Inrush current in a transformer

in article S_wIb.353$ snipped-for-privacy@news.getnet.dk, Cubus at snipped-for-privacy@sol.dk wrote on

12/31/03 1:50 AM:

The data-label does not provide sufficient information to answer your question.

The inrush current depends, in part on the leakage and magnetizing inductances of your transformer as well as the load. The exact closure phase can be extremely important. Resonant charging can cause peak currents much higher than you might otherwise expect. Core saturation can also contribute.

In short, you need a fairly accurate equivalent circuit for you configuration, including the nonlinearitie.

Bill

OK, but it was n ot a bad question. In thepast (like maybe up to about

5-10 yearrs ago), youcould go to the IEEE and get a typical value for Inrust current. Now, in this era of high efficiency" transformers, I;'m told theat some of them have an inrush of 10 to 14 times nominal nameplate amps. (Same thing for the newer motors.)

So what do people do when they are trying to perform studies that will give them an idea of how their system is supposed to respond to the new loads - so they know how much capacity they need to add to the system? Waiting for the actuals from the mfr. seems to be the wrong way to go - and you need to know WELL in advance of receiving the transformer or motor, as the lead time on system components is long, too.

So what rules of thumb are we supposed to use now?

Thanks, folkjs.

HR.

Er, if you have the transformer and Siemens can't tell you, can you measure it?

--s falke

"Repeating Rifle" skrev

What im actually looking for, is some rules of thumb, not the exact calculated value.

Purpose: to be able to estimate what size of mini-circuit breaker would be nessesary on the primary side without tripping when the transformer is powered on.

A smaller transformer was replaced with this one and a 10 A type C mini-cicuit breaker tripped almost everytime power was turned on.

The maximum inrush current would be determined by the inductance of the coil with an air core. The inrush current depends upon a number of things and so can not be computed.

1 Shape of the Hysterieses loop. 2 Point on the hysterieses loop when the transformer was last deenergized. 3 Available margin of spare volt seconds designed into the core over the required. 4 The applied voltage. 5 The series impedance of the coil and the source.

So you see, it is basically impossible to compute what the inrush current will be at turn on. It depends on to many things that have happened in the past and to measure those things will change them.

Er your wrong.. The inrush current can be vastly different each time you energize the transformer and never ever repeat itself.

Not quite correct. I have set up a circuit in the laboratory, where one can get the same inrush current time after time. However you do need :-

(a) Constant Supply voltage

(b) A load on the transformer significantly greater than the magnetising current

(c) A circuit-breaker that breaks accurately at current zero.

(d) Point on wave control of the closing circuit breaker.

With a set up like this you can plot the distribution of the magnetising inrush peak with closing angle.

Items (b) & (c) together will guarantee that you have the same remanent flux each time, although for a single phase transformer it could be in either of two directions.

John

You can get an upper limit by measuring the DC resistance of the primary winding. (Divide the peak input voltage by that value). This works because that's all that's left when the core saturates. Toroidal power transformers seem to be particularly bad.

Best regards, Spehro Pefhany

Yes with the setup below, you could measure the inrush current and get repeatable readings almost.

But you need a bit more than you outline. Your lab setup must always turn the power off at the same phase zero crossing and must also turn it on at the same phase zero voltage crossing.

Turning the power off at a Zero current crossing does not insure repeatable power measurements. This only helps if you have a large gap in the core.

If you have a square loop material then 0 current crossing occurs at or near max flux density. If the power is turned on with the voltage trying to drive the flux further in this direction there will be saturation and a large current flow. If you turn the power off at 0 flux you will still get a large surge if the power is turned on at 0 voltage in either direction as the core will have to go from 0 flux to max flux (with the core saturated) as it will have to hold twice the volt seconds that a well designed transformer is designed to withstand.

To avoid a surge, you would have to turn to power off at max flux and then apply a voltage that drives the flux in opposite direction. There would then be no saturation as the core could absorb the full VS's going in the opposet direction.

