Iso trafo at sub panel garage studio

[Dan Elleson]

Howdy!

I live in a rural setting with sketchy power and am constantly looking for a way to improve my little recording studio in my garage. At the moment I get noise and interference from the house (something to do with a water pressure tank system) and would I’d like to isolate the studio from the house. I have a Tripp Lite medical grade 1800w box that I use to protect my console from power brown outs etc (roasted two power supplies in the past… argh). The Tripp Lite works well but I was hoping to have a similar concept wall mounted (inside or outside) that would isolate the the power pre-sub panel. I’ve yet to add up my kVa’s yet but prob something around a 5kva I’d guess.

Also if I go this route can I add an isolated ground post iso transformer?

Any tips?

Thanks,

Rob

20 years ago when I was battling with interference, I went down this same rabbit hole and came to the conclusion that I needed to have the mains earth /ground ripped up and done properly, or that I needed to isolate the power in my room and build my own interior system. I went on to work as a studio tech for over 10 years and learned a lot more about it, and actually that was a ridiculous conclusion. At least for a single room, all you need is for gear to be correctly interconnected, with thought for different ground potential where equipment is connected to power, and what type of ins and outs you're connecting, unbalanced or balanced, which units suffer from what design issues, etc.

The categorical go to solution for interference in your situation, IMO, is a central earth/ground bus for all your interconnects, which is conventional in properly build control rooms. With a few exceptions that require understanding and planning, you need to connect all the earths /grounds of the lines going to all your gear, to a single point in the central patch panel. It's more complicated than that of course, but that's the most categorical solution, and that's what I would do first, if you haven't done that.

Papers on earth /ground systems and audio interconnects by Bill Whitlock are very useful for properly understanding and solving these problems too.

Just my 2c

BTW don't ever just add a ground stake unless you've done the math on voltage gradients and EPR around your mains system and your building. You can blow everything up or get killed when lightning strikes otherwise.

 

[abbey road d enfer]

The categorical go to solution for interference in your situation, IMO, is a central earth/ground bus for all your interconnects, which is conventional in properly build control rooms. With a few exceptions that require understanding and planning, you need to connect all the earths /grounds of the lines going to all your gear, to a single point in the central patch panel. It's more complicated than that of course, but that's the most categorical solution, and that's what I would do first, if you haven't done that.

Every "categorical go to solution", is a simplistic answer to a complex problem.
The concept of star ground as a panacea to any audio problem including ingrown toenails is not realistic.
Star ground is a justified obligation in terms of safety, but it's not optimum in terms of audio interconnects, where gospel is ground follows signal. It's not easy to achieve in a complex set-up; that's why some of the smartest minds advocate mesh connection, when others propose a separate earthing system similar to those used in medical equipment. They're often right, both.

 

[Dan Elleson]

Every "categorical go to solution", is a simplistic answer to a complex problem.
The concept of star ground as a panacea to any audio problem including ingrown toenails is not realistic.
Star ground is a justified obligation in terms of safety, but it's not optimum in terms of audio interconnects, where gospel is ground follows signal. It's not easy to achieve in a complex set-up; that's why some of the smartest minds advocate mesh connection, when others propose a separate earthing system similar to those used in medical equipment. They're often right, both.

Totally. I agree with you, simplistic schemes exist for good reason tho. I don't think saying to hire an engineer is much help, and outside of pointing to Whitlock's papers, what else can be communicated in short time on a forum. I dunno. Just felt the need to point in the general direction, rather than let someone rip up their mains.

 

[Brian Roth]

Somewhat on/off topic are the extreme "grounding" systems required at a radio or TV station location which has a co-located high power transmitter and a tall tower for the antenna. In my region, the metal towers are often 1500 to 2000 feet in height above the dirt. It becomes a crazy world with thick/wide copper straps and/or fat braid "running everywhere" to:

1. Protect the facility's equipment against lightning strikes.

2. Keep RFI out of the audio (or video) equipment.

However....That's all above my pay grade...lol.

Bri

 

[FIX]

Danger Will Robinson, free internet advice. Caveat lector: I describe rural US mains power.

When talking about house power from a utility pole (in the US) that is considered single phase (just one leg of the three legs phase spread 120' apart on high power lines).

