Auto transformer questions

I am looking for a 240 x 400 V transformer for a European machine and have come upon an autotransformer with this basic spec, claimed to have a Y high voltage side. I can't find an official spec sheet for the transformer. The machine on the 400V side needs to have a neutral, which I assume means Y configured. Since this is an autotransformer, doesn't that mean that the LV side must also be Y configured? And that the center neutral can be (and in fact must be) connected, so that neutral on each side is the same potential? Since the 240 3Ph has a funky leg, won't this cause problems on the secondary, which will now also have a funky leg, WRT neutral?

A delta Y configured transformer seems like what I need, but 240 x 400 is not common. Curious if I can use the autotransformer.

Autoformer

Swarf Rat;

I just went the opposite direction so I presume YES!

I have a Supertec OD/ID grinder whose control transformer was every imagained voltage in and 120 Vac out. That transformer was dead.

Lacking a suitable autoformer (which could be any transformer with the requisite ratio's on either primary or secondary) and having only simple needs I put in a 2:1 to get 120 from my 240 line.

I just had to break out the Wye and feed both primary and secondary off that Wye.

There is probably some reason why this cannot be done, but I will toss the idea out there for someone else to shoot down.

Why not get three single phase isolation transformers? They are common.
Hook the transformers up delta on the line side, and wye on the load side.

By the magic of the square root of three you would have a 240 volt delta to a 416 volt wye. You would need to make sure the machine can handle that voltage + / - 5%.

Now that is a peculiar idea. I think I have seen isolation transformers with trimming taps, that would correct the difference. If I found a 3 phase, I might be able to rewire the connections and have it all in one box.

Bentworker said:

There is probably some reason why this cannot be done, but I will toss the idea out there for someone else to shoot down.

Why not get three single phase isolation transformers? They are common.
Hook the transformers up delta on the line side, and wye on the load side.

By the magic of the square root of three you would have a 240 volt delta to a 416 volt wye. You would need to make sure the machine can handle that voltage + / - 5%.

Click to expand...

swarf, couple comments for you from a 30+ year transformer designer and seller.... u didnt say size (kva) or I would give u bllpark cost of a new auto and new iso for reference....

autoxmfrs have no pri & secondary or different sides; name pri and sec on aut xfmr is simply terms used to make them easier to understand; each leg is simply a tapped coil.

if this is an older european machine, 416 v may be too high for it; they used to be 380v and stdized on 400 a few years back; older machines had servos that had a hard time with higher voltages and some would fault or fail on 416v....

conversely i doubt there is a 400v machine that will not run just perfectly on 380v too; so in looking for a xfmr, I would also keep eye out for 220->380v models too.

u did not mention 50 vs 60hz? should be no problem running 60hz into that machine but u should be sure there is not something unique that needs 50hz supply too (u can run 50hz things on 60 hz and they will not overheat, motors will run 5/6 the speed tho,, but u cannot go the other way without consideration -- someone in europe cannot run a 60hz machine on their 50hz w/o worrying about overheating stuff). only thing I can think of right off for u would be if there is any speed critical motor on that machine that will cause performance issue if run 5/6 speed....

a 3 ph auto xfmr by design is a WYE unit - must be. cant have delta & y, cant easily be delta - in all my years we never built a delta auto xfmr cuz it doesnt make sense. so the y auto u found is normal.

u can envision an auto xfmr easily: a coil with a tap on it. for 3 phase, hook one side of 3 of thee together, and u now have a 3 ph autoxfmr - u can choose to ignore the center point, or call it neutral and use it if u want. in any case, the 400v is the end point of the coil and the 220 a tapped part 220/400 of the way from the 400v end.

neutral not being same on ur 220 side and 400v side? as u can see from last paragraph, that is not possible. u MAY have reason for concern- depends on where ur 220v 3ph comes from. if 220v 4 wire system (ungrounded delta) then you should be ok using an autoxfmr. if ur 220v has grounded corner or grounded centertap, u may have an issue using an auto xfmr as u say since u may want/need to ground the 400v neutral also - again depends on where that 400v neutral goes and what it is used for. If siemens drives, the neutral also is grounded so u may have issue.

if in doubt, buy an isolation xfmr instead.

Mike, thanks for all that info!

The machine in question (I haven't bough it yet) is a 10 year old German machine, input is supposed to be 30KVA at 400V, they say don't go less but maybe a little more, drives are Siemens and use the neutral. I have 240V 3phase "high leg delta" so it seems to me an auto transformer (which as you say must be Y) will result in 400v "high phase delta". Without knowing what is on phase B, this seems quite dangerous for a machine than uses neutral. It must be a 50/60 hertz machine since it was run off a transformer before (why the dealer didn't collect the transformer when he collected the machine is a mystery...)

A couple of questions: how far off rated voltage on the primary can I go? For example if I find a transformer with the right turns ratio, but the primary is 208V, can I run this at 240? As long as the core is not in saturation is the impedance related only to the frequency?

