Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Phillip Graham]

bobkatz wrote on Wed, 15 December 2004 16:09

Agreed, but the neutral is ALWAYS grounded here in the U.S. except for certain types of hospital installations where floating is permitted. There are advantages to that, also. The safety ground should also trip the breaker.... Please let me know if any of this explanation does not make sense or does not seem as ideal to you as it could be! Thanks, BK

Bob, did you read my reply above Sahib.  It speaks to both issues.

Floating would ONLY be allowed if the device has no ground (is double insulated).  If you float the neutral with respect to a local ground, and somehow a connection between them develops (a fault) the potential on the ground can come up quickly because of the ground resistivity and V=IR.

I'm honestly not sure if your proposed setup will meet code, but it will be safe if you follow my additional suggestions.  Good luck finding someone in Orlando who knows the code well enough to say for sure.

[bobkatz]

[quote title=gtphill wrote on Wed, 15 December 2004 14:43]

bobkatz wrote on Wed, 15 December 2004 10:30

Glad to help.  This sounds pretty good with two minor caveats.  One is to insure that you can't accidently tie the grounds from the two systems together, either by clearly using orange outlets, or removing ones referenced to the original ground.  Obviously, because you know what you are doing this won't be an issue, but would not be idiot proof in a general installation.

Fortunately, I've built and installed isolated ground studio systems in the past, and I insist on measuring opens with an ohmmeter prior to doing the last bondings! It is definitely not foolproof. There are some models of IG outlets that have  jumper in them that has to be removed, and can screw up your whole day if you do not! There are some models of buss bar that appear to be floating but really cheat to the chassis by some silly route, so I fully follow what you are saying. Thanks.

Quote:

The second point is that I would use GFCI breakers on the secondary side of the isolation transformer.  You are relying on your new ground rod (bonded to the center tap) to act as a current sink in case of a short.

Why? I'm bonding the neutral to the ground inside the new sub box, so similar to as you described to me on the phone, the neutral is the one that carries the return current for the hot phase, and really ensures we'll blow a breaker, and the ground is really a safety. Anyway, no harm in using GFCI breakers as long as I can keep my clean ground clean and isolated from the dirty ground of the box.

Quote:

a building.  Subpanels, including temporary power distribution seen at concerts, is to keep neutral and ground isolated.

But I'm breaking this rule because the sub panel will contain an iso transformer, and therefore I can bond my NEW floating neutral to the ground that I also create within this new box, eh????

Comments greatly desired. I'm putting it into my own FAQ for my own use as well. Thanks, Phil.

[Phillip Graham]

bobkatz wrote on Wed, 15 December 2004 16:26

gtphill wrote on Wed, 15 December 2004 14:43 The second point is that I would use GFCI breakers on the secondary side of the isolation transformer.  You are relying on your new ground rod (bonded to the center tap) to act as a current sink in case of a short. Why? I'm bonding the neutral to the ground inside the new sub box, so similar to as you described to me on the phone, the neutral is the one that carries the return current for the hot phase, and really ensures we'll blow a breaker, and the ground is really a safety. Anyway, no harm in using GFCI breakers as long as I can keep my clean ground clean and isolated from the dirty ground of the box.

If the neutral should come unbonded to your ground, a standard breaker trips only if the resisitivity of the ground rod is low enough to sink the trip current of the breaker.

If however there is only a single ground, and that ground is bonded to the non-isolated neutral at the main panel, then then you are relying on the low resistivity of the power company's neutral to provide a sufficient current path to trip the (standard) breaker should something come unbonded.  The power company's neutral returns to either a pole transformer or load station, and should have a lower resistivity that your local rod.
We won't even touch the issue of voltage differences between the power company ground and the service entrance ground...

The caveat to ultraquiet power with the single ground point method is you are now relying on the integrtity of the zero potential reference at the primary building ground (+ your 100ft of wiring) to sink neutral "trash" currents.  Great if the neutral is a superconductor, but somewhat less than great with real wire.  NEC is about safety, not ultra-clean audio.