House power is considered 3 phase delta or 3 phase Y prior to the pole transformer that brings in to a street or neighborhood. At that point, it's called a single phase, or double pole, which may be incorrect, but in the home's panel it's called 2 phases, as you have 2 opposing phases. If one phase of the delta was used without the step down transformer, the power in a house would be 208. The neutral is usually from the center of the Y, which becomes your neutral. If you are being fed from a delta configuration, there is no neutral. The step down transformer may have a center tap that is connected to the center of the Y, or may be reliant on a balanced load at the house end. If the Y configuration is used, their is 208 between any two phases, but 120 between a single phase and neutral.

You never know if the power feeding the transformer to your house comes from a 600 volt 3 ph line, a 34,500 volt line, or somewhere in between.

My point is that the power coming in is not only dirty, but shared by several homes, and the neutral moves around depending on the different loads. This is the source of literally all gorounding issues, isn house wiring was not designed for audio, only lights, a toaster and a fridge.

At a famous studio in LA, the back studio had power shared with an apartment building, where the building actually took the power from their side of the pole transformer. You used to have to rotate the 24 track to null the hum. I've also seen a studio in Belgium that has it's own HV feed from the power company, then a large cage and a step-down. You could hear the noise change as the city closed up businesses around 6PM. They refused to install ISO transformers as the facility guy claimed that it was "wired correctly".

Generally a center tapped step down transformer winding with opposite polarity single phase power on the two ends of that winding. The center tap gets bonded to neutral in the panel and earth (cough) ground... The earth ground is probably of more interest to lightning than audio circuits. EGC or "safety ground" has a primary function of blowing fuses or tripping breakers if the hot lead goes rouge and energizes a chassis or whatever.

The center tap is not only shared among all homes connected to the transformer that takes the 3 phase power and converts it to dual phase (or double pole) 240, either on a pole or in the ground, but the MGN grounding shares the center tap with the delta center. Thats where the problems start. If each house had a transformer, it wouldn't be so bad, but thats not the case unless you live way out...

? indeed I replaced my water inlet pipe with PVC after it sprung a leak.

You many times see a ground wire from the panel to a water pipe.

There is probably more than one kind of panel... my house still has fuses.

Yikes.

Be very careful messing with mains power, people can and do get killed.

JR

Thats why I said to use an electrician.

 

[abbey road d enfer]

The center tap is not only shared among all homes connected to the transformer that takes the 3 phase power and converts it to dual phase (or double pole) 240, either on a pole or in the ground, but the MGN grounding shares the center tap with the delta center. Thats where the problems start.

It shouldn't be a problem with proper grounding.
The disturbances on neutral are not relevant if the whole system is referenced to it. That what it should be with neutral connected to earth at the panel and a good earth continuity; the system should be as clean as if it was flying at 30 00 ft, because there would be no common mode noise. Unfortunately, earth continuity is often non-existant. Many electricians consider a system is good when the lights light up and the fridge is cold.
I believe the problem at case is one of loss of line regulation. An isolation xfmr would not fix it.

 

[JohnRoberts]

For human safety I advocate use of GFCI outlets these will protect even without ground wires. My house uses 2 circuit wiring so no effective EGC or safety ground anywhere.

Years ago I has some excitement (literally) when my hot water heater rusted through the heat element housing and energized my hot water supply. I felt a tingle while testing the water temp with my finger before climbing into the shower. After replacing the killer hot water heater I bonded the new one's chassis to my fuse panel with some decent gauge wire.

Prior to hurricane Katrina I had my own house transformer, but that ended up down in my yard after a huge cottonwood tree pulled down my power pole and lines. I recall the utility workers dug up and carted off a bunch of dirt from my yard where the transformer fell and leaked oil. Now I share a single transformer with two other neighbors. I hadn't thought about it but indeed I now share a common neutral with both neighbors.

Several years ago when I replaced my in wall kitchen oven I tried to hire a local electrician to wire in a proper service with EGC ground lead between my panel and oven. After he declined to even answer my phone calls I wired in the oven myself using the original wiring (the oven hook up instructions showed a recommended wiring digram for that).

The voltage up on the power poles in the street near my house is 7kV or 12kV. There is sub station a mile+ up the road with much higher voltage coming into it ( an educated guess based on spacing of the 3 wires and height of the transmission poles.) I also notice that power poles include an extra wire running below the hot lines, earth grounded I am pretty sure).. The wires coming from the transformer to my house is only 230VAC.