I think the right way to do this is with a Delta-Y, then connect the secondary neutral back to the building neutral? or building ground? If the machine is internally connected neutral to green ground, then it will already be referenced to ground I guess.

I would like to hear what a new isolation transformer costs, I suspect it is out of budget.....

30kva iso xfmr today is around $ 2500 typically IIRC after an afternoon of enjoying the Christmas present cases of beer

If you want official quote to be sure, PM me and I will get it Tues when back at work.

Siemens drives have a special requirement of "3% or less impedance." very few folks other than OEM mfgrs know this. this is because they regen to the input line rather than dump power into DB resistors on decels. Without the low Z, the noise generated is horrendous: they short the input 400v line to neutral at 2khz rate to make this happen to their liking. This Z is about 1/2 the "normal 5-6% impedance" of off the shelf xfmrs. An auto xfmr has typically 1/2 the impedance of an iso.... all that said, I can tell u of case in NY once where CMI sent a siemens drives machine to a customer who used 6% Z xfmr and it caused all the clocks in the 3 story building to run backwards when ever the machine was turned on - until they bought proper iso xfmr from us

ur 240V 3phase "high leg delta" does not jive with me; what does this mean?

how high can u go? to be sure, get the SELLER to answer that question for you.... my experience with siemens drive is that if you go to 10% high line u may be in trouble... but that is just a personal opinion as I am not a siemens rep..... If it were my machine, I would not go over 405v on a 10yr old design or I would expect drive failures every year until it was reduced to 390v.... 10 yr old designs were not very forgiving on high dc bus voltaqes on the igbt transistors. but again, that is just my opinion based on 30 yrs of screwing with them....

swarf_rat said:

....
For example if I find a transformer with the right turns ratio, but the primary is 208V, can I run this at 240? As long as the core is not in saturation is the impedance related only to the frequency?

Click to expand...

missed this question in my last reply, so will give .012cent worth.....I will break it down to 2 questions:

can u run a 208v rated xfmr at 240v input?

PROBABLY but no guarantee.... we just had new 400kva xfmr design at a customer who put 495v into the 460v tap and promptly overheated the xfmr to the point of having varnish drip out of it.... reason? this particular xmfr design was too close to the hi end of flux density; the core losses alone caused the xfmr to overheat even tho it was running almost no load - after 10 hours on being turned on. customer retapped to +10hi tap and core temp promptly dropped to normal temp and they are good to go. But moral of the story is it may be that running 240 into a 208v tap MAY overheat the xfmr.... there are TWO things that heat a xmfr: core losses and load losses.... any xmfr design can be made to have low core losses and higher I^2R (load) losses (more typical when in Nema 4 enclousre) , or high core losses and low I^2R losses, or equal.... depends on the mood of the xfmr designer at the time he designed it or his spec requirements.....

Second part of question about impedance (Z): Z has nothing to do with voltage input over or below rating so this is not valid question.

mike_kilroy said:

ur 240V 3phase "high leg delta" does not jive with me; what does this mean?

Click to expand...

This is the 240V that is really 220 single phase with a third stinger leg. The two normal legs measure 120V to neutral and the third leg is about 200V to neutral. The phasing is arranged so that to a delta load it looks right. It seems to me that this cannot be connected to a Y primary without some strange voltages on the secondary. And since an autotransformer has to be wired same on both sides (doesn't it?) and I need Y on the secondary then I can't use an autotransformer?

You can probably use it OK functionally, but you can't ground the "virtual neutral". There is no actual neutral for the 3 phase, only for the 120V

For any other voltage that is a problem for the electrical code, the code specifically requires grounding. But there is an exception for 240-208 transformers, so the high leg/lighting tap/wild leg/stinger 240V is OK with autotransformers for getting 208, or at least it was the last time I looked it up in teh code.

A delta-delta wired autotransformer should be able to get you 208 (or 400) from 240. But you do not get a neutral leg. I apparently need the neutral leg, which I think means Y secondary. And I think that means a Y primary because I cannot see how you would create an autotransformer that is delta - Y (or Y - delta). Unless it had a different turns ratio for the B (high leg) phase. If you start with 208 "real 3 phase", then things are easy.

You wire the autotransformer Y. The Y point goes noplace, it is a "virtual neutral". You ignore teh "neutral" on the 3 phase for "high leg" 240V, because connecting it will cause problems due to the already-present 120V neutral.

if you have 240V 3 phase with 1 phase grounded, same deal.

Works fine, just as with motors that are wired Y with no connection to the Y point. It has no path to ground, but doesn't need one if the load is balanced, as I assume it is in this case.

With an autotransformer, since by definition the "secondary" and "primary" share the very same wires, whatever way the the "primary" is wired the "secondary" is also.

Yeah, that is my understanding. But the Y connection point on the 240 primary side is going to be bouncing around compared to ground, or the other phases, no? And on the 400V secondary too, so on a machine that uses it expecting it to be somewhere near "neutral" might be expensively disappointed. 3 phase motors wouldn't have a problem, as they are a delta load. But if the servo drives are expecting 200V from each leg to "neutral", then.....

