What's new
What's new

Using a 15kva 240V/480V dry transformer to run 460V VFDs

darronb

Aluminum
Joined
May 19, 2003
Location
Reno, Nevada, USA
Howdy all,

I bought a used 15kVA dry-type 240V / 480V transformer earlier this year. I will be using it to power 460V VFDs on several machines. I like to buy used 460V VFDs because they are often much cheaper on eBay than 200V-class models.

I bought a “10HP” GE/Fuji 460V VFD that GE tells me is fine on single-phase input, when maximum output current is de-rated 50%. The VFD’s full rated output is 18A. So my 7.5HP 4-pole lathe motor should run well on this VFD, as the motor is rated at 9.5FLA at 460V. I will rarely need to run at full load.

I used my 115V variac and a 1.5kVA control transformer (wired for 1:4 step-up) to gradually reform the VFD’s caps, just as a precaution. The seller told me it had only recently been removed from power. Though I have not yet run a motor with it, it powered up nicely and seems to be in working order. Not bad for $50…

Today I powered up the big (230+ pounds!) 15kVA transformer for the first time. After staring at the connection diagram plate on the transformer’s case, and the existing connections inside, I was quite confused. Neither connection pattern shown on the diagram seemed to match what had been done inside. Instead, it seemed that the existing connections were a combination of the two diagrams shown. So a little doping out with a voltmeter let me know that the transformer seemed to be wired for 1:1 voltage, perhaps for use as an isolation transformer. Is there a common need for an isolation transformer with this large a capacity?

After re-connecting to match the 480V/240V diagram, I wired up a 230V input to the low side. I don’t yet have the contactor needed for my high-current 230V circuit. So I used my 115V variac and 1.5kVA control transformer (wired for 1:2 step-up) to energize the 15kVA transformer from a 115V 20A circuit. Energizing with the variac at 0V, and then smoothly turning up the voltage allows powering up this huge transformer from an ordinary 115V outlet without a huge inrush current.

This method of testing a large transformer does have a distinct danger, though. If the power fails while energized at full voltage, and the variac is left on at full voltage, the inrush current spike when the power returns will definitely pop something. Hopefully just the 20A breaker on the 115V circuit, but I wouldn’t count on it…

All seemed to go well. No bad smells or excessive heat, just the “pleasant” hum of a large transformer. My 115VAC wattmeter indicated about 1100VA current flowing once 480V was reached on the high side of the 15kVA transformer, and about 210 watts. Some of the watts are being burned in the variac and 1.5kVA control transformer, of course. I don’t have any way to measure watts at 230V or 460V, though. Does anyone have an idea about how many watts are consumed by an idle 15kVA dry transformer?

I will shut off the contactor on the input side of the transformer when it’s not needed. Other than the huge inrush current load on my domestic power when energizing this transformer, is there any problem with frequent (several times a day) power cycles on a large dry transformer?

Full Load Amps on the 240V side of this transformer is 62. I ordered a 75-Amp contactor, which should be adequate. Can I get away with powering the transformer with a 60A breaker in my main panel? I will never be running the transformer at a full 62A, other than during the moment of inrush current. Any idea how much inrush current there is on a transformer this size, and if a 60A breaker can handle it? I will use a 100A breaker if needed, and appropriate gauge wire. The transformer will be located just a few feet from the main panel, so the high-current wire run will be short.

I’ll be able to use much smaller wires on the load side of the transformer, since none of my individual 460V loads will exceed even 20A.

15kva_transformer_1.jpg


15kva_transformer_2.jpg


15kva_transformer_3.jpg


15kva_transformer_4.jpg
 
Some of the tranformer gurus will likely pipe in here and give some real calculations on your setup.
I use the same idea but on a 50% scale. I suspect you will have problems with the 60 amp circuit but may be able to get it to work if you can get a time delay breaker (I can not remember the correct name for them). If you use an ordinary fast trip house type breaker I think you will trip the breaker while saturating the transformer. The 100 amp circuit will probably work but you need time delay fuses at the transformer to protect it and you will need them on the 460 volt side also to protect the transformer from surges or transients coming back from the Load side. See the National Electric Code (NEC) for the specifics of the fuse sizes allowed. Your public Library should have a copy of the NEC. I do not have one with me at the moment.
This setup works nicely for 460 volt VFDs. You have one discrepency in you statement about the output amps of the VFD. The amps stay the same as the nameplate shows but you can not put as much power into the VFD with single phase so you must derate the rated POWER output by 50%. Thus you can expect to run a 5 hp motor from a 10 hp 3 phase input VFD. The maximum theortical output would be 59% but most VFD manufacturers say to use 50% due to ripple in the output casused by the single phase input. This assumes you stay within the rated input amps and the rated output amps.
On the output side you can put some 3 pole fuse blocks and fuse for smaller wire to run to your tools if you want to run wire smaller than the amp rating of the transformer. I think the output side will be about 30 amps and you may want to run something smaller to the tools but you must have fusing (breakers) to protect the wire.
 
