Is there any way of reducing the surge when powering up a transformer?

Peter. · Feb 7, 2015

I have a 3kva transformer and when I power it up for the first time it sometimes knocks out the trip in my workshop. Other times it knocks out the trip in the consumer unit in my house. I'm well-used to the phenomenon because it happens on site sometimes when I have to plug in my 240/110v transformer into customer's power supplies. Invariably, powering it a second time after re-setting the trip causes no problem.

The one in my workshop however is wired as a step-up autotransformer and is only powered when I use my lathe. It's an incredible annoyance having the whole workshop plunged into darkness because the thump of powering it up has upset the breaker. Is there anyway of dampening the effect, or am I stuck with swapping out the breakers in both panels?

Jim B. · Feb 7, 2015

You could try a "Thermistor". These are devices with a negative coefficient of resistance with temperature.
When cold they have a relative high resistance and then when current heats them up the resistance falls to a low value.

Here are some potentially suitable ones if used in parallel to achieve the necessary constant current.

4pcs NTC 47D 15 Inrush Current Limiter Thermistor NTC 3Amp 47 Ohm CL 80 | eBay

Others may be more suitable.

The real Leigh · Feb 7, 2015

Jim B. said:
You could try a "Thermistor". These are devices with a negative coefficient of resistance with temperature.
When cold they have a relative high resistance and then when current heats them up the resistance falls to a low value.

Current-limiting thermistors work well if the load is constant and predictable within perhaps 25% max delta.
If not, your actual voltage will vary over a wide range since the resistance changes with current.

- Leigh

SBAER · Feb 7, 2015

I don't think a 3kva transformer should draw enough current to plunge the shop into darkness. Is it single phase or three phase (is it on the other side of a rotary phase converter)?

Larry Dickman · Feb 7, 2015

there are systems available which use a battery bank and a power inverter, specifically for this reason. Around here, where you have a money grubbing greedy ass electric company, they bill you a surcharge for you peak usage. If you have a large compressor, for example, this can run into the thousands per month, for the surcharge. With this system in place, it draws temp power from the battery bank and prevents a startup surge from showing up on your meter.

The real Leigh · Feb 7, 2015

Larry Dickman said:
Around here, where you have a money grubbing greedy ass electric company, they bill you a surcharge for you peak usage.

Although I don't doubt for a moment that you have a money grubbing greedy ass electric company...

Realize that to a certain extent peak-usage surcharges have a legitimate justification.
The utility's infrastructure must be designed to handle those peak loads, specifically the peak current, without voltage drop.

So the equipment costs rise as peak current requirements rise, even though those are transient events.

- Leigh

Onepass · Feb 7, 2015

Of course there is an easier fix. When you turn on the transformer it needs to fill up with electricity but you aren't letting enough get into it. Just like a hose that is too small, you get velocity but little volume and that causes the breaker to think too much juice is going out so it trips. It is so simple, you just have to make sure all of the power cords are straight with no links. It is also best to run them down hill if you can. ( that is why the electric poles have the wires so high up)

TDegenhart · Feb 7, 2015

I have a 3KVA transformer feeding a RPC, but the transformer is energized separately from the motor. Occasionally, it will trip the breaker. The problem is that when the transformer is shut off the flux in the core could be near a maximum. If the transformer is high efficiency, there will be little in the way of air gaps in the core to demagnetize the core, plus low winding resistance does little to limit the inrush current. When power is reapplied, if the initial flux is high and of the opposite polarity, there will be a pulse of current needed to reset the core. There are two ways to fix this, one is a circuit the primary that oscillates on power down and gradually demagnetizes the core, much like the ones on mag chucks for surface grinders and second, or to energize the core through a resistance as noted, fixed or variable, and then shorting it out for efficiency. For small transformers, a thermistor and a few other components can be used very effectively. A Google search using the search term "reducing transformer inrush current" will also show other methods.

Controlling the Inrush Current Required by Large Transformers

Tom

TDegenhart · Feb 7, 2015

Onepass said:
Of course there is an easier fix. When you turn on the transformer it needs to fill up with electricity but you aren't letting enough get into it. Just like a hose that is too small, you get velocity but little volume and that causes the breaker to think too much juice is going out so it trips. It is so simple, you just have to make sure all of the power cords are straight with no links. It is also best to run them down hill if you can. ( that is why the electric poles have the wires so high up)

Methinks you are nipping a wee too much.

