208V/480V Transformer In Rush Current?

My shop has 208V 200A 3PH service panel. I’ve got several 480V machines that I supply through the following transformer:


I feed the 208V side of the transformer through a 100A breaker in the main panel. I leave that breaker off most of the time, turn it on about once a week to run a 480V machine. When I turn that breaker on it usually kicks right back out several times before it stays on. Occasionally the 200A main will trip. Once “on” there is never a problem.

Why should there be this huge in rush of current? Aren’t inductors (transformers) supposed to resist changes in current? That’s why they call them chokes?

What can I do avoid this? Would it be better to feed the transformer through a disconnect and leave the breaker on? Why would that matter? I need some kind of resistor to limit the current which switches out of the circuit after a few seconds. Is there such a thing?


Sent from my iPhone using Tapatalk

The larger the transformer the bigger this peoblem is, takes a couple minutes for the dc bias to work its way out of a gigawatt sized transformer. Takes my 2kw torodal variac about 3 seconds.


Simple way to fix the problem is add a second relay or momentary push button switch, to power up the transformer through a resistor, for your unit, 3 stove top heating elements will work fine. Push the button wait 2 seconds and throw the main disconnect.

There is a wikipedia page on this. Inrush current - Wikipedia

How many amps are you actually pulling through this transformer under load? I'd personally use some zero crossing SSRs on there. Simple control, and avoids the magnetizing inrush which occurs on sudden voltage transitions.


Basically transformers aren't perfect, and their own core acts as a winding of sorts. It's too poorly "linked" to the primary to respond to the smooth 60hz sine wave, but if you turn the breaker on while mains power is near it's peak (which is guaranteed with 3 phase) at least one leg will draw roughly 5 times FLA. (Average value that I find in my own projects. It varies). That ~1000A half cycle is enough to trip the short circuit detection on the breaker, which trips based on the magnetic field in a coil in the breaker and not on the temperature of the heater.

The resistor approach is also valid, but use contactors. I used stove coils on my RPC, but mounting and connecting them is inconvenient. The cost of each is similar depending on what you can scrounge.

In any case I advocate not using the breaker to routinely turn the transformer on and off, as I've had plenty of them fail on me in that sort of service.


If you choose to got the SSR approach, you will need a small DC 12 or 24 volt power supply (very cheap) a switch, and three SSRs rated to a bit over the current you intend to pull through the transformer.

The first cycle will still draw a bit more power due to magnetics, but not too much. There are advantages and drawbacks to both systems that I don't think really apply here, but I'm glad to guide you through building either system.

It is not a problem. It is physics. It cannot be avoided. It is normal. It should be compensated for through design. A very easy and inexpensive solution is a simple light bulb in series with the primary winding and a shunt switch parallel to the bulb.The use is then a start position, with the bulb in circuit, then a run position with the bulb out of the circuit. Simple solution.

steve-l said:

It is not a problem. It is physics. It cannot be avoided. It is normal. It should be compensated for through design. A very easy and inexpensive solution is a simple light bulb in series with the primary winding and a shunt switch parallel to the bulb.The use is then a start position, with the bulb in circuit, then a run position with the bulb out of the circuit. Simple solution.

Click to expand...

"It is not a problem" *offers solution*

Would a "soft start" device work in this application? From the description of the device it seems like it would, but I only see them discussed in regards to a motor, not a transformer.

adh2000 said:

Would a "soft start" device work in this application? From the description of the device it seems like it would, but I only see them discussed in regards to a motor, not a transformer.

Click to expand...

Yes, but it depends on the particular details of the soft starter.

Strostkovy said:

Yes, but it depends on the particular details of the soft starter.

Click to expand...

Care to make a specific suggestion?

adh2000 said:

Care to make a specific suggestion?

Click to expand...

Primary resistor soft starter.


They are simply one resistor in series with each phase, and a timed contactor to bypass them after a short delay.

Strostkovy said:

Primary resistor soft starter.

Click to expand...

Yes, that's what I want. This is helpful, thanks. But I'm having trouble converting this to a product or part number. My transformer is 45kVA, so I want a 60HP soft starter? Sound right? Is it reasonable to feed this transformer from a 200 amp panel? There's lighting, dehumidifier, etc on the panel as well. It's what I've got so hope to make this work.

adh2000 said:

Yes, that's what I want. This is helpful, thanks. But I'm having trouble converting this to a product or part number. My transformer is 45kVA, so I want a 60HP soft starter? Sound right? Is it reasonable to feed this transformer from a 200 amp panel? There's lighting, dehumidifier, etc on the panel as well. It's what I've got so hope to make this work.

Click to expand...