I forgot one other item:

It is a good idea to put load reactors between the 460 volt VFD and the load motor to prevent high frequency transients from damaging your motor. this is espically true when you have old and hard to replace special motors. If you have new inverter duty motors then do not bother with the load reactors.
You may be getting the idea by now that it is about as expensive to do this as it is to buy 230 volt VFDs. I only use the 460 volt set up for motors that must have 460 volts.
 
Thanks for the info, toolnut.

I'll have to do some searching for a 60A time delay breaker. I realized after thinking about it a bit that although the transformer may start every time on a 100A circuit, it would not be properly protected by the over-rated breaker.

On the output side, my plan is to feed the output of the transformer into a 460V sub-panel. It would seem that if I have a main 30A breaker in the sub-panel, that would protect the transformer from problems on the load side, correct? I would install smaller-value breakers in the sub-panel to feed each 460V circuit.

Is “600V” rated wire OK for a 460VAC circuit? I ask because I know the peak voltage on a 460VAC circuit is higher than 600V.

As for the need to de-rate the VFD, I spoke on the phone with a GE engineer. I asked him specifically if I could run only a 5HP motor on my “10HP” VFD, or if halving the maximum output current was the proper method. I was told to de-rate current, not HP. It seems that the “horsepower” ratings of VFDs are not very precise, anyway, as two “5HP” VFDs from different makers will have substantially different max. currents, and the same is true with motors. VFD HP ratings seem little more than a ballpark guide, with max. current being the proper way to rate them.

I should not need the load reactor between the VFD and my lathe motor. The motor is an almost-new dual-voltage US Motors VFD-rated 7.5HP 4-pole. It was $80, and was needed for my lathe, anyway, so it was not an extra 460V-related expense. The VFD was $50, as mentioned, which in my experience is way less than a similar 230V model would have sold for.

The 480V/240V 15kVA transformer cost $75, so it plus the $20 75A contactor and some other related parts are the main 460V-related expenses. I will save a few bucks by needing smaller gauge wire from the 460V sub-panel. I would have been running dedicated 230V circuits for each tool, anyway. There are other tools which will eventually get 3-phase motors and 460V VFDs, so the transformer and sub-panel are a one-time expense, to be amortized with each new circuit added.

Truth be told, even if I had not been able to save as much on the VFDs by going to 460V, I just think this transformer setup is really cool. I love to tinker (safely) with such things, and thoroughly enjoy finding different ways to accomplish a project.

Darron
 
So many questions, such a little response box...

60A breaker should be fine, although if you ever fully load the transformer, that will likely trip on you. Breakers are really only rated for 80% of their nameplate (long story), so if you want to fully load the transformer, use an 80A breaker, then size your conductors accordingly. The inrush of a transformer is rarely high enough for long enough to trip a circuit breaker (but it is often high enough to blow a non-time delay fuse). So size the breaker for the load, then size the cables for the breaker, don't worry about inrush unless you prove you have a problem later. By the way, all breakers manufactured since 1997 are now "time delay" (used to be called HACR), because UL changed their rules to make them all like that. If you don't know when a breaker was made, look for the "HACR" designation somewhere on the label or body. If it doesn't have it, it isn't. If it's new out of a box and doesn't say HACR, it is anyway.

The full load current rating of a 15kVA xfmr on the 480V side is 31.2A, so yes, a 30A CB would be OK, but the same 80% rule applies there as well. I would go with a 40A if you need all the capacity.

600V wire is what you need to use. Ratings always refer to RMS voltage, not peak.

Bingo on the current vs HP ratings issue.

VFD rated motor should be OK, as long as your distance from the VFD to the motor is not more than about 50' (some say 25', but that would apply to a non-VFD rated motor IMHO).

It sounds as though you have a sweet deal there. I don't recommend it to everyone, especially if you are not comfortable with electricity, but it sounds like you know enough to be safe.