Tom

Scottl · Feb 7, 2015

A company I worked for had the same problem with a 1kVA transformer in a power supply some years back. I put some thermistors inline with the primary as a temporary fix. We hired a consultant with a reputation for solving such problems for the long term fix.

The issue turned out to be with the type of steel used in the core. The design used grain oriented silicon steel which can retain enough magnetism so that if power is re-applied at nearly the same point in the sine wave as it was discontinued the transformer behaves as if it has no core. At those times the inrush can be massive, we measured 300A at one physical location and nearly 600A at another.

The "stiffer" the line is, ie. the closer to the distribution transformer, the worse the problem will be. Longer wires tend to dampen inrush somewhat. The problem was initially discovered by one of our overseas customers. They were using magnetic circuit breakers with a short time constant. Thermal breakers are much more forgiving since the very short inrush event does not usually raise the temperature enough to trip them.

I would start by checking the breakers and see if they need to be replaced by units more forgiving of inrush currents.

Edit: The major problem with thermistors is if there is a brief (less than 2 cycle) interruption in the supply. Since the thermistor has already reached operating temperature it provides no protection at all in those cases.

Peter. · Feb 7, 2015

Thanks for the replies, I will look into thermistors. So long as I remember to wait a few moments if I ever turn the power off and on again I should be ok.

SBAER said:
I don't think a 3kva transformer should draw enough current to plunge the shop into darkness. Is it single phase or three phase (is it on the other side of a rotary phase converter)?

It's a 3 kva constant, so larger than what you would normally call a 3kva. Physically it's the size of a 5kva.

motion guru · Feb 7, 2015

Peter,

Make sure there is no load on the transformer when you power it up. I had a similar situation where I powered up a 5HP 240V PRC, then powered up a 15kVA 3-phase transformer connected to the RPC to get to 480V 3-phase, then I powered up a 480V VFD . . .

If I tried to power things up with the transformer connected to the RPC - I'd blow the breaker every time . . . if the RPC was started up OK, and then tried to power up the transformer with the VFD turned on - again the breaker would blow. So I had a very repeatable start-up sequence, first the RPC with no load, then the transformer with no load and then the VFD with no load . . . all this was on a 50A breaker.

Mark Rand · Feb 7, 2015

There is a reasonable explanation of the cause here. ignore the last two paragraphs, they're utter bollocks unless you're in the habit of making transformers out of spring steel instead of transformer iron.

Your situation is made worse by the fact that you're using a constant voltage (ferro resonant) transformer. These run partially saturated all the time and can jump a bit when started...

The solution is probably quite simple. If you've got breakers/MCBs from the normal suppliers, they'll be class B, which is good for mostly resistive loads. Find the nearest electrical wholesaler (or go to someone like CPC online) and replace the ones you've got with class D, which are designed with the high inrush current of motors and transformers in mind.

JST · Feb 7, 2015

Thermistors are a bad idea unless you bypass them after they do their work, with a relay/contactor.

At that point they are functioning as a tough resistor. But they have limits. Normally, they are rated for the capacitance and mains voltage. Might not be easy to characterize the transformer inrush sufficiently to match a suitable one to the problem.

Alternately, you can put an ordinary resistor in-line, sized for the current surge, which you can assume is essentially limited by the resistance for the first portion of the cycle..... Power is not as important as current capability.

Bypass it with a relay/contactor, so the resistor is in for a couple cycles and then the contactor closes to short it out. You can size the resistor to limit current to a suitable amount, but I believe with no further info, I would measure the no-load magnetizing current, and size resistor in ohms to allow about 80 to 90 % of mains volts with transformer unloaded. That will be an impedance of about 10 to 20%, and should be good enough to keep surges under control.

You will get two smaller current surges, instead of one near-shorting of the line.

mike_kilroy · Feb 8, 2015

Resistor value can be your voltage divided by your breaker size... or higher r... that stops break trip.

I assume this is 230v? You can put 2 light bulbs in series, wall light switch around this. Power up, when lites dim flip switch.

