Assistance choosing a 3-phase transformer to supply a CNC machine

[Nescius]

Hi everybody,

I'm a mechanical engineer trying to wrap my head around choosing a transformer for a new CNC mill coming into our shop. We're a small operation, so I have ownership of all the technical aspects of the purchase, nearly all of which are squarely within my professional experience. The 3-phase power, however, is a new frontier.

Of course, a real electrician will perform the installation, but I need to have a fundamental understanding of the situation in order to get quotes for the work. Also, the machine represents a significant investment for our shop, so we're proceeding very methodically.

Documentation provided by the machine tool maker seems weak, at least to a layman. A call to the applications department of a transformer manufacturer was little help. Most surprising is the amount of bad and dangerous advice online regarding machine tool wiring.

I greatly appreciate any education that you can provide. Please correct any errors of terminology without mercy. So, here's our situation:

[li]Power company will supply our facility with 277/480 3-phase 4-wire wye.[/li]
[li]Machine tool specs name a range of acceptable input voltage that is nominally referred to as 220 3-phase.[/li]
[li]Machine peak power is 40kVA, maximum continuous power is 28 kVA. I'd guess that most people never use these machines at anywhere near these levels; we certainly won't.[/li]
[li]We're considering a transformer rated for 75kVA to support future machine purchases.[/li]

The following points are taken directly from the machine tool documentation:
[ol 1]
[li]The ground wire is required for operator safety and correct operation.[/li]
[li]A separate ground wire must be connected to the chassis of the machine.[/li]
[li]The ground wire must be supplied from the main building ground.[/li]
[li]Do not use a conduit as a ground wire.[/li]
[li]Do not use a cold-water pipe or ground rod to supply the machine ground buss[/li]
[li]Power can be Wye or Delta type.[/li]
[li]You must ground the power source. One leg or center leg for Delta. Neutral for Wye.[/li]
[/ol]

I wish that their language was a bit more precise or that they provided a diagram. In any case, as a layman with little knowledge of the applicable NEC requirements, I draw the following conclusions:
[ol 1]
[li]Transformer secondary could be either wye or delta.[/li]
[li]With a wye secondary, the center of the wye should be bonded to the equipment grounding conductor (EGC).[/li]
[li]With a delta secondary, either the center of one phase or the corner should be bonded to the EGC.[/li]
[li]The machine does not require a specific voltage between any phase and the ground lug of the machine.[/li]
[li]The EGC should be a wire and should run directly to the ground lug of the machine.[/li]
[/ol]

These conclusions lead to more questions:
[ol 1]
[li]It seems like there is no purpose for a "neutral" coming from the secondary of the transformer. Is this correct?[/li]
[li]Is there a reason to choose a particular primary arrangement, delta or wye?[/li]
[li]Is there a reason to choose a particular secondary arrangement, either delta or wye?[/li]
[li]If a delta secondary is chosen, since I don't need a particular phase-to-neutral voltage, is there a preferred method of bonding to the EGC, either center-of-phase or corner?[/li]
[/ol]

This post is getting awfully long, so let me stop here for now. I have attached a diagram of my understanding of the situation. I welcome harsh criticism of that as well.

I appreciate your help!

 

[LionelHutz]

I would use the wye secondary. It keeps the 3 phase voltages nice and balanced. The electronics in the machine will certainly appreciate it, even if the manufacturer claims it's not necessary.

 

[itsmoked]

While I agree generally with Hutz, if there is no place to land a neutral in the machine then there isn't much use for a neutral. Most machine tools do not have neutral landing posts. Some made for 200V (208V) do and that's so they can use the neutral and a hot to provide 120Vac to run all the controls without including a control transformer (cheap bastards I'd like to strangle half the time). (Mostly European tools.)

If you indeed have three phase and are not using two windings (open delta) but three in a large floor mounted transformer, I'd go with that with one winding center grounded providing 120 to each end of the center tapped winding.

That point on the grounded winding should be connected to the safety ground (bonding) it should be done there and only there. That secondary side of the transformer is called a "separately derived source". All separately derived sources must be referenced to something, preferably ground. Whatever that point is in the separately derived system becomes the neutral of that new system. From then on it is treated as a neutral and not a ground. The safety ground must continue as the safety ground in the separately derived source space.

As for grounding:
They're making much to do about nothing with regards to it.
Yes, you definitely need one.

It should come from the service entrance where somewhere very close an actual ground rod must be in place. Other grounds are available like UFER grounds which tend to be better than ground rods. A UFER ground consists of metal encased in concrete that is in direct contact with earth. Because of the immense surface area a UFER ground typically outperforms a mere copper plated steel rod by an order-of-magnitude. They do need to have a strong capable connection though, so unless you have a purpose-built one or you have big bolts projecting from the concrete, perhaps bolting down the walls with the bolts tied to rebar in the slabs you will probably have to use a rod.

Often the ground wire can be smaller than the phase wires but I'd probably stick with the same size.