Well, you are feeding it through a 100 amp breaker, so I would use a soft starter rated to ~36kva, or about 50HP.

As far as part numbers go, I'm not sure. I'd build something like this instead of buy it, so I have no idea where to buy a complete unit.



I have a better idea of how to build something that would do what you need, easier and cheaper, and be more convenient for your application. You will need:
A 3 position switch
A 50 amp contactor
Strips of steel (possibly stainless) to make resistors
A 100 amp contactor

The first position of the switch is off.
The middle position just turns on the contactor with the resistor.
The final position turns on both contactors.

The idea being you can just turn the switch on and off, and the soft start action happens inherently as you turn the switch. No timers, no fancy controls. I can draw up a schematic for you and help you source parts for this approach if you would like. It also has the advantage of a smaller initial surge, as you aren't limited in resistance by a hypothetical motor load.

Are there code issues with wiring it directly to incoming power and using slow blow fuses for wire protection?

Hillside Fab said:

Are there code issues with wiring it directly to incoming power and using slow blow fuses for wire protection?

Click to expand...

"Code" is primarily to reduce risk of fire or electrocution. Minimalist approach of using "approved" goods in one of the many "approved" manners.

It carries a built-in "fiddle" to keep itself OUT of the nitty-gritty of technological "progress". Or lack, thereof.

Simply put all your experi mental-case s**t inside the enclosure of the load "system". Oversized JEDEC cabinet, even a wire-cage ROOM if you must...

Then, so long as wire TO it is sized and protected at or above the "rated" requirement? Code closes one eye to what goes-on inside your "system".

A Fire Marshall or an Inchoorance claims adjuster who has to investigate what TF it was as kilt you .. and/or burned the whole city block to the ground ...is "somewhat less cooperative".

So it still "pays-well" to NOT f**k it up!

Part of the reason folks recommend stove heating elements or light-bulbs, not hand-cut strips of steel nor Iron fence wire.

Hillside Fab said:

Are there code issues with wiring it directly to incoming power and using slow blow fuses for wire protection?

Click to expand...

Yes, whatever wire you run to your fuses must be protected by an upstream current protection.

*Maybe* you could get away with using the same gauge wire as your service entry and tying directly into where it feeds your breaker, but I doubt it. I've never personally daisy chained feeder connections but I feel like there are a lot of weird gotchas, such as now having two service entry panels, doubling wires in terminals, etc.

thermite said:

Part of the reason folks recommend stove heating elements or light-bulbs, not hand-cut strips of steel nor Iron fence wire.

Click to expand...

To be fair, if a resistor gets hot it's just doing it's job.

A high wattage lightbulb in a socket is probably the easiest, plus you can screw in different things. The flash at contact might be fun too. I was of course envisioning a cut shape much like a current sensing shunt, but it is likely not wise to assume everyone will take the same amount of care in its construction.

My brain is still a little stuck on my RPC fiasco so I've been thinking in terms of motors, not stocky inductors. (Now that I understand how fire pump motors are used, it was dumb of me to think a "pulled from service" motor wouldn't be fried).

Strostkovy said:

To be fair, if a resistor gets hot it's just doing it's job.

Click to expand...

"To be fair.." if the bare metal strip electrocutes you instead of burning your hand?

It's "just doing its job".

it was dumb of me to think a "pulled from service" motor wouldn't be fried).

Click to expand...

Conversely, I EXPECT the "NOS" Reliance Duty Masters I've been buying to HAVE "cosmetic damage". Returns or write-offs from damage in-transit from the look of them, s**t lousy OEM packing crate (that means YOU, Reliance...) and worse palletizing (same, again, OEM) the trigger.

As-in cracked CI fan housing, bent sheet metal fan housing, busted-off Zerk fittings, missing shaft keys, bent foot-flanges, missing peckerhead covers - or the whole peckerhead.. rusted face mounts from careless storage...

ALL of which are cheap and easy to remedy.. and I get a motor that is virgin-clean, virgin wires, unused bearings, and in general electrically perfect.

For about $250, plus freight. Mind, sometimes freight costs more than the motor, and never very much LESS.

I did say "Reliance Duty Master?"

Heavy f**kers. Most weigh more than I do. I'm good with both sides of that!

thermite said:

"To be fair.." if the bare metal strip electrocutes you instead of burning your hand?

It's "just doing its job".



Conversely, I EXPECT the "NOS" Reliance Duty Masters I've been buying to HAVE "cosmetic damage". Returns or write-oof from damage in-transit from the look of them, s**t lousy packing crate and palletizng the trigger.

