residential isolation xfmr

Status
Not open for further replies.

jeremysterling

Senior Member
Location
Austin, TX
This is a picture of a 15KVA 240V to 240/120V "isolation" transformer that I installed in the attic of a single family dwelling as part of a remodel. (Yes, me and an apprentice lifted this 170 lb. XFMR through the attic access). The plan is for primary and secondary ungrounded conductors to be #6. My feeders will be on 60A bkr. My secondary grounded conductor will be a #4 (HO's preference).

The service is approx 80' away. I'm using 1" smurf and FMC to the existing main panel and new sub-panel (approx 30' away.)

In the existing main panel, I see the existing cold water GEC (looks like bare #6).

Do I need to install a bare #6 terminated with the GC at the neutral taps on the secondary of the XFMR, run outside of the feeder conduit, and crimp it with like a c-hypress on to the existing cold water GEC (versus landing it on the neutral bar) in the existing main panel? Would this be the best and easiest way to bond the new SDS GC?

Would a green#10 be OK for the EGC ran in the feeder and secondary conduits?

If I'm bonding the GC at the XFMR, the neutral bar in the sub-panel should not be bonded to the sub-panel at the sub-panel?

I am also concerned the HO will see the green #10 to the EGC bar and think he's not "well grounded." I think I may install as large of EGC that will fit with my other secondary conductors, perhaps a #8 or even #6 (which may overfill my 1" smurf). Would you do the same?

I hope this is detailed enough to paint an accurate description of the install.
 
This is a picture of a 15KVA 240V to 240/120V "isolation" transformer that I installed in the attic of a single family dwelling as part of a remodel. (Yes, me and an apprentice lifted this 170 lb. XFMR through the attic access).
looks close to being a 450.21(A) Violation
The plan is for primary and secondary ungrounded conductors to be #6. My feeders will be on 60A bkr. My secondary grounded conductor will be a #4 (HO's preference).

The service is approx 80' away. I'm using 1" smurf and FMC to the existing main panel and new sub-panel (approx 30' away.)
Violation 240.21(C), I believe
In the existing main panel, I see the existing cold water GEC (looks like bare #6).

Do I need to install a bare #6 terminated with the GC at the neutral taps on the secondary of the XFMR, run outside of the feeder conduit, and crimp it with like a c-hypress on to the existing cold water GEC (versus landing it on the neutral bar) in the existing main panel? Would this be the best and easiest way to bond the new SDS GC?
See 250.30(a)(7)
Would a green#10 be OK for the EGC ran in the feeder and secondary conduits?
yes
If I'm bonding the GC at the XFMR, the neutral bar in the sub-panel should not be bonded to the sub-panel at the sub-panel?
at transformer or panel., not both
I am also concerned the HO will see the green #10 to the EGC bar and think he's not "well grounded." I think I may install as large of EGC that will fit with my other secondary conductors, perhaps a #8 or even #6 (which may overfill my 1" smurf). Would you do the same?

I hope this is detailed enough to paint an accurate description of the install.

Only one question. Why ??
 
As Gus asked, what is the purpose of this "isolation" transformer? Especially given the fact that you can't really isolate anything and comply with the NEC rules. These rules will require that the primary EGC be connected to the secondary grounded conductor.
 
looks close to being a 450.21(A) Violation
I'm good there by at least 6":smile:

Violation 240.21(C), I believe
Yeah, good catch, I said 30', probably closer to <25'! I may have to put a breaker enclosure or fused disconnect in the attic, darn it!:mad: Maybe, not.:roll:

See 250.30(a)(7)
OK here's where I ask the AHJ what he wants to see:wink:

yes at transformer or panel., not both
I got that concept down cold:cool:


Only one question. Why ??
Cleaner power?:grin:
 
As Gus asked, what is the purpose of this "isolation" transformer? Especially given the fact that you can't really isolate anything and comply with the NEC rules. These rules will require that the primary EGC be connected to the secondary grounded conductor.

I've tried to explain to the HO exactly that. Rack mount clean power could have been a better alternative. Not really my expertise.

The purpose is immaterial to the install. Sorry I can't say more than that.