Years ago a lawyer tried to hire me as an expert witness to testify that voltage from a 230V house drop arced over to his client's metal ladder and shocked him. I told the lawyer that his dumb ass client must have touched the wire because 230V wouldn't jump across a spark plug gap, let alone feet to an aluminum ladder. I didn't get hired.

Be careful about listening to free advice (even mine). I hope you have better luck than I did hiring an electrician.

JR

 

[FIX]

I though they spaced those wires apart so birds wouldn't become fuses... LOL

Their was a study by the feds on the harmful emission of magnetic fields from power lines, and of course the whole study said their was no possibility that power lines could cause problems with people's health BECAUSE THE MAGNETIC FIELD FROM THE DC HIGH VOLTAGE WIRES POSED NO RISK... I kid you not.

 

[JohnRoberts]

I don't think the magnetic field is the major risk from extreme high voltage.

Ionizing radiation OTOH can affect human cells, so keep sticking that cell phone in your ear.

===

We routinely see squirrels blow the fuses protecting local mains high voltage power distribution. Actually not much to see of the squirrels after such events but they conduct enough current long enough to blow the fuses.

Since I am prepping for hurricane Ida losing electrical power is a topical subject. I just put several days worth of beer into the refrigerator.

JR

 

[abbey road d enfer]

Risks related to high voltage distribution is a subject of controversy.
VHT DC lines are balanced, which results in very small electrostatic field since at normal distance the field emitted by one conductor is compensated by the other.
In virtue of the proximity of the conductors compared to the length, the magnetic field is also very small.
There is no scientific proof of the dangers of 50/60 Hz AC radiation, although it is undeniable that MHz and GHz radiations can generate enough Joule effect to fry animal cells.

 

[terri faron]

https://www.nfpa.org/codes-and-stan...ds/list-of-codes-and-standards/detail?code=70
and not forget Isaac Physics.

https://www.aes.org/standards/comments/drafts/aes48-xxxx-190121-cfc.pdf
https://www.aes.org/e-lib/browse.cfm?elib=120
use a separate duct for the earth distribution to minimize the noises.

 

[ccaudle]

use a separate duct for the earth distribution to minimize the noises.

Your profile does not show a location, but I assume what you refer to as "duct" is what would be called electrical conduit in North America.
It is not very practical in domestic construction, almost all houses here are build with non-metallic insulated cable, usually referred to by the most popular brand name, Romex.
Non-metallic actually isn't too bad, because the safety ground connection is centered in between the line and neutral conductors, so the magnetic fields mostly cancel and don't induce much voltage in the ground conductor. Apparently the aluminum metallic clad (armored cable) is slightly better, but I am not sure of the mechanism there. The armored cable can be used in US domestic wiring (I guess technically conduit can be used, but it is not economically feasible in most wood frame houses). Both are significantly better than wire pulled in conduit without any twisting of line and neutral together.

The measurements and descriptions are in a paper authored by Bill Whitlock and Jamie Bell in AES preprint 8234, but Jensen Transformers has provided a copy on their web site if you do not have an AES membership:
Ground Loops: The Rest of the Story

 

[abbey road d enfer]

use a separate duct for the earth distribution to minimize the noises.

I understood that as "don't put audio cables in the same conduit as mains distro", but now I have doubts.
I don't see how separating the ground wire from the two others can result in noise reduction. In normal regime, there is no current in the ground wire.

 

[shabtek]

duct for communications cable runs IME is ferrous wireway , 60HZ power runs in emt/conduit.

...Both are significantly better than wire pulled in conduit without any twisting of line and neutral together.

huh? @ what HZ? better how?

Non-metallic actually isn't too bad, because the safety ground connection is centered in between the line and neutral conductors, so the magnetic fields mostly cancel and don't induce much voltage in the ground conductor.

magnetic fields?

(conduit is superior in every way except cost)

I am with Abbey; likely you have regulation issue.

+1 on GFCI
120vac supply (black in US) and return (white in US) currents should be same

 

[ccaudle]

I don't see how separating the ground wire from the two others can result in noise reduction.

better how?

Obviously neither of you guys read the paper I linked. Whitlock and Bell spent 9 pages explaining it all, with pretty pictures and everything.

magnetic fields?