As-in cracked CI fan housing, bent sheet metal fan housing, busted-off Zerk fittings, missing shaft keys, bent foot-flanges, missing peckerhead covers - or the whole peckerhead.. rusted face mounts from careless storage...

ALL of which are cheap and easy to remedy.. and I get a motor that is virgin-clean, virgin wires, unused bearings, and in general electrically perfect.

For about $250, plus freight.

Click to expand...

Well it's connected to all of the other live stuff. No matter what you do here it needs to be in an enclosure. There isn't enough energy dissipated to warrant venting. You can get just as hurt with the rest of the contactors and terminal blocks.

Though to be clear, light bulbs are continuous duty and fool resistant. I highly suggest the lightbulb route instead of resistors. Having dealt with stove elements, they are a nightmare to mount and their connectors suck. I would not suggest them. Also, lightbulbs only require a contactor around 10 amps or whatever.


The smashed fan cover is a guarantee anytime I buy a motor. The particular motor I bought was 40hp pulled from service through surplus center. I trusted surplus center so much from past experiences that I didn't even consider checking the windings because I thought for sure anything not working was my doing. I should have known the squealing bearings were an indication of prior overheating, and didn't even consider the paint flaking off of the primer was also related to that. Speculating, I think it was a motor pulled off of a pump that went bad, and the bad pump killed the motor too. I wasted a lot of time on that. But, now I have a rotary phase converter capable of dual idlers *and* capable of starting and synchronizing a synchronous motor. In a building I rarely visit, producing three phase I no longer need. Anyway, that's not related to this thread and I've probably hijacked enough of them.

Strostkovy said:

I didn't even consider checking the windings because I thought for sure anything not working was my doing. I should have known the squealing bearings were an indication of prior overheating, and didn't even consider the paint flaking off of the primer was also related to that.

Click to expand...

Well "pardon me".. but that's f****g comical for "trust"!

I'll presume you are handicapped by having no sense of smell, either?

Half a century or so @ five packs of smokes a day, yet I can still detect polymer fume or phenolic char well-enough to avoid getting "skunked" on electrical goods in general.

Quick and dirty schematic. Symbols aren't standard and there are diagonal lines, but it gets the point across. The two images are the same. One is on imgur to display better, the other is hosted by the forum and is smaller but will be preserved for posterity.

I'll offer a few thoughts, contrary to the others given here, and what you speculated in your first post.

Your transformer is too big for the size service you have, using breakers instead of time delay fuses also works against you. Evidence of this fact is that your main breaker trips on inrush current.

Further using a transformer in reverse is also working against you in two ways. First is with a delta winding on the output side, you only have two choices in ways to use it. Ungrounded or corner grounded, neither of which is good or desirable.

Second is that the inrush current is magnified when energizing it in reverse of it's designed connection.
The standard way transformers are wound places the input winding closest to the core, when used in reverse the input winding is further from the core which increases the inrush current spike, and extends the time period of the surge current.

Transformer cores retain residual magnetism when turned off, no matter how long they sit. This residual magnetism has a relation to the sine wave powering it when disconnected. When re-energizing the transformer, if the sine wave is at a different position in time, that when it was turned off, that increases the surge current to overcome the difference in timing.

For standard transformers the normal inrush current is 6 to 10 times the full load current, for the first couple cycles till things stabilize, Essentially it's akin to a short circuit when first energized, similar to a motor but with a larger surge current than a typical motor winding. Oversized and used in reverse just complicates it further.

The NEC stipulates that the over-current devices feeding a transformer be 125 to 175% of its full load current, to prevent nuisance tripping on start-up and still protect the windings and supply conductors. It also requires the same level of over-current on the output side of the transformer.

Your 45KVA unit is rated at full load current of 54A @480 and 125A @ 208. To prevent nuisance tripping for a 45KVA @208V you would require a 188A supply circuit minimum (125Ax1.5). You would require a fused disconnect with fuses sized at the full 200A of your service for reliable operation.

A better solution would be to check your equipment loads that are run at the same time and figure out what is the smallest transformer size you need to get that done, and the do the math to see if your existing 100A breaker can handle that load. If not the you would need to do something different with your service panel, since many 200A 3Φ panels will only accept a 100A breaker.

A 25KVA transformer of the correct type 208 delta input with a 480Y output would be about the largest size you could reliably start off of a 100A breaker. That would provide you 30A @ 480V to supply your equipment. And it could be properly grounded for you and your equipment's safety.

Side note, a breaker that has been tripped many times will become out of calibration, and will likely trip more readily with continued use at or near it's rating. Exacerbating your original problem.

I would advise against a Rube Goldberg solution, and obtain the proper equipment to stay within your means and have a safe reliable solution.

SAF Ω