Your proposed second ground rod configuration truly separates the neutral (ignoring any galvanic coupling in the soil), but then needs a GFCI-type breaker to insure the resistivity of the soil doesn't prevent enough current flow to trip a standard breaker in the case of the neutral becoming unbonded.

Quote:

But I'm breaking this rule because the sub panel will contain an iso transformer, and therefore I can bond my NEW floating neutral to the ground that I also create within this new box, eh????

Yup.  This is exactly what the power company is doing outside your house, with the added assurance of a low resistivity ground plane that they test in some manner.  If you could insure the resistivity of the new ground, you'd have no issues with standard breakers.  GFCI is just covering your butt.

For an object lesson, go ahead and short your hot and ground on the secondary side of the transformer, with a breaker of course, both with and without the neutral bonded.  Standard breaker should trip in the bonded case, and might trip in the unbonded case.  If it does trip in the unbonded case, you have a good low resistivity ground.  Just don't pee on the ground rod during this process

This whole process of a separate "service" panel and new ground rod is legal under the NEC if the panel is located on a separate structure (e.g. guest house), even though it's technically a subpanel of the main house.  I'll dig up the section that's in.

Quote:

Comments greatly desired. I'm putting it into my own FAQ for my own use as well. Thanks, Phil.

No problem.  I'm not really breaking new ground here, just apeing my teachers.  Once you grasp the take home message of my original post, and realize you can set up an e-field potential without a current having to flow, you've got 90% of it.

[John Klett]

Quote:

Did you find the isolated ground bar and the means to mount it at Home Depot also?

nope - but I did't look.  I go to the nearby electrical supply place periodically to buy spools of VU-TRON portable cordage, isolated ground bars, just plain old ground bars, various connectors for large wire, twist locks and other stuff I burn through when doing work for people.  I just have that stuff.  Home Depot MIGHT have isolated ground buss bars...  don't know.

[bobkatz]

gtphill wrote on Wed, 15 December 2004 16:53

Just go ahead and short your hot and neutral on the secondary side of the transformer (with a breaker, of course).  If it trips a standard breaker, then the resistivity of ground is low enough.  If it doesn't, then use GFCIs.  Just don't pee on the new ground rod during this process

Don't you mean "go ahead and short your hot and ground on the secondary side of the transformer?" Basically, we're testing the integrity of the safety chassis grounds in our equipment, right? So, I short hot to ground at some distant point in the studio, the current runs back through the 12 gauge green wire, which goes back to the new box, where it is bonded to the neutral and also goes to the ground rod. So it is the bonding point there to the neutral and the current carrying capacity of the 12 gauge wire that I'm really testing, not even current to ground. The ground cable is just supplying a potential, and in reality, even if it is high resistivity and in an arid climate, we're still safe. Follow?

BK

[Phillip Graham]

bobkatz wrote on Thu, 16 December 2004 11:19

gtphill wrote on Wed, 15 December 2004 16:53 Just go ahead and short your hot and neutral on the secondary side of the transformer (with a breaker, of course).  If it trips a standard breaker, then the resistivity of ground is low enough.  If it doesn't, then use GFCIs.  Just don't pee on the new ground rod during this process Don't you mean "go ahead and short your hot and ground on the secondary side of the transformer?" Basically, we're testing the integrity of the safety chassis grounds in our equipment, right? So, I short hot to ground at some distant point in the studio, the current runs back through the 12 gauge green wire, which goes back to the new box, where it is bonded to the neutral and also goes to the ground rod. So it is the bonding point there to the neutral and the current carrying capacity of the 12 gauge wire that I'm really testing, not even current to ground. The ground cable is just supplying a potential, and in reality, even if it is high resistivity and in an arid climate, we're still safe. Follow? BK

Yes Bob, that's correct, I was typing too quickly.  If your neutral to ground bond is good, then you are set.

