,

based on what you have said, you may need several types of ground isolators, but no matter which ones you use, get them from http://www.jensen-transformers.com/ I cannot think of a better quality product for the job, and if you need help Bill Whitlock is a ground loop isolation genius (took his class at CEDIA...WAY smart guy)

my .02

I agree with the above, check to see if the cable tv line isn't the cause by unhooking the catv feed to the system. Most often in my neck of the woods it is the culprit. If you find the source it can save you having to throw a bunch of money at the problem to cover all of the ground paths.

2nd that BUT he did say three phase and generator backup= potential NIGHTMARE.

I would sit down with the electrician and see if you can't get as much of the equipment on the same circuit first, same phase second and then take a couple of thousand dollars worth of Jensen transformers and start playing. Good luck.

that's what I was thinking when I read the post the first time.

I agree with everyone above.

1st. take off just the cable TV coming into the house. Xantech makes a ground breaker.

The electrician answer is always the best. It could be "Phase" and probably is.

Mike

,

By "not hooked up", do you mean the incoming coax is not attached to the cable box, or that the cable company hasn't connected a feed outside?

I'm always surprised when electricians sometimes give me the "Homer Simpson" blank look when I request outlets on the same phase.

Last week I watched very closely when one of out techs asked about phase in an existing Boardroom with an equipment rack and (3) remote displays on (4) separate outlets... same situation with (3) phase & a generator.

I'm not kidding - the electrician visibly flinched, blinked twice, then told us the phase couldn't possibly make any difference as long as all the outlets were all "hot".

I've run into some *very* sharp electricians, but I really think that some of these guys don't understand this.

Larry - isn't ground potential & ground loops a fairly basic concept? Is there a better way to communicate or phrase this?

Well, I wasn't going to comment, but since you asked...

No, asking to keep a group of circuits on one phase isn't a difficult concept, but I can understand why some electricians don't understand why it matters, and I'll admit to at least understanding why they feel that way.

Even many electricians I've worked with understand the mechanics of their work, but they have a relatively poor grasp of electrical theory. They tend to think that all electricity behaves like DC, such as when applying Ohm's and Watt's Laws to circuits.

There is a big difference between DC and AC when it comes to implementing impedances, because AC includes reactances (capacitance and inductance combined) as well as DC resistance; reactance and resistance combine to make impedance.

One of my best friends is an EE, and we have discussions that leave most people lost. We've delved into things like skin effect, bi-wiring, speaker-wire size, the everlasting RG-6 vs. RG-59 debates, etc., down to the molecular level.

As far as ground loops, we tend to think of all grounds as equal, so one circuit's EGC (equipment grounding conductor) should be no different than another. In theory, they should be, just as all neutrals "should" be. Also, power supplies "should" provide full isolation.

When it comes to 2-wire power cords, there should be no connection to provide a loop-causing current pathway that parallels the interconnect shields. Normally, the incoming power goes through the power switch to the transformer secondary.

It's like the isolation transformers that ground the center tap and have 60 volts to ground on both wires. Since the equipment's own transformer does the same thing, what difference does it make what the voltage-to-ground is on either end of the primary?

The secondary (center-tapped for a split supply) feeds a rectifier and then a filter cap(s) and voltage regulator circuit. Either one end of the rectifier's output is grounded (for single-ended supplies), or the center tap (for full-wave supplies) is connected to the chassis.

Most equipment keeps the power and signal grounds separate, except at a single point. This is to keep internal ground loops to a minimum. This gear usually has insulated-ground interconnect jacks unless there is no chassis conductor between the jack and the circuit board.

So far, there's no real parallel pathways for ground lops, so where do they come from? The easiest to isolate (pardon the pun) is an incoming cable-TV coax, a common problem because they're connected to earth (or should be) externally to the premises.

Now, if we have 3-conductor power cables, the ground loop is all but a certainty. We can try broken-shield interconnects and balanced connections. Still, internal equipment layout has something to do with it. Power and signal grounds can be kept isolated.

Okay, that's a lot of typing, but I'll close with restating that relative voltages on separate transformers' primaries ought not matter. Nor should phase-angle differences, because of the isolation and rectification of the power supplies, but somehow they can.

Multiple ground paths can add noise, to a point, after which they can reduce noise. If you have a multitude of low-enough-impedance grounds, the separate chassis effectively become one. On large rack-mounted systems, the rack itself can help or hurt with ground-loop-indiced noise.

You have to experiment sometimes. One solution does not work in all apps. That's why some "pro" gear comes with ground-lift switches on inputs and/or outputs. Okay, I'm really doen this time. My fingers are hungry, and it's dinner time. Ciao!

,

I'm convinced you understand the theory *way* better than I do, but my real-world experience is that the odds of having to spend time & money chasing ground loops has gone down since I started requesting same-phase circuits.



This is so common in our area. We have so many houses in the hills built on dry, rocky dirt that I don't think a ground rod driven in to backfill by the cable providers really acts as a ground. Bonded to *what*? I've measured some pretty high voltages between the CATV "ground" relative to the panel ground.

Only a little off-topic here, but a lot of our projects have only AC ground rod(s) driven into the same dry rocky backfill... no "Ufer" ground, with only plastic incoming water pipes, etc.

[Link: psihq.com]

Am I wrong in thinking these homes are basically relying on the ground at the power transformer?


OK, I understand that they "ought not matter", but they sure seem to. Especially on Friday afternoons before major holidays.


We had (3) of my best smarty-pants techs take 5 hours to chase down a nasty hum/buzz to a rack-mounted power amp. The actual fix took 10 minutes and a buck worth of isolated shoulder-washers from Middle Atlantic. Learn something every day.


Keep talking Larry - I love being educated on AC issues & appreciate the time. I do *not* want to make the electricians jobs any tougher than it already is by spouting bogus info or useless requests.

Barry

So, Bob, is this house built of vinyl sitting on a rock outcropping two hundred feet above the water table, or what?

Excellent points, everbody.

But consider something counterintuitive: add more actual GROUND grounding to the equipment (not to the power) and see what happens. The one home I have run into that had a generator, with the situation worsened by the new homeowner having his personal electrician remove it, had a horrible hum in their one system. It was not the cable ground. After some hours of mucking about, I told him to tell his electrician that the house did not have a good ground.

That got us nowhere. It should have. We could measure very little resistance from the panel ground to the water pipes, and he did not go any further than that.

I solved the problem eventually by connecting one particular junction of cable, satellite and components to a 12 gauge wire going directly through the wall to a metal water pipe. A garden hose bib, I believe.

I have also seen a multi-room, TVs in each room, system have bad hum on the video when it was partially assembled. When ALL the connections were made, the hum was gone! This argues for more ground connections, not less. In fact, Larry says (it WAS Larry, right?) and he is right that audio gear has signal grounds and power grounds and they are carefully tied together in one place. Early amps did very poorly with TVs because they used an audio-appropriate star grounding system that was susceptible to current coming in on the cable.

But these days most amps don't have those hum problems and don't have star grounds. Instead, they have gone in the other direction: instead of carefully devising and routing the grounds, just ground everything everywhere with a ground plane. This is the equivalent of spiderwebbing ground wires from each component to each other component! You can't do that, of course, but adding some grounds (that do not tie directly to power ground) will probably help.

That may even make the phase problem go away.

OK, but with poly/plastic incoming & internal water lines, where's the ground?

We see electricians lay a copper ground cable alongside the poly/plastic water line in the trench sometimes to compensate, but with dry rocky conditions, is that really an improvement?

Even assuming ground rods start off as effective when installed, where are you years down the road as they rust and bonding connections corrode?

So... how do you test & define a "poorly grounded" house?

Spending some time with the theory is important--which is why so many electricians have no clue. It's not merely an issue with performing the rather difficult mathematics--which most of all us would also fail-- but the concept. AC circuits are not affected by gound loops so electricans almost never have to deal with them. They have trouble understanding open neutrals and harmonics on 277v circuits. But I also spent 10 minutes with the EE specifying a #1 neutral for a 277v lighting control job many years back.

Ground issues are the result of eddy currents. What produces eddy currents? Almost anything. Sometimes what we see is the result of a poor ground but I think this unusual. More often it is the result of not having the many multiple gounds bonded. But as my arch conservative friends at Jensen Transformers explain balanced power is only a theoretical solution. The problem is that it is impossible to perfectly wind a wire around an iron core and if those windings in the transformer are imperfectly wound an induced current is the result

I am unclear why you think dry, rocky dirt is a poor conductor? There is no better conductor than "earth." You may need an 8 foot grounding rod but I have never heard--though that in of itself says little--about different types of soil conducting electricity better or worse. You need to go down into that "soil" 8-10 feet but I do not see why the contents of the soil would make ground less of a conductor. The purpose of bonding to metal plumbing pipes is precisely to take advantage of the fact that plumbing pipes by code are burried in the earth sufficiently deep to be used as effective grounding conduits for stray voltages.

How else are eddy currents induced? Most all grounding issues come via the power grid. Some are induced by the line amplifiers cable companies use to power their signals down those thousands of miles of coaxial cable. Some are the result of too many grounds not bonded together ( as in separate grounding rods not tied together ) and in theory some could be the fault of an improperly secured saftey ground in the electral panel. Actually, most electricans are very good at this and I cannot recall a signle instince where this has been the fault of a gound loop.

On more than a few occasions the problem is single ended circuits or poorly implemented grounding schemes by manufacturers. There is a reprint ( or used to be. I am not going to see if it's still on Jensen's web site. ) of a paper given at an AES convention of pin1 problems resulting in ground issues. The only fix for this is not to use that gear. That is hardly useful when you have gear in place that was manufactured with grounding issues. Certainly, different methods of grounding circuits can lead to eddy currents in a rack which is why we used to ( but I don't any longer ) use copper gouunding cable in racks that tied each piece of equipment to a single grounding point.

One of the real issues with ground loops is that, what ever the cause, they can spread by all sorts of methods thoughout the system. Interconnected subsystems and equipment will carry and the interconnected equipment often amplify those signals.

Most homes do not have multiple panels. Those that do often have them in the same machanical room. Only on commercial jobs and the largest homes ( over 15 thousand squar feet ) here in NY do we see decentralized subpanels. Here the stray voltage will be the result of poorly designed grounding scheme. The longer the distance from the bonded ground the more likely an eddy current will be induced. As this will have no baring on AC circuits electricans pay not mind to it--just as many of us pay little attention to wiring boilers or other arcane matters to which electricans need to be mindful. Their equipment does not amplify and when they are involved in jobs where this is important an EE has been hired as a consultant to define and put on a plan an appropriate grounding scheme.

The fact is that there is no simple method of discovering the cause of a gound loop. Certainly, experience will tell you to check the CATV feed as impoper equipment removing entirely the 60 Hz power signal from that line is often the cause of thise ground loops. More often the problem