Isolation transformer inrush current protection

[Simongyork]

Hi, I acquired a 240v primary to 240v secondary (2x70 + 2x50v) at 1.7kva. Toriodal transformer from ebay (from a PA system).

As I am in the U.K. this is great for my needs.

When I use my variac to test it, slowly bringing it up to voltage, it behaves.

But if I leave it for a few days, then just turn it on, the house breaker trips.

If I use a dim bulb tester, the bulb for glows 1/4 bright for a split second, then goes off and the transformer behaves

I know that slow blow fuses are available, but I don’t think that is a safe option.

I think I need some inrush current protection, but I am confused about how to tackle this so that I can safely use this.

There is a calculator online I tried, but I only have the 240v and 1.7kva parts of the equation and am not sure how to work through the rest.

Can anyone advise?

Thanks

Simon

[Ian.M]

Large torodials are notorious for high inrush currents due to core saturation during the first half-cycle.  It depends on where in the cycle you switch on and whether residual magnetism is opposing or aiding the flux buildup.  Without precise phase control of the previous switch-off and the current switch-on, you might as well flip a coin to guess whether or not it will trip for any particular switch-on.

A suitable NTC thermistor designed for inrush surge prevention, and rated for >7.5A continuous current would do the job.  You need to know the no-load magnetising current so you can choose an appropriate one that will stay low resistance when the secondary is unloaded.  It will run hot, so its placement with respect to surrounding parts should allow for that.  However it offers no protection if you switch off and quickly switch back on without waiting for it to cool down and also increases losses and gives the load side a higher source impedance.

If yiu want to get fancier, put a time delay relay across the thermistor to short it out after switch-on so it cools down ASAP.  The ultimate is to add another time delay relay so the secondary side load is only connected after the thermistor has been bypassed.

[exe]

You need a soft-start circuit (usually based on a resistor, bypassing relay and delay timer). There are many of them, you can google, or just buy a ready one. I'm not sure about NTC at such loads... But I didn't do any calculations.

[Ian.M]

The trouble with a single relay + power resistor soft start is its difficult to handle high initial secondary load currents.   If you aren't using a NTC thermistor, for a 1.7KVA beast like this you'll most likely need that second (load-side) relay.

[rstofer]

https://en.wikipedia.org/wiki/Inrush_current

The inrush of a standard transformer may be 10 to 15 times full current.  For a torroidal, it is up to 60 times full rated current.  Your 1.7 kVA transformer has a full load current of 7A at 240V so the maximum inrush COULD be as high as 420A.  In the US, a standard molded case breaker has an instantaneous trip at 10x rating so we could conceivably trip a 40A breaker so we would need to think in terms of a 50A breaker for a 7A transformer.  Amazing...

Note that US breakers trip on both thermal (long time) and instantaneous currents.  The instantaneous trip unit is a simple solenoid.  The current goes through the winding and, if it creates a large enough magnetic field in the pole piece, the breaker trips - instantaneously.

The maximum inrush occurs when the transformer is energized at zero crossing (the slope of the voltage curve is steepest (dV/dt).  Maybe you can come up with a sophisticated way to use triacs to turn on much later in the cycle, somewhere around 90 degrees from zero crossing.

I think you're going to have a hard time finding a suitable protective device with a full load rating anywhere near 10A.

One way or another, you're going to have to control the inrush current.

http://www.instructables.com/id/Controlling-the-Inrush-Current-Required-by-Large-T/

Google for 'how to control inrush current for torroidal transformer'

[Simongyork]

Thanks for this. It is like flipping a coin as sometimes it causes no issues at all, so I can see that your explanation neatly fits the scenario I am facing.

Might be a great video post for Dave Jones to cover; inrush current calculations and protection.

So using the variac is a bad idea, given the current being pulled?

[Ian.M]

Do you have an ammeter on your Varac?  (If not, its worth adding one in series with the wiper, preferably a small analog meter with a redline marked on the scale at the Variac's rated current.  A small DC 1mA panel meter, scaled '0' to '10' with an appropriate shunt resistor, and a preset to trim it in series with the meter coil, connected between + and - of a high current bridge rectifier will do nicely after calibration against a DMM)

If you ramp up the voltage slowly enough to stay under the Variac's maximum current rating (which does *NOT* increase when its outputting a lower voltage), no problem - but the maximum current you can draw from the secondary will be limited by the Variac's current rating if its under 7.5A.   Also it ties up your Variac, which may well be needed for soak-testing some other project at high or low line voltage.

[Benta]

I can confirm the experience that you and others have with large transformers. In my case it was a 1 kVA isolation transformer that kept tripping the breakers (13 A). After two replacements (I thought the transformer was faulty) I began to see the light.