I edited my original post to clearly state that I was talking about the worst case situation of a short with neutral somehow unbonded to ground, and fixed the error you pointed out.

Accidental unbonding of neutral and ground (at the primary service panel) is surprisingly easy to do in portable power applications, but that's probably not a concern I needed to mention for the situation at hand.

[Phillip Graham]

So, the people that are still awake reading this thread are probably wondering "why can't we bond ground to several points along a neutral?"

The answer is that would connect the ground run in parallel with the neutral, causing the ground wire to carry some of the neutral current (I), and therefore creating a voltage drop along the ground wire by V=IR.

Ground wires are only meant to carry a fault current (e.g. an internal short), and the NEC doesn't allow them to carry any "normal" current. From an audio standpoint this is good, as the ground wire will be at the same potential everywhere if I in V=IR is zero.

[Barish]

ssltech wrote on Wed, 15 December 2004 16:44

So you can run 220V using two lives, with no neutral requirement, but a part of the code (like "the regs" in UK house wiring) requires that all live conductors pass through steel conduit, and that the conduit must be grounded... Keith

No wonder. I don't know if any of you has ever seen two live phases collide. You see a big blast and then the fire running down the wire with 200 mph and no fuse blows while this happens.


If it happens in your hands then you may look for a plastic surgeon if you are still alive of course.



The subject moved on quite fast while I was trying to sort out CaPE II stuff over there in Marsh but I'll reply to Bob's question about what advantage an online UPS can bring into an audio setup.

bobkatz wrote on Sun, 12 December 2004 03:27

You'd have to prove to me that there is an "audiophile-quality" UPS. Plus, those things make lots of acoustic noise. I'd have to put the UPS that's on the analog leg out at the breakers near the Iso transformer. The whole purpose of the Iso transformer is to reduce common mode noise coming in, so I'm not sure why a UPS would be needed to further reduce noise.

A true online type UPS is totally isolated from the mains and ıt regenerates a pure sinewave from battery so it provides maximum isolation from any kind of noise and interference coming through the mains. Unlike the line-interactive UPS' which only kick in if there is a brown-out or black out, at which point it actually causes clicks and pops into the audio equipment, a true online type UPS is always on and its output is in no way linked to the voltage at the input.

There are many applications of this in recording studios and it is proven to bring the best results among all. We don't use one here at the moment because our recording environment is not that critical, but I've seen the results it yields and it really does make a huge difference. Since noise is a very important issue to deal with in mastering environment I suggest you give it a try (with an Online type, and not any other type) and see the difference before you decide.

A further reading about this subject  and some feedback on case studies is here from a very experienced respectable AE/systems integrator that I know. It's a shame he doesn't hang out here in PSW (as far as I know).


Regards,


Mahcem

[bobkatz]

Barish wrote on Thu, 16 December 2004 15:45

A true online type UPS is totally isolated from the mains and ?t regenerates a pure sinewave from battery

Hi, Barish. Is the harmonic distortion of the models you are used to less than 3%? I'm really skeptical that the typical "inverter" circuit is anything but a glorified square wave, but I'm prepared to be a believer. I HAVE encountered some very "purist" AC generators of this sort, but they are VERY VERY expensive. Make me a believer  


BK

[Barish]

Okay Bob, see what I can do. The Director General of APC's Turkey Subsidiary is my classmate from technical college years back in Istanbul, so I'll talk to him to provide me with the scientific documentation. For all the other audible evidences, you need to hear the difference yourself

Gimme a few days, he's a busy man, he can't say no to me but then he ends up working until 2am in the morning.

[bobkatz]

This topic (thread) has been a world of wealth for me!