I know what you mean by the tinkering comment. My home office looks like a lab right now.
 
Re inrush current, it depends on the design of the transformer, low copper resistance and high flux density makes for high inrush current, the more conservatively rated transformers tend to be better mainly due to the lower flux densities they run. Jraefs advice is good, try it first, it is not difficult to fix if it is a problem though.

Loss on a 15Kva transformer anything between 200W and 750W.

If your 115V wattmeter presents a resistive load on a separate voltage sense line then a simple voltage divider say 4:1 on the voltage input will do the trick, just multiply the measured watts by 4.
 
Jraef, thanks for the circuit breaker info.

I read elsewhere that depending on where in the AC cycle the transformer was when switched off, and where in the cycle it is when re-energized will make a BIG difference in the inrush current. The figure I saw was up to 30X nameplate FLA current when everything is "just right" (or just wrong!). That would put the theoretical maximum inrush for my 62FLA-rated transformer at not far from 2000 Amps. Woohoo!

I did notice quite a difference in the small inrush current between individual starts from my variac. Even with the variac set to minimum, there was a measurable (and slightly audible) surge when energizing, even at the fraction of a volt that the variac supplies at it's minimum setting.

My 7.5HP motor will be only a few feet from the 460V VFD. I'm trying to find some VFD motor cable, but it seems most dealers sell by the large spool only, and I don't need much. I just finished installing a sensorless-vector 1HP 230V VFD by my drill press, and all I had was a piece of some old orange 14-ga. stranded extension cord. I used the three conductors (colored white, black, green- typical 115VAC colors) inside the extension cord for the three motor phases, and ran a separate ground wire from the VFD chassis to the motor frame. Works OK, and certainly better than the ancient 4-conductor cable I had on there before with my old VFD. That old cable had severely crumbling rubber-like insulation, so bad it would fall right off the conductors when they were bent more than a little...

I've seen references on this board on the need to over-rate a transformer when the load is a motor. I'm guessing that this only applies to motors that are started across-the-line. A VFD-run motor with a slow enough ramp-up time should not have the over-current spike that would seem to me to be the reason for the need to over-rate a transformer. Is this correct?

Helicalcut, you say that “Loss on a 15Kva transformer anything between 200W and 750W.” I assume this means the loss at full load? I’m wondering what the watt loss is with no load, since I will at times be leaving the transformer energized with no load for perhaps a few hours at a time.

I’ll have to see about modding my wattmeter for use at higher voltages. I actually have two wattmeters, the little “Kill-A-Watt” meter shown above, and a much larger one in a large steel enclosure that would be easier to get into and modify.

Darron
 
That worst case (I've never heard of it being more than 20x BTW) inrush would be for 1/2 cycle, i.e. no more than 8ms. Nothing acts that fast, so don't worry about it. That's the stuff of engineers' nightmares, it's pretty transparent to the rest of the world.

No matter how slowly you increase the voltage, you still need to magnetize the core. That's where the inrush comes from. You can mitigate it, but never eliminate it.

For that short of a run, I would just use metallic seal-tite flex conduit (not the new all plastic kind) and machine tool wire or THHN. Twist your wires together before pulling them in. Make sure the metal part of the conduit is grounded on both ends. Don't use unshielded flexible cord, it will act like an FM broadcast antenna.

Yes, over-rating a transformer is for full speed starting. Don't worry about it with a VFD.

Total losses in a distribution transformer, assuming an older non-energy-efficient design, are typically around 3% at full load. So on your 15kVA transformer, that amounts to about 450W. Some of the losses are magnetic and therefore continuous, i.e. regardless of load. So they will be there if powered up with nothing on it. They typically represent about 35% of the total full load losses. There are additional copper losses that are load related, i.e. the amount changes with load. At full load, they will represent about 65% of the losses. So compared to full load, if your total losses are 450W, the magnetic portion would be about 158W. Hey, pretty close to what HelicalCut estimated!

It will be interesting to see how that compares to what you read in real life.
 
158 watts, huh? That sounds like it should be just about right on the money. Using my 115V Kill-A-Watt to measure, the total watt load with the variac, 1.5kVA control transformer and 15kVA transformer in series was about 210 watts.

I know from separate measurements that:

The variac uses about 5W at no load.

The 1.5kVA control transformer is 49W no load.

210-5-49=156 (wow- you're good).