Classically you're supposed to run wire 20% larger than the equipment plate to cater to the 80% NEC rule of over-sizing wiring so that it's not loaded more than 80% of rating for continuous use, which a machine tool could conceivably be left to do, running long periods at high loading. Some of the shops I service machine tools at run their machines unattended all night long and even over weekends.

As I said 'classically' that's what you have to do to be 'code compliant' especially if you will be inspected. That said, many of the places I service wouldn't conceive of wiring their machines to the 80% rule and in fact wouldn't even consider 40% of the plate rating. They know their process and how loaded the machines are. I see dozens of machines in lineups with 60A 480V plates wired with 10AWG and 20A breakers. While in shock at such behavior I open the panels pull out a clamp-on and measure 5A. They inform me, "That's our biggest part we subject the machine to".

My attitude is they're protecting the wire more than adequately. If they pull more, they'll drop the breaker and have a CNC mess to clean up, but it's not a fire hazard. It's actually more hazardous to plumb 60A to problem than 20A. But, you have to assess your situation knowing your..situation. If you do piece work and not production, down the road someone may want to hog out some huge chunk of tool steel and the machine might need a lot of power.

Your drawing looks fine. I'd stick with the center path as shown.




Keith Cress
kcress -

http://www.flaminsystems.com

 

[LionelHutz]

Keith, why is the TX required to produce 120V center tapped when the machine doesn't need it. I could be as dismissive of you recommending that as you were of me recommending the wye connection.

 

[itsmoked]

The center tapping wasn't for use with 'the machine' just as more useful 'for things' than a corner grounded. Though, with 480V it wouldn't matter. Now that I think of it the OP stated 480 but diagrammed 240.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Nescius]

Thank you both for the replies.

LionelHutz, your reasoning for a wye does make sense; it certainly can't hurt that the 3 phase voltages are equal with respect to ground. I also read a source (random guy on internet) that claimed insulation life might be prolonged by avoiding the higher potentials above ground that result from a delta secondary. Is this anything more than a hypothetical concern?

Whatever that point is in the separately derived system becomes the neutral of that new system. From then on it is treated as a neutral and not a ground. The safety ground must continue as the safety ground in the separately derived source space.

This has me puzzled a bit. If the neutral of the secondary is bonded to the EGC inside the transformer enclosure, and two wires common to that point exit the enclosure, it's pretty clear what's a current-carrying grounded neutral conductor and what's the EGC. However, if there's only one wire exiting the enclosure, no need for a neutral conductor, and no place to attach one at the load, is it accurate to say there is no neutral conductor and that the single wire in question is the EGC?

I apologize if this question seems obvious. The project is a unique mix of theory, terminology, codes, and best practices. I appreciate the help.

 

[itsmoked]

if there's only one wire exiting the enclosure, no need for a neutral conductor, and no place to attach one at the load, is it accurate to say there is no neutral conductor and that the single wire in question is the EGC?

There is no neutral then and you have the typical 3 phases and a safety ground.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Nescius]

That makes sense. Thanks again.

 

[waross]

Contact the power company and request a 120/208 Volt service, or a 120:240 Volt four wire open delta, or closed delta.
You mentioned possible future expansion.
It amy be well to provide a system with a neutral available.
It may get expensive if you get a machine in the future that requires a neutral and none is available.
Do not under any circumstance allow a Wye/Delta connected transformer.
Wye/Wye, Delta/Wye, Delta/Delta are okay.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Nescius]

I did not communicate with the power company in the early stages of this project, but my understanding is that 277/480 4-wire wye is the only option for us.

The neighboring business had existing 3-wire delta service (not sure of the voltage) and they had to agree to accept this new service as part of our installation. The neighbors readily agreed, as they have experienced some equipment failures that were blamed on the 3-wire delta.

"Expansion" for this particular facility would be realistically limited to one additional CNC machine. There simply isn't room or floor weight capacity for more or bigger machines. Buying an oversized transformer could support this one hypothetical additional CNC.

Regarding Wye/Delta transformers, what is the concern? In my research, I've glossed over some comments centering on the topic, usually having to do with using a delta/wye transformer in reverse and whether or not to connect the primary neutral to anything. I didn't study further, as none of the transformers I've been shopping for have a wye primary. Perhaps that's a clue.

 

[itsmoked]

Loss of a primary line can cause severe voltage crazyness on the delta side as it tries to 'correct' for the missing phase on the WYE side.

Phase loss on the primary causes backfeeds, circulating currents, overloads, burnouts, and/or blown fuses.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

You may be aware that a small voltage unbalance in the supply to a three phase motor may cause a large increase in the motor current on one or more phases.
An unbalanced voltage applied to a wye/delta transformer bank with the primary neutral connected causes a similar effect but often much more severe.
If you float the primary neutral you may have damage caused by over voltage transients during primary switching.
This is the short list of issues.
The full list takes pages and pages.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[waross]

You must be sharing a transformer with other customers.
120/208 Volts is a common, standard voltage across North America.
Large apartment buildings, malls, large super markets, large schools, the list goes on of the normal applications of 120/208 Volts.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[waross]

The neighbors readily agreed, as they have experienced some equipment failures that were blamed on the 3-wire delta.