Yes, current flowing in wires generates a magnetic field. Another wire placed in proximity to a changing magnetic field will have a voltage induced. Current flowing in line and neutral can induce a voltage in the ground wire if the spacing from line to ground wire and neutral to ground wire is not exactly matched. If you have three-pin power cables on equipment, that voltage on the ground wire is a voltage between chassis, which will cause current to flow through the cable shields.
I don't think that causes problems in most domestic systems, or even in most commercial installations, but it obviously can in system with large current flow (see the end of section 4 on page 8 of the linked paper). The example there was of a system with 650mV noise between the console shield connection and the amplifier shield connection. A change in the geometry of the power wiring dropped that to 3mV.

 

[abbey road d enfer]

Obviously neither of you guys read the paper I linked. Whitlock and Bell spent 9 pages explaining it all, with pretty pictures and everything.

I did, and I may be blind, butnowhere I see anything that says performance is improved by separating the groundwire from the other two.

Yes, current flowing in wires generates a magnetic field. Another wire placed in proximity to a changing magnetic field will have a voltage induced. Current flowing in line and neutral can induce a voltage in the ground wire if the spacing from line to ground wire and neutral to ground wire is not exactly matched. If you have three-pin power cables on equipment, that voltage on the ground wire is a voltage between chassis, which will cause current to flow through the cable shields.

Actually ground loop current depends very much on the area of the loop (and the resistance of the wires, of course). That's why hum due to ground loops can be eliminated by lifting the earth connection. I don't recommend it at all, for obvious safety reasons but we all know it works.

I don't think that causes problems in most domestic systems, or even in most commercial installations, but it obviously can in system with large current flow (see the end of section 4 on page 8 of the linked paper). The example there was of a system with 650mV noise between the console shield connection and the amplifier shield connection. A change in the geometry of the power wiring dropped that to 3mV.

Yes that was obtained by twisting live and neutral, not by separating earth.

 

[Dan Elleson]

Somewhat on/off topic are the extreme "grounding" systems required at a radio or TV station location which has a co-located high power transmitter and a tall tower for the antenna. In my region, the metal towers are often 1500 to 2000 feet in height above the dirt. It becomes a crazy world with thick/wide copper straps and/or fat braid "running everywhere" to:

1. Protect the facility's equipment against lightning strikes.

2. Keep RFI out of the audio (or video) equipment.

However....That's all above my pay grade...lol.

Bri

Exactly, and it's all very carefully calculated by engineers, taking into account those strikes, rising ground potential, current paths, etc. Adding ground stakes naively is dangerous.

 

[ccaudle]

I did, and I may be blind, butnowhere I see anything that says performance is improved by separating the groundwire from the other two.

It follows from the general principle that magnetic flux due to current in a long wire decreases proportional to distance, and therefore the voltage induced in a second wire will be inversely proportional to separation distance.

Actually ground loop current depends very much on the area of the loop (and the resistance of the wires, of course).

That is true for magnetically induced current, but in this case the magnetic induction is within the power wiring, which you cannot control easily. That loop area is set by the geometry of the power wires, and once the voltage is induced between two ends of the power feed, it is straight Ohm's Law, there is a voltage difference between two chassis, the current flowing will be determined by the resistance to that flow. That current will be flowing in the shields of the signal interconnects, so for unbalanced connections you would want the resistance to be a small as possible to reduce the voltage in the signal returns, and for balanced connections you would want the geometric symmetry of the twisted pair within the shield to be as good as possible so that any voltage induced into the signal conductors by that current flow is as fully common mode as possible (very little common mode to differential mode conversion; there will always be some small amount due to parasitic capacitance imbalance).

Yes that was obtained by twisting live and neutral, not by separating earth.

"On the hunch that this voltage difference was due to
magnetic induction in the power interconnect wiring,
I asked the staff to twist, as tightly as possible, all the
current-carrying wires (3 phases and neutral) and lay
the safety ground wire next to the bundle a few
inches away."

Twisted line and neutral, and increase separation to the ground wire.

 

[abbey road d enfer]

It follows from the general principle that magnetic flux due to current in a long wire decreases proportional to distance, and therefore the voltage induced in a second wire will be inversely proportional to separation distance.

OK, got it.

 

[JohnRoberts]

That may all be true and have nothing to do with the OP's noise issue.

It is the nature of web discussions for people to tell you everything they know, not necessarily answer your question (me included).

Work to understand what your specific problem is, don't just shotgun a bunch of possible remedies.

JR

PS: Bill's paper on ground loops is worth reading, not saying that is your problem.