Darron
 
Aargh- one more question...

Assuming I place the 15kVA transformer close to the main panel, I will need only very short wire runs.

Using this wire size calculator, and entering 5-foot length, 240V, 1-phase, copper, 62A, I get a required wire size of only 14-gauge!

Can this be right? The calculator page mentions a maximum voltage drop of 3%. Would the slight extra impedance of 14-ga. wire over, say 8-ga. reduce the inrush current a little? (I know we've determined that the inrush is not going to be a problem, I'm just curious...).

Darron
 
Any increase in supply impedance will reduce inrush current.

Voltage drop is irrelvant for wire size calculations on short runs, thermal issues dominate, My guess is that 14Ga will burn off the insulation at anything over 40A.

I see the wattmeter in the picture doesn't have the voltage sense brought out to a separate terminal.

My watt estimate was for a loaded transformer but Jraef seems to have got a better estimate
 
Howdy all,

I don't know if thread resurrection is frowned upon here. I apologize if it is.

I am finally getting to the installation of this 15kva single-phase transformer. It will be used to step 240V up to 480V. It will be located about 10 to 12 half-circuit feet from the 200-amp breaker panel. I am trying to determine the best wire size to use. There is a wide range of opinion on this when doing a Google search. I will be using copper wire, since the short run means the (much) higher price of copper will not break the bank. I also prefer the smaller gauge of copper, for any given current capacity.

I've looked at voltage drop calculators, which based on the short run and max current of 62A recommend a 14-gauge wire :nutter:. Not that I would consider such lunacy, but perhaps some silicon-insulated finely stranded 14-ga. copper wire made for powerful RC model motors would survive, provided it was near nothing flammable...

Calculators based on the NEC are confusing, since there are so many variables to take into account. My less than fully educated guesses give a result of #3 copper required :eek:

The various seat of the pants recommendations tend to call #6 copper good enough for a ~60A circuit. Some call for #4 copper.

I will be using a 60A breaker in the panel to supply this new circuit. I realize a typical HACR breaker can only sustain 80% of rated current continuously. After the very brief inrush current, my continuous load may never reach the 62FLA capacity of the low side of the transformer. I have some 80A HACRs if I need to change the breaker. This would be getting back into marginal wire safety range if I ran #6 copper, I suspect.

I think I can get away with #8 copper for the grounding conductor. There will be no neutral.

I have a 75A contactor that will energize the transformer. The 480V output side of the transformer will be protected by 480V-rated 30A fuses. I'll get the type that's rated for motor starting, just to be safe against nuisance pops (I forget what they're called).

The load will never be an across-the-line motor start, since a 480V single-phase motor is probably as rare as hen's teeth. Instead, I will run one or more VFDs, each (de)rated for operation on 480V single-phase input. As far as I know, inrush on a VFD is minimal, since there's some sort of current limiter built in that prevents the discharged caps from sucking up all the juice that the line can supply.

So there we are, I'll be buying back the copper I recently sold to the scrap yard, in new, shiny form. The question begs, how thick and beefy must it be?

Darron
 
Think about this...

Will you be running that transformer at that HIGH a load? If not, how high will you be running it?

If you're not running more than 30A, you could concieveably make your branch circuit run on 10AWG stranded... the only thing that you'd have to worry about after that, is inrush.

The way I would do it (assuming you're not gonna be loading it silly heavy), would be to pull a 50A range plug service from your panel, put a breaker in it accordingly, and use #6 as any other 50A service would.

Then, put your transformer, and a few other controls on it, feed it from an off-the-shelf 50A range-cord. That way, there's no question about your method of disconnect, should you be challenged by a codus investegamos.

Next: Use TWO contactors for operating this bugger... make ONE of 'em your basic across-the-line, and make the OTHER one just like it, with exception that there's a pair of light bulbs or resistors (or both, in parallel) across the contacts of the second.

When you power it up, power up the across-the-line (so the transformer will initially 'start' with resistors in series of the primary, and then a short time later, a timer relay (3 seconds?) will close contacts that bypass the resistors, and you'll have full-load connection.

This would also keep inrush in check in the event that you accidentally have a load on the secondary (like VFDs, or some other largely-capacitive input) when powering up.
 
huh?

forget about the load when picking wire size - pick wire size that will not melt if a short happens - ie., the proper wire size is to carry the current rating of your circuit breaker! dont screw with anything smaller! even at home!
 
Mike's right...

