 0.1 A. Theoretically if I toast the transformer the power supply will shutoff the output thus limiting the damage to small puff of smoke and a toasted transformer (this is a wee small tansformer).

I also want to limit the current on the primary side bacause when the "insulator" under test does breakdown I don't want to burn it up. In fact if this does work I will also install a resistor in series so when current starts to flow above a certain level it will trigger off the power supply unit.

I have been looking for a very small capacity (power) 10X step up transformer but have yet to find one. Something in the nominal 1-5 Watt range and 1000V output and <$100 would be nice. #### MJHanagan Apr 19, 2012 7 Because of the type of material being tested it needs to be done with an AC voltage. #### MJHanagan Apr 19, 2012 7 Ah ha! I think I finally found a low-power step-up transformer. Seems to be used for CCFL's. Here is the spec sheet: http://www.cooperindustries.com/con... Sheets/BUS_Elx_PM_4303_CCFL_Transformers.pdf Seems like it they have a maximum output voltage of 1350V RMS which is plenty and with only a nominal 20 V input. The output power ranges from 2.5 to 14 Watts so if the sample avalanches the damage can be limited this preserving the sample for a possible recovery and retest (one of the main goals of this test). Best part is they cost <$20! Perfect.

Unless I'm missing something obvious in the datasheet this device seems ideal.

#### BobK

Jan 5, 2010
7,682
I hope you noticed this line from the datasheet.

• Frequency range from 40 to 80 KHz

Don't go putting 60Hz AC on it!

Bob

Replies
30
Views
2K
Replies
7
Views
667
Replies
3
Views
2K
Replies
13
Views
1K
Replies
2
Views
915