It is generally bad form to demand a balance to neutral, as with for instance an aux power circuit for controls..... While some *might* hook it L-N, that is crummy practice, using a transformer to isolate it, pri on 2 stable lines (in your case the 120-0-120 if that is of interest) and secondary re-referenced to machine ground as a "separately derived source".

I understand CNC are different, and may demand a good neutral for various reasons.

In that case, I suggest bagging the autoformer idea, and instead going with a delta-Y transformer, with Y on the machine side, connected neutral to ground as a "separately derived source".

You COULD use a Y-Y, if you had one, but presumably not a delta-delta if you need a neutral.

Since you need 400V, a 240 delta secondary properly reconfigured to be Y would give you 415V nominal on the high side with 240 delta input. 400V 3 phase is 230 to neutral/ground, which is why the europeans use it.

Also 3 single phase "4-way" (240/480 to 240/480) could be configured any way you want, and could, if properly sized at 1/3 of required demand each, do 240 to 400 wired as primary delta and secondary Y.... you use the 240 connection both sides.

You need 30 kVA, so 3 x 15 kVA "4-way" would do it.... they may be more common than the 3 phase version

I agree that the neutral requirement is not great. Seems like it has to be referenced to ground, and it may have that connection inside the machine already.

Not sure I have seen the "4-way" you are talking about. Is this an isolation transformer with a 480v primary and 240 tap, and a 480 secondary with 240 tap? So i would wire them delta-Y, using only the 240 taps? If I could find some with trimming taps I might be able to get pretty close to 400V.

swarf_rat said:

I agree that the neutral requirement is not great. Seems like it has to be referenced to ground, and it may have that connection inside the machine already.

Not sure I have seen the "4-way" you are talking about. Is this an isolation transformer with a 480v primary and 240 tap, and a 480 secondary with 240 tap? So i would wire them delta-Y, using only the 240 taps? If I could find some with trimming taps I might be able to get pretty close to 400V.

Click to expand...

remember the siemens require a solid neutral to work properly.... also see if seller or machine nametag states tolerance on the 400v, as 10 yr old siemens may have been 380 with 400 as a max and 415 may stress them - just something to ask to see if u can get real answer for that machine.

I think most '4 way' xfmrs I ever saw were just that with no taps at all.

A 4-way has 4 identical windings.... Common types are either all 120V, or all 240V. They are available in sizes at least up to 15 kVA, which is more than you need.

The 120V winding types go 240:240, or 120:240. The 240V winding types go 240:240 or 240:480. You can use either voltage, because all you need is the 240V setting.

Wired Y, the 240 output gives 415V 3 phase IF you have a real 240 coming in. Otherwise proportionately more or less

I never see any 4-way units with taps, but your machine should tolerate some variation. Normally they are required to tolerate +- 10%, which would be 360 to 440V based on a nominal 400V. In that case, any voltage from 208 to 254 should be OK. The 415V is just under 4% high, and should not be an issue.

And, with transformers in series, the net series impedance will be up, and under load the voltage will fall somewhat more than if you could connect direct.

As always, check the manual for the actual power requirements. It may be much more fussy than I think, in which case even your basic service stability may or may not be sufficient to satisfy it. Lots of the "farm"/wild leg/"lighting tap/etc 240 services are "open delta", using only 2 transformer windings, which is not nearly as stable in voltage.

If machine voltage input is based on 230V to neutral, nominal would be 400V (398, actually) and 6% high would be 424V, 10% high 440V.

If based on 220V to neutral, nominal would be 380V, 6% high would be 400V, and 10% high would be 418V, not much margin.

My power seems to be about 243V on the A and C phases, and around 200 on the B. On my Deckel I'm getting 380 on two phases and about 375 on the other.

The machine in question says 400V on the placard. That is all I know. I will have to find out if that is +/- What.

There seems to be a set of 240/480 four way transformers near me and even have trimming taps. But they are 50KVA. There must be a disadvantage of Too Big. This is about 5 times bigger than I need. Would these be hugely less efficient, burn more idle power, or what?

swarf, i checked some old siemens drive specs today and found the common 611 model was rated 400/415v, +/-10%! so guess my comment about watching input voltage being high was off base for them. but I do recall YEARS of close work with Cincinnati supplying their transformers for siemens since around 1980, and back on the pre 480v input models their engineers being VERY concerned over slightly high voltage inputs to them. years like that. they felt they did not hold up well at all at much over 400v input. so. sorry for my possibly crying wolf; take it with a grain of salt

but they do still have a very strict rule of 3% impedance or less on the input transformer, and as you see from the posts here, 5-6% is common and even higher for 3 sep ones used to make the 3 phase. guess this is why we still sell so many transformers to siemens themselves and cincinnati still.

mike_kilroy said:

u can run 50hz things on 60 hz and they will not overheat, motors will run 5/6 the speed tho,,

Click to expand...

6/5 speed

Bill