Yep, I guess I should have qualified that...

Since your'e running VFDs downstream of the transformer, once you've limited inrush current for starting the transformer up, size your breaker for some reasonable amount, and then wire accordingly.

How much power do you plan on pulling overall?
 
Since your'e running VFDs downstream of the transformer, once you've limited inrush current for starting the transformer up, size your breaker for some reasonable amount, and then wire accordingly.

How much power do you plan on pulling overall?

For now, I plan to run just the one 10HP VFD (derated by 50%). But I'd like to be able to load the transformer to 100%, or close to it, should the need or desire arise. And I'd rather not have to rewire with a heavier gauge at that point, just to save a few bucks when buying wire this first time. I'd also rather not use heavier wire than necessary for safety and performance.

Therefore, I'd like to size my input cable and breaker accordingly. So I'm still wondering whether #6 Cu will do, or do I need #4 Cu for a 62A load? I'm guessing that I should go for #4, since I will likely have to run an 80A breaker to handle a 62A continuous load. Unless I can get a 60A breaker rated for a continuous 60A.

I'll set up something like the inrush damper suggested if nuisance trips become apparent in testing. If trip ye not, stalwart breaker, rush on in and out of there, you zippy little electrons!

Thanks,
Darron
 
Next: Use TWO contactors for operating this bugger... make ONE of 'em your basic across-the-line, and make the OTHER one just like it, with exception that there's a pair of light bulbs or resistors (or both, in parallel) across the contacts of the second.

When you power it up, power up the across-the-line (so the transformer will initially 'start' with resistors in series of the primary, and then a short time later, a timer relay (3 seconds?) will close contacts that bypass the resistors, and you'll have full-load connection.

If inrush does become an issue, I thought of a modified version of your method to limit it:

I could wire a 60A or greater switch between the panel and contactor. I think I have a big switch like that around here somewhere.

I could then set up the contactor with its contacts paralleled with resistors or lightbulbs of proper impedance, as you suggest. It would then be effectively a fully manual system, as opening the contactor would still let some dangerous level of current flow through the resistors, so manually opening the switch would be necessary to render the circuit "dead". I suppose that the contactor would be rather unnecessary, since a second switch with resistors would do the same thing as the contactor.

Depending on how loud the contactor and transformer are when on and idling, I may want bright indicator lights in the room with the transformer, and in the room next door with the machine tools. This way, I won't leave the shop with the transformer still powered, sucking up 150+ watts all night (or week). I can get inexpensive high-power LED drivers that will run from 240VAC, and connect them to the output side of the contactor with low-value fuses (250mA should do nicely). A 1-watt colored LED is VERY bright, so when placed in the right spot, it should be an instant reminder of power status.

Darron
 
Sure!

That'll all work. If you're planning on running it to full load, then definately wire it heavy.

Another thing you COULD do, but may not be sensible in your application, is put the transformer close to your 240v panel, and then pull your 480 wire in conduit to the shop... that way you'll be avoiding long stretches of large-gauge wire. May not be suitable for your situation, but it's a thought.

As for operating- if you make it totally contactor/delay timer operated, you could put a motion sensor in the shop... and make it so that if you turn off the lights, or leave the room for more than an hour, it shuts it down...
 
If you're planning on running it to full load, then definately wire it heavy.

Another thing you COULD do, but may not be sensible in your application, is put the transformer close to your 240v panel, and then pull your 480 wire in conduit to the shop... that way you'll be avoiding long stretches of large-gauge wire. May not be suitable for your situation, but it's a thought.

As for operating- if you make it totally contactor/delay timer operated, you could put a motion sensor in the shop... and make it so that if you turn off the lights, or leave the room for more than an hour, it shuts it down...

The wire run from the panel to the transformer will be only 12 to 15 feet (half circuit length). I'll just run #4 Cu, and stop sweating the details. That way, if I ever need to run it to full load, I'll have the capacity.

Can I run "Romex" type plastic-sheathed #10 Cu cable for the 480V side, since it's 30A? Or do 480V conductors have to be in conduit to meet code? I will fuse at 30A on the 480V output side of the transformer.

I already have some NEMA L8-30P 480V single phase 30A plugs, so I can wire up an outlet or two for temporary VFD/motor testing. Now I just need to find a good deal on some NEMA L8-30R receptacles.

I like the motion sensor idea. I'll have to scrounge up something.

Darron
 








 
Back
Top