The branch circuits will be dedicated IG 5-20R duplex installed in PVC and FS boxes! I know, I know, why do I need IG recepts when the EGC is clean all the way back to the sub-panel?
 
I think I would just have found a way to put it back at the main service.
The whole installation seems very odd.
What are you guys trying to accomplish?
What will the temperature be in that attic during the summer?
 
081221-1002 EST

jeremysterling:

What is "dirty power" in this application?
How does an "isolation transformer" clean up this "dirty power"?

I agree a hot attic is a poor place for the transformer.

.
 
081221-1002 EST
What is "dirty power" in this application?

On this project, any power upstream of the 15KVA XFMR is arbitrarily labeled "dirty" just for identification sake, and not a fundemental description of the POCO's power quality. No power quality assessments have been made.

How does an "isolation transformer" clean up this "dirty power"?

I am just a wireman, neither engineer nor salesman. If I could easily answer that question with lecture or charismatic conjecture, I would not have to crawl around in attics. The equipment was provided by the HO (who already owned the XFMR which I think should have been left to gather dust in his storage).

I agree a hot attic is a poor place for the transformer.
.

The architect and HO spec'd the location. The XFMR is NEMA 3R. The location was debated between 4 locations. Outside near service, outside near xfmr fed sub-panel, inside near sub-panel (in a hallway), or the attic.

I have seen several posts concerned about hot attic. This is Texas and it does get warm in attics here. Electrical rooms here get warm in the summer, too. There is an attic vent directly above the XFMR, so I'm not too worried about the XFMR combusting itself or anything around it. The XFMR looks hermetically sealed. What, me worry? Yes. I am more concerned with audible noise being heard in rooms underneath the XFMR.
 
081222-1010

jeremysterling:

What is an isolation transformer?

From Google:

An isolation transformer is a transformer, often with symmetrical windings, which is used to decouple two circuits. ...
www.en.wikipedia.org/wiki/Isolation_transformer
This produced a reasonable discussion.

A transformer used to reduce or eliminate noise and create the equivalent of a dedicated or isolated ground circuit. ...
http://es.tripplite.com/support/glossary/detail.cfm?mode=1&range=5
This description implies more than it should by the way it is worded.

The secondary windings of an iso-transformer have no physical electrical connection to the primary windings.
www.sbspowerquality.com/grap/pqglos.htm

An isolation transformer is not a solution for all types of power line noise. If you do not know from what you want to protect, then just inserting an isolation transformer in the line may or may not be of any use.

In a basic sense all transformers that have no low resistance connection between primary and secondary can probably be classified as an isolation transformer. Without a Faraday shield between the windings there may be a substantial amount of capacitive coupling from primary to secondary. An ordinary transformer is unlikely to have a Faraday shield.

If you connect any of the secondary wires back to primary side wires you may loose the isolation you thought the transformer provided. So how you connect the EGC to the secondary side may be important relative to solving whatever is the imagined problem. Common mode noise adds to both incoming lines. Ground path noise falls into the this category. If you run an ECG from the isolation transformer secondary center tap to the EGC on the primary side you may introduce a lot of noise that is on the primary side EGC. If instead the secondary EGC is run by itself all the way back to the service entrance ground bus some of this noise may be eliminated.

A normal mode noise to a large extent will not be removed by an isolation transformer. This is noise that is part of the incoming signal.

If you do not know what you want to protect from, then it is difficult to determine how to build the protection.

All electrical insulating materials have a maximum absolute temperature limit. A transformer when powered and loaded produces substantial heat internally. This heat input produces a temperature rise in the transformer from its outside surface. The amount of rise is relatively independent of the ambient temperature. If the rise is 80 deg with an ambient of 70 deg the internal absolute temperature is 150 deg. By independent I mean that if you raise the ambient that it does not have much of an effect on the rise. Thus, the same transformer loaded the same in an ambient of 150 deg will have an absolute internal temperature of 80 + 150 or 230 deg. The absolute internal temperature of the transformer is the temperature rise from internal heating plus the ambient temperature. The life of an electrical device above some value, maybe 70 deg F., will degrade more rapidly as the temperature rises. Therefore, a hot attic is not a good location.