That is one of the possible issues with a wye/delta with a floating primary neutral.
Utilities often use that connection because the issues become much worse if they connect the neutral.
Generally found in very old legacy installations.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Nescius]

Thanks again. This stuff is actually quite interesting.

 

[jraef]

Power electronics do not like delta power sources. So for something like a CNC machine that will have 3 phase servo amplifiers and possibly VFDs for spindles, you want a wye secondary with a solidly grounded neutral. Your machine will not use the neutral so you would not need to run it out of the transformer, but you want that wye secondary. The only possibility of not needing it is if the machine mfr used 220V single phase input for all of the servo amps and VFDs by tapping two of the three phases, balancing out the current draw among the loads. Possible, but not probable. I only mention it because you seem to imply that this is a small machine.

Common off the shelf distribution transformers will be 208Y120, but you can order one with a 240Y136V secondary if your machine can not tolerate the 208V input. Again, you will not use the L-N voltage so the fact that it would be 136V is irrelevant. This type of transformer configuration exists SOLELY for this purpose; supplying 230V machines that contain power electronics.

The 480V primary of that transformer would be delta, even though your system is wye.

If you CAN’T wait for that transformer and can’t use 208V, it is possible to use 240V delta, but you will have to get into your (3phase) servo amps and VFDs to ensure that the protective elements inside of them are not referenced to ground. Some NFPA privide simple ways to accomplish this by removing jumpers or screws, some require you to disassemble the units to get access to wires that need to be cut. Doing so sometimes voids the warranty, voids the UL listing (if any) but ALWAYS removes the protection features these devices provided. Do do at your own risk.


" We are all here on earth to help others; what on earth the others are here for I don't know." -- W. H. Auden

 

[waross]

If you must go delta, come back for a simple method of developing an artificial neutral or balanced grounding point.
Three small lighting transformers will do it.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Nescius]

Well, we have purchased a transformer: 75kVA Delta-Delta with center of one phase tapped.

I hope it doesn't seem like I ignored the advice of some who gave reasons why a wye secondary might be better.

We got lucky with a supplier right around the corner from our facility. Price and availability were appealing. After agonizing about it for far too long, I finally committed.

Some additional points that the collective might find relevant:

[ol 1]
[li]The machine has its own transformer inside. Could this be why the machine specs freely allow both delta and wye supply?[/li]


[li]Older documentation shows that the machine builder, Haas, has offered external transformers in the past for use in facilities with 480V service. These were Delta-Delta configured. (This is a separate, external transformer in addition to the internal, integral transformer.)[/li]
[/ol]

 

[waross]

Well, if life hands you a lemon, make lemonaid.
On the positive side:
1. You will have true 120:240 Volt single phase available if you need it.
2. The center tap will provide a good grounding point. I would use the center tap for a ground before using an artificial neutral.
3. Delta/delta avoids the problems seen with a wye delta transformer.

BUT:
Beware of jraef's advice.

If you CAN’T wait for that transformer and can’t use 208V, it is possible to use 240V delta, but you will have to get into your (3phase) servo amps and VFDs to ensure that the protective elements inside of them are not referenced to ground. Some NFPA privide simple ways to accomplish this by removing jumpers or screws, some require you to disassemble the units to get access to wires that need to be cut. Doing so sometimes voids the warranty, voids the UL listing (if any) but ALWAYS removes the protection features these devices provided. Do do at your own risk.

If you find that you need a symmetrical ground point come back for details on a "Roll your own" artificial neutral.
If you use an artificial neutral you will not be able to use the center tap to supply 120 Volts.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[itsmoked]

Probably 80% of the Machine tools I service are fed delta or open delta. If they're older than a couple of years I've never seen one that needed the servo drives modified or checked. YES, I see lots and LOTs of new VFDs that scream check-or-I'll-die, because of newer built-in EMI mitigation. If it's a newer CNC certainly check the manual for allowed power connections. In the last couple of years, maybe eight or so, they've all started to include an entire chapter on power supply requirements. Very helpful actually. I suspect they know anything and everything can be out there for power and don't want unhappy customers or blown stuff to deal with and harming their reputations.


The internal transformer you're talking about Nescius is not large is it? About the size of a basket ball give or take? You'd need one at least the size of the biggest motor to have it be a 'system transformer'. What you;re seeing is the "control transformer" which is there to run all the single phase equipment in the CNC machine. All the controls are usually 120Vac. The computers and PLCs etc. are 120 single phase powered and if the machine is fed 3Φ 240 or 480 then they need to transform that down and create a separately derived isolated single phase supply. Depending on which supply a machine gets fed the user/installer needs to confirm the taps are set on the control transformer for the circumstance.

Thanks for coming back. Let us know how the rest of it goes.

Keith Cress
kcress -

http://www.flaminsystems.com