Lessons learned: 500 VA is just about the largest toroid transformer you can connect directly to a 230 V, 13 A circuit. 800 VA EI types also work. For anything above that, you need starter circuitry.

I played with the idea of a Triac starter that would turn on the transformer at peak voltage, but never got around to it, as I'd found a 500 VA type in the meantime.

[Benta]

One way or another, you're going to have to control the inrush current.

http://www.instructables.com/id/Controlling-the-Inrush-Current-Required-by-Large-T/

I now remember why I put that "Instructables" website on my no-go list (quoted from the link above):

Large transformers have a huge current demand when they are initially turned on. This is because, until the magnetic field and inductive resistance builds, they are essentially short circuits.

Complete gobbledygook...

The problem is asymmetric flux, which drives the transformer into saturation on the second (and following) half-cycles, resulting in immense peak currents.

[Brumby]

I had a big transformer some years ago that had that problem.  I fixed it by putting a 250W incandescent light in series with the primary - and a switch to short out the light.  With the switch open, apply power and let the light regulate the current for a second or two, then close the switch to apply full power to the transformer.

Inelegant for sure, but it was enough to get around the problem.  I didn't get any fancier because it was only a short term need.

I did have a more involved solution in case more control was needed ... it consisted of a number of extra lights that could be switched on in sequence, paralleling the original globe.  Apply power and one light is lit; switch in a second light to allow more current .... then a third and so on.  Finally when all the lights were lit, short them out to allow full power to the transformer.

I never had to go there - besides, I didn't have the goggles, rubber gloves, lab coat nor a maniacal laugh that would just have to accompany such a procedure!

[woodchips]

Why not just replace the MCB in the circuit powering the transformer with a D rated one, that is designed especially for transformers?

[Ian.M]

The O.P's in the UK, so are you, so you should know better.
The 'nanny state' regulations (Part P of the Building Regulations) make it prohibitively expensive to do a simple breaker swap even *IF *a D rated breaker is available, approved for use in the O.P's consumer unit.

[IanMacdonald]

The O.P's in the UK, so are you, so you should know better.
The 'nanny state' regulations (Part P of the Building Regulations) make it prohibitively expensive to do a simple breaker swap even *IF *a D rated breaker is available, approved for use in the O.P's consumer unit.

Actually, more racketeering than nanny state. Designed to restrict work to the large trade cartels, and make life difficult for the small contractor. Thankfully doesn't apply to Scotland.

The inrush problem arises, paradoxically, if the power is turned on at or near the zero voltage point. This is because the core has to take up a whole half cycle worth of energy in one direction, before the flux starts to reduce again. If the power is turned on at peak voltage then the core only has to absorb half as much before the zero crossing is reached and the flux starts to reduce.

[exe]

The inrush problem arises, paradoxically, if the power is turned on at or near the zero voltage point. This is because the core has to take up a whole half cycle worth of energy in one direction, before the flux starts to reduce again. If the power is turned on at peak voltage then the core only has to absorb half as much before the zero crossing is reached and the flux starts to reduce.

I think there is (also) another effect. When turning at peak voltage, inductor tries to "prevent"  sudden voltage change (change from 0 to Vpeak). So it "smooths" it, big reactive resistance (or whatever it is called).

[rstofer]

About 6 decades back, I used to make arc lights using the carbon rods from batteries.  I would put a flat iron in series to limit the current.  Interesting project for a 6th grader...

So, let the flat iron absorb the inrush and put a switch in parallel across the iron to get to full power.  This thing just screams for a 3 position switch: OFF - START - RUN or OFF - ON - ON

They actually make such a switch and Mouser has them in stock rated 15A 250V:
http://www.mouser.com/ProductDetail/Carling-Technologies/2GG51-73/

[Kleinstein]

Such a large transformer should have an inrush current limiter of some kind just a slower fuse is not an option, as the peak current might well blow the next level fuse and cause nasty transients in other circuits. How bad it is depends on the size and type of transformer. So there is no fixed number for the ratio inrush current to nominal current. This tends to get larger the larger the transformer.

It depends on the load, but often the slow start also help a load on the secondary side to start soft. So you might want the resistor for the soft start to be a little more powerful than just starting the transformer. The resistor can be rather low value - it does not take that much to limit the peak current to something in the 20 A range, so the fuse can stand it for a short pulse.

Starting at peak voltage is better than nothing, but for some transformers this might not be enough - they can still trip a fuse or burn the triac if residual magnetization is too large.

[edpalmer42]

This topic has been discussed a few times on this forum.  Search for <transformer inrush> or <transformer surge>.

Ed

[Zero999]

https://www.digikey.com/product-detail/en/ametherm/SL22-10008/570-1034-ND/749874