Chip Chester · Feb 8, 2015

+1 on the motor-tolerant breakers. Or add an electronic switch that acts at zero-crossing on the input waveform, if it's the case of "where in the waveform you shut it down" vs. "where in the waveform you turn it on." Crydom themselves don't recommend it for power factors of < .75 though, but they're cheap enough to try, or to use to control something else with a consistent delta-t(ime) for actuation.

http://www.crydom.com/en/Tech/Newsletters/Solid Statements - SSRs switching types.pdf

Chip

9100 · Feb 8, 2015

Where did this idea that the core retains the flux come from? Core manufacturers go to great lengths to anneal cores to minimize residual magnetism. When the current is turned off, a proper core will fall back to an almost neutral state. The anneal process requires a soak at annealing temperature long enough for the heat to penetrate to the center of the core, expresses in time per inch of depth, then the cool down takes hours, gradually ramping down, often an overnight cycle. After that, they must avoid any plastic deformation, which will work harden the material. One of my customers who made a lot of toroid mag amps and transformers had a standing order that any core that was dropped went straight into the scrap bin.

The problem is applying voltage at a peak instead of starting at zero and letting the magnetism ramp up.

I vote for either a resistor with a relay to short it after a couple of cycles or zero crossing starting. Because the idle current of a transformer is much lower that max rating, the resistor doesn't have to be all that large and the relay won't have to deal with large surges or break high currents, so it may also be small.

Bill

mike_kilroy · Feb 8, 2015

9100;2482171Where did this idea that the core retains the flux come from? The problem is applying voltage at a peak instead of starting at zero and letting the magnetism ramp up. [/QUOTE said:
Ya, it SEEMS logical what you say. I believed it too for my first 20 years of BSEE in industry life!

But it ain't so. Seems illogical that a transformer turned off 3 days ago will retain enough polarity to matter, and that yes, it makes 'sense' that it should be where in the cycle you APPLY power.... But it ain't so.

Weird how some stuff just doesn't seem to follow apparent logic or common sense.

Google it and after a few hours of reading, this seemingly illogical idea will sink it.

It 100% is a factor of where it was turned OFF, not where it was turned on.

Scottl · Feb 8, 2015

I wouldn't have believed it either, until the guy with the PhD from MIT and years of power supply experience proved it to us. More tolerant breakers should solve the problem unless something else is wrong with the circuit. Normally my first advice to people in such situations would be to contact the transformer manufacturer and ask for advice.

mike_kilroy said:
Ya, it SEEMS logical what you say. I believed it too for my first 20 years of BSEE in industry life!

But it ain't so. Seems illogical that a transformer turned off 3 days ago will retain enough polarity to matter, and that yes, it makes 'sense' that it should be where in the cycle you APPLY power.... But it ain't so.

Weird how some stuff just doesn't seem to follow apparent logic or common sense.

Google it and after a few hours of reading, this seemingly illogical idea will sink it.

It 100% is a factor of where it was turned OFF, not where it was turned on.

Scottl · Feb 8, 2015

A lot of companies also deal with individual installation issues. They don't want someone bad-mouthing their product when the issue may be something like improper use or the wrong circuit breakers. I have found many companies quite cooperative even for the one-offs.

thermite said:
If/as/when one is designing a volume-production product .... lots of things to look at.

WHEN one has a smallholder shop with a one-off need? Sodabunchlightbulbs ....and thermistors.

Electric hob/range heaters are meant to live their longish lives directly across the line with nary a cooling fan, and actually heat-TRAPPED under cookware. Wattages range from just under 1K to my 'canning' elements 9K. Barely even get warm seriesed-up for a second or so.

Cheap, cheerful, durable, and a shunt/bypass switch or relay much the same.

KISS. End of problem. Regardless of the 'why' of it.

At sub ten bucks per, I can't even be BOTHERED to do the math.

Got similar ones up in the kitchen to fry the proverbial OTHER fish with.

Bill

Is there any way of reducing the surge when powering up a transformer?

Titanium

Hot Rolled

Diamond

Hot Rolled

Titanium

Diamond

Hot Rolled

Diamond

Diamond

Diamond

Titanium

Diamond

Diamond

Diamond

Stainless

Diamond

Diamond

Stainless

Diamond

Diamond

Similar threads