In any case load current plays on part in surge currents.

Your correct Spehro but you must also include the line impedance. Torriodal cores are generally worse because they are, in general are nade from square loop materials or at least squarer than the material used in E-I cores like Selectron. Also E-I cores in general have some effective gap which brings the flux down from saturation closer to 0 flux so when the power is turned on there is at least half a BH loop to climb.

Embedded software/hardware/analog Info for designers:

From the label, it looks like about 3KVA capacity transformer (230VAC,

14A output when 2ndary windings are connected in series). What voltage is being applied to the primary side (and across which pins)? What kind of load is connected to the transformer? The more details you can provide, the more quickly we might be able to arrive at a possible explanation (& solution).

Yes, we know this. But one of the basics of engineering is to provide a conservative (i.e., worst case) estimate and base all calcs on this value.

You see, that way your system will usually be in better shape than you calculated, and your system will never fail due to a higher value of inrust current! (Relays won't operate unexpectedly; etc.)

It would be more helpful if you would try to supply information of assistance.

HR.

OK, I guess I should have said "If you cannot help, you don't have to kill any electrons."

A collection of motherhood statements which appear to sound impressive but mean nothing is of no help to either the original poster, or to me.

HR.

Umm, he did help. He said that it cannot be predicted.

Huh? Being a microprocessor hardware type, I hadn't considered that this wasn't predictable. I learned something here, even if you were unable to grasp the English. I suppose this is the reason "inrush limiters" are used in the very large power supplies I've come across (and the cause of much angst in a previous life).

in article snipped-for-privacy@news.telus.net, Rowbotth at snipped-for-privacy@telusplanet.net wrote on 1/1/04 3:17 PM:

How do you kill an electron without using positron bullets?

Bill

BMFH?????

A highly tecnical question was asked. There is no simple way of telling whqt the inrush current will be. It is not my fault if you can not understand the answere. Perhaps I should have answered you The grass is green because. Those mother hood statements are not impressive. They are the fact. Perhaps I should have started you with basic electronics 101.

---------------- This is a power transformer- it will definitely not have a square loop BH curve as the material used should have as small a loop area as possible. Core gaps in such a transformer are not deliberate (and unwanted) and the effective gap in the EI core will have little effect. Some transformers use a toroid made of a continuous strip of transformer steel to eliminat such gaps- I do not know whether any significant improvement was achieved by this construction. - however speculation on the core construction is simply that- speculation.

The case that Spehro gave is a good upper limit estimate.

The reason for the dual kVA rating is not apparent. Air cooled vs. oil cooled, possible autotransformer operation - insufficient information.

-- Don Kelly snipped-for-privacy@peeshaw.ca remove the urine to answer

As I said power transformers for 60 hertz are generally made of 17 mil selectron or some such material

I realize that EI, DU and all the other types the desire is to keep the gap as small as possible but it is always there and it does tend to drop the residual flux down to a very low value.

Some transformers use

I have never heard of a power transformer made with a torroid. Power transformers in general have heavier wire sizes and they are hard to wind on a torroid. I used to work for Arnold who made torroid winders.

When I designed I generally used some sort of nickle core which are square looped. Some times for converters you want softer curves so the saturation currents are not so high.

But I had developed teqniques to use very square loop and was able to prevent the flux from creeping up the loop on one side of the other.

In fact on one mag amp controlled converter for one satelite we went all through tests and passed every test and then I knowticed that the wave did not have dead zones on both sides pulses.

We checked and found that there was an open connection and half the converter was not working. The flyback on the current drive was providing the drive to the open side. Because the transformer had base emmiter and external return diodes, both sides of the wave shape were symetrical.

Mag amps work out nice in converters as they are Volt-Second devices and you can get open loop response to the input changes which cancel out input variations. We used a lower gain amplifier with 12db role off and had an inherrently stable circuit.

Been in orbit and still working now after about 35 years.