Now I realize I have neglected to ask for your opinions on the topic of induced and mechanical noise. The transformer will be, let's say, 2 to 4 kva, and mounted outdoors inside of a standard metal (steel?) breaker cabinet. The cabinet will be affixed to the concrete block wall of my house. I have the choice of putting it anywhere along that side wall, but the best place for a perfect "star center" and minimal loop area happens to be directly opposite my 1/2" analog recorder and sensitive analog preamps.

So, there's the question. I'm concerned about the noise issues. If I hear the transformer physically vibrating through the block wall or if it puts hum into my low impedance tape heads or the input transformer of the MR70 I will definitely cry. I know the safe thing to do is mount the transformer farther down the wall opposite the next room. This means more conduit, greater loop area and so on. Am I being over cautious here?

BK

[Phillip Graham]

bobkatz wrote on Sat, 18 December 2004 14:35

So, there's the question. I'm concerned about the noise issues. If I hear the transformer physically vibrating through the block wall or if it puts hum into my low impedance tape heads or the input transformer of the MR70 I will definitely cry. I know the safe thing to do is mount the transformer farther down the wall opposite the next room. This means more conduit, greater loop area and so on. Am I being over cautious here? BK

Yes.  Move the xformer.

Stucco'ed 8" block, furring strips, and 1 layer of gypsum (presuming your house is like the ones in Wekiva) is going to have STC in the low 50s, if you're lucky (and don't have any windows).

On a related topic, do you know the NC or RC specs of your listening space?

[bobkatz]

gtphill wrote on Sat, 18 December 2004 17:44

Yes.  Move the xformer. Stucco'ed 8" block, furring strips, and 1 layer of gypsum (presuming your house is like the ones in Wekiva) is going to have STC in the low 50s, if you're lucky (and don't have any windows). On a related topic, do you know the NC or RC specs of your listening space?

Sonufagun, a custom noise rating from Phil! I'll make sure the transformer is at a distance. The neighborhood is in a cull de sac dead end. No traffic. This was all foreseen by me when I bought the house.

It's dead silent 99.99% of the time in the mastering room unless the Fedex truck drops in outside. I measure better than NC30 with my Audio Toolbox. You can hear a pin drop in here, no fans anywhere either.

BK

[Phillip Graham]

bobkatz wrote on Sat, 18 December 2004 22:04

gtphill wrote on Sat, 18 December 2004 17:44 Yes.  Move the xformer. Stucco'ed 8" block, furring strips, and 1 layer of gypsum (presuming your house is like the ones in Wekiva) is going to have STC in the low 50s, if you're lucky (and don't have any windows). On a related topic, do you know the NC or RC specs of your listening space? Sonufagun, a custom noise rating from Phil! I'll make sure the transformer is at a distance. The neighborhood is in a cull de sac dead end. No traffic. This was all foreseen by me when I bought the house. It's dead silent 99.99% of the time in the mastering room unless the Fedex truck drops in outside. I measure better than NC30 with my Audio Toolbox. You can hear a pin drop in here, no fans anywhere either. BK

NC30 in a house, that's pretty good, and better than I expected.  What version of the audio toolbox?  Have you done an RC?

Your STC is probably only still in the mid 40s, so be careful with the xformer.

[bobkatz]

gtphill wrote on Mon, 20 December 2004 13:48

NC30 in a house, that's pretty good, and better than I expected.  What version of the audio toolbox?  Have you done an RC?

There are occasional little hits if a car goes by, but as I said, cars don't go by that often. I have a 47 dB custom made door and there's only one window. All the noise comes in through the window, though only the low freq's as it's a double sealed kind of thing for those Florida air conditioned houses.

What's an RC? Version of Audio toolbox? You mean they got the NC wrong in an early version?

BK


Your STC is probably only still in the mid 40s

[/quote]

Or worse when you're talking mounting the transformer physically to the outside wall and it's single-wall construction. I'll just mount the transformer 12 feet further down opposite the next room. The loop area for the 110 volt conduit distribution will be a bit higher but I guess no big deal.

BK