.
 
Well #6 AWG would be considered 50A wire so #10AWG Copper would be good as an EGC. Are you running the secondary through non-metallic conduit? That would be a reason to run an EGC wire.
 
The heat in that attic will make this a poor and unsafe install. Even with the vent above it. Please make sure you use series wired smoke detectors with battery back ups.
Don't take this the wrong way, I do wish you well.:wink:
 
I would bet your homeowner is someone who thinks they know a lot more about electricity than they really do. Perhaps one of those stuck-up engineers that are always getting trashed around here, or a do-it-yourself person that has installed a few receptacles and therefore thinks they know everything.

Sadly, although I maintain that I am not stuck-up, I am an engineer that has installed a few receptacles. :) I don't think I would do something like this, but I wouldn't place any bets on it either. :D:D
 
081222-1010
If you connect any of the secondary wires back to primary side wires you may loose the isolation you thought the transformer provided. So how you connect the EGC to the secondary side may be important relative to solving whatever is the imagined problem. Common mode noise adds to both incoming lines. Ground path noise falls into the this category. If you run an ECG from the isolation transformer secondary center tap to the EGC on the primary side you may introduce a lot of noise that is on the primary side EGC. If instead the secondary EGC is run by itself all the way back to the service entrance ground bus some of this noise may be eliminated.
However under the rules of the NEC (with a couple of rare exceptions), if the secondary voltage is one that is required to be a grounded system, you are required to connect the primary EGC to the secondary grounded conductor. This is not a direct connection, but indirect as both the primary and secondary EGCs must connect to the transformer frame and case.
 
081222-1211 EST

don:

I was not addressing NEC issues, but rather what the noise problems might be.

The secondary side EGC must be connected to the secondary side center tap for it to perform its load side function.

It seems appropriated that the transformer shell be connected to the primary EGC. This would mean that capacitive currents from the primary via the core or enclosure would mostly flow thru the primary EGC. If a single Faraday shield existed, then it would go to the primary EGC. If a second shield existed, then the shield closest to the secondary would go to the secondary EGC.

Does the NEC require the secondary center tap, EGC, be connected to the primary EGC directly at the transformer, or can that secondary EGC run back to the main grounding point at the service entrance? If it can, then I believe it will introduce the least noise to the secondary side.

.
 
Gar,
My main point it is not possible to provide "isolation" and code compliance. The secondary ECG connection cannot be run back to the primary EGC at the service. Both the primary and secondary EGCs must be connected at the transformer.
 
081222-1915 EST

don:

What are the restrictions on the bond of the load EGC to the supply side EGC? Is it the distance between the two points? The length of the wire? For example could the bond be one hundred feet of wire wound into a coil with a powdered metal core and the two terminals of the coil 3" apart?

Suppose the inductor is allowed for this bond, then this alone will provide substantial impedance at high frequencies. To this could be added a supplemental grounding conductor and electrode.

In the application of the original post what is really expected of this "isolation transformer" and is it of any value? It seems the home owner needs to supply more information. jeremy may know but can not tell for proprietary reasons.

.
 
Gar,
My main point it is not possible to provide "isolation" and code compliance. The secondary ECG connection cannot be run back to the primary EGC at the service. Both the primary and secondary EGCs must be connected at the transformer.

If you read my OP, you see I am asking for guidance on tying the secondary's Grounding Electrode Conductor (not the EGC, I know where that goes) back to the structure's grounding electrode (at the service).

Quote: Pierre C Belarge: If there is no structural steel, which it seems there isn't, you will need to connect to the water pipe 5 feet of where it enters the building.

Thank you Mr Belarge, I'll inform the HO this is where I need to bring the grounding electrode conductor. I did not see a ground rod anywhere and I'd just as soon not drive one.

I understand at the end of the day all green, white, and bare wires in this structure will be electrically continuous.

"Step-up","step-down", "buck-boost", etc. do not apply so "isolation" won out. We can call it "the" transformer from here on out.

(gar, quote:) (the) transformer is not a solution for all types of power line noise. You are correct gar, the end objective of this install is to reduce power line noise.
 
Status
Not open for further replies.
Top