Perplexed With a Delta/Y Transformer

It sounds like an unintentionally corner-grounded transformer (i.e., ground fault). If the output is nominally 230 and balanced, each leg to ground should be ~133 volts.

I discovered the xfmr is relying on the nipple to the cabinet for bonding. There is no equipment ground. Not even from the 480 feed panel.

I just reread this and and was curious to if it applied but it appears that this is referencing a delta config. My Primary is Delta but my secondary is a Wye. Both line and load sides of the xfmr are ungrounded.

http://static.schneider-electric.us/docs/Circuit%20Protection/Molded%20Case%20Circuit%20Breakers/2700DB0202.pdf

I am going to load test 1 pair of devices at a time this sat AM. I will remove them all and reconnect one at time along with some basic tests. I'm thinking a mhegger my come in handy.

So on the wye side you have no neutral, yet you are referencing individual phases to ground?

What is the reference if you have nothing set? I assume the wye side is at least grounded? Is the equipment grounded/bonded? What is the grounding arrangement for your factory if you use no neutrals? (it sure sounds like you have single-phase loads, though).

How about providing a 1-line for your lineup?

Yes, that is true...keep in mind I'm a little green with transformers. The neutral is not being utilized. I have 3ph 480 in 3ph 230 out. Grounding? heh i just determined that there is no grounding! This is my first order of business! I know that this xfmr has been here for 10 plus years and whoever installed it didn't even run a bond to it's adjoining cabinet, it has no ground coming from the distribution center either. I'm glad no one has died from this thing yet.

All we are doing is balancing lets say (30) 2 wire elements rated at 230V between 3 phases. On start up (cold) each phase draws appx 100 amps until they reach our desired temp setting with a controller that speaks to an thermal coupling and an ssr for each zone.

A qualified Electrician, working in the UK to BS7671, would at least measure the earth loop impedance on the secondary side of the transformer before sanctioning its energising so as to have a high level of confidence that the circuit protective device(s) operate so as to disconnect any fault. The problem appears, from here, to be insufficient or missing bonding between the secondary neutral terminal on the transformer and local earth/ground. In its absence, a single phase-to-earth fault would do two things:

1. It would not cause a circuit protective device to operate and disconnect the fault, and
2. It would raise the voltage-to-earth on the other two phases, putting additional stress on the insulation of all conductors connected to each of those phases.

This is not an issue for persistent exclamation marks. Get it fixed before it causes injury/fatality/property damage. If in doubt, consult a qualified Electrician without delay/now/this morning/before tea break.

It was definitely caused by not having XO (Neut) Grounded. As we had device failures downstream it would make the voltage flip flop between 1 and 3. This was a beauty! Thanks for your help, i really appreciate it :)

I am going to do this ...thank you!

If it does blow, it may indicate a neutral/ground fault downstream of the transformer, which is what will give the varying phase voltage described earlier. Once that fault is disconnected, make the bond permanent, then re-energise each circuit one at a time. The one that blows the individual circuit protective device is the circuit that needs remedial action; simply remove the neutral-to-ground fault, then re-energise.

Once that is done, consider installing a residual current circuit breaker in the leads between the transformer and the individual circuit protective devices.

If in doubt, consult a qualified Electrician.

I'm going to do this testing this Saturday. Thank you for your feedback.

The problems in the secondary of the transformer not the devices it powering. At some time the output of L3 is shorting to chassis ground. That's when you get your readings of 230V to ground from L1 & L2 and L3 reads the 3V. The short sounds like the heat in that winding under load is expanding something to cause the short. By the voltages given I would check wire at the terminal board.

Warning if it is shorted to chassis ground the device is not bonded to earth ground or it should have trip out. So be aware that there is the chance of a shock from touching any metal on the transformer.

I think your right.

Why would they use a delta/wye when the are not using the neutral? Is this the right transformer for this application?

Just because someone hooked up some wires and things sort of worked afterwards does not make them a knowledgeable electrician.

They may have been licensed but that in itself still does not mean they knew what they are doing.

The chassis is grounded and we haven't had any issues with blown fuses unless a device literally shorts out which they do. in that case we do blow a fuse.

There is a bonding strap that connects the chassis to the isolated inner framework of the transformer cores. There is nothing else connected other than the 3 Line feeds and the 3 output feeds. Not even a ground wire.

These installers were hacks. The primaries were run 300ft with no ground and it looks like the voltage is most likely floating because the xfmr is wired incorrectly. This is most likely it's way of reacting to an unbalanced laod. This xfmr may have been like this since it's inception 20 years ago.

This is just stunning! Even if the downstream panels are all bonded "correctly" you are risking great hazard by having this system powered. Since there is no reference to the separately derived system your over-current protection devices will not function as intended. There is no way that the ground and neutral which should be delivered to each panel is going to work effectively. Conduit bonding? Worthless.

Ya... hard to believe. I still have a hard time seeing runs with no bonding conductor pulled in. Using the raceway as a bond is pretty much no better than wishing it was bonded.

That's awesome good ole days stuff.

When transformers are wound it is normal to find that certain leg winding to be closer to the ground plane than the rest of the windings.. Unless it is a specially designed and manufactured to be a well balanced in terms of interwinding capacitance, insulations, etc., like those specially built low-leakage isolation transformers!

This leakage problems becomes more pronounced when its core or casing is grounded! Leakage current due to (stray capacitances) the physical construction of the winding /and its insulation as well as with the laminated core when inserted into the coil form!

Therefore, It is very possible that there is really no electrical short in the transformer in question! The erroneous voltage terminal readings maybe cause by an increased current leakage of the transformer? Increased leakage capacitance, due to accumulated dirt, surrounding humidity where the transformers are located? Intermittent voltage levels may not only be due to the poorly grounded casing or ungrounded wye, but may also be due to the kind of tester being used?

After checking the grounds and finding that i do have decent continuity from each cabinet and transformer to ground I'm not fearing electricution. I notified management that we need to pull a ground from the distribution center to the secondary and ground and bond each piece of equipment individually.

I'm still a little foggy on whats next but I'm going to isolate the loads and do some testing before reconnecting them one at a time.

If in doubt, consult a qualified Electrician.

Hi Laurz,

If your transformer is good it is definitely the neutral output is not connected or connected well --this is common problem it will show good on low to no amps then will spike when you put a draw on it I once got a thousand volts on my 220 line from an ungrounded new utility side 4160 transformer with delta primary wye secondary.

The system all tested normal on my side with lights and small outlets but when I started running a 20 hp motor to open my doors (Lets just say they were Real big doors, wherever they were) it spiked and cooked half of my devices.

Expat 07

As far as fearing electrocution you would still need a fault, but that would be easy to come across...but the cabinet of the SDS xfmr is not referenced to the center tap. Until that is fixed (like you said), you won't be good.

Regardless, The system should be recomissioned by a qualified electrician.

That was common in Canada also - but it was restricted to industrial environments with full time electrical staff.

Variable speed drive systems until the last few years where also ungrounded Y secondaries. It has just been the last 10 years or so that code changes require solidly grounded (preferred) or ground fault protection and ground fault indication.

Before slapping on a neutral ground connection, I would suggest testing the transformer. Disconnect primary power, disconnect all three secondaries at the transformer, measure the resistance from each leg to ground. It should indicate meg ohms. If it is a couple ohms, then buy yourself a new transformer. If it checks out OK, connect the neutral to ground. Now check the outgoing phases from the transformer. They should also all read meg ohms to ground. If phase 3 reads low, then you need to troubleshoot the downstream equipment to find your ground fault.

If you ground the neutral, I recommend secondary line fuses - then in the future you know which downstream phase shorts to ground. It also provides better fault protection for your cables.

Remember - disconnect power first!

If you use a megger, be sure to have all personnel away from the associated equipment - it can be shocking.

- This is electrical installation work - and by many jurisdictions requires a certified or licensed journeyman to work on the equipment.

(I had one customer who could not understand that the electrical engineer had to design and specify the equipment, but was not allowed to install the equipment, but had to certify the equipment was correctly installed. By the time the job was finished both the electrical installer and engineer were frustrated with the customer complaining they where milking it for money.)

I pretty much agree with what you have said.

I have worked in a few facilities with such ground fault indication systems. A serious problem was when a ground fault was indicated and not investigated right away. A subsequent fault, if and when it happened would be of much greater impact as it would end up as a phase to phase fault and not a phase to neutral fault.

You said...

"If it checks out OK, connect the neutral to ground. Now check the outgoing phases from the transformer. They should also all read meg ohms to ground".

Once the neutral of the secondary is tied to ground, an ohmeter check of the outgoing phase connection to ground will now read the resistance of the secondary winding. Which will be way lower then the megohm range you mentioned.

Ok I had some confusion in the instructions - the secondary leads where disconnected for the first test, and I thought they were still open, so the test was to check the out going leads while disconnected from the transformer.

Perhaps my error then... I thought you wanted the check done from the outgoing phase connection on the transformer to the ground.

So as it stands, my neutral is floating. You are both suggesting that I should connect the secondary neutral bus to ground? It makes sense.

I will let you know tomorrow.

My plan is to isolate the xformer and meg the primary and secondary in and out feed terminals.

Depending on the results I will run a ground wire and do some bonding then reconnect the existing wiring including a wire from the neutral bus to ground.

I will also test each input leg to the loads with an ohmmeter for general purposes.

This does seem like the most logical approach. I think once the neutral is connected to ground it should stabilize the load line output voltage. It appears that the load may be unbalanced and that is what is causing this issue. God only knows how long it's been like this.

I have quoted out a new transformer and have a Rockwell Service dude coming in on monday.

Yikes $$$$$$

No problem, we are on the same page.

Yes, the "Y" connection (starpoint) on the secondary side either has a loose connection or other high impedance fault.

In a "Y" xfmr the "B" phase voltage will be very unstable if the starpoint is not grounded properly.

This is known as a "wild leg" transformer condition.

Exactly what was wrong!

I have isolated everything.

I megged the xfrm and it's looking good. Once I make repairs on the load side I will connect the neutral to ground with a fused jumper. At present when I meg XO to grnd it is isolated. Starting with a fused jumper should be a profitable attempt for a test.

So far I have found 2 Extruder Barrel heaters that are shorted with no blown fuse. Of course the meg-ohmmeter could be reading potentials in the wiring etc.

I also found a short in the 230vac side of the xfmr in the Die. Time for some repairs and I'll return with my report!

thanks for all your help guys!

This is a classic thread with tons of good responses.

Make sure there is no fuse permanently in the jumper between the XO and the grounding conductor back to your ground grid or in the bonding conductor between the XO and the frame/case of the transformer.

Glad you appear to have been helped by the back and forth on this thread. There is a good group of folk here with vast knowledge, even if we do get rangy by times and argue among ourselves... CR4 is a great resource.

Yea,

The fuse is history, I made up a 1/0 jumper and closed it up. She's runnin good today!

Nice work!

Management will be along shortly to write you up for actually having done something that solved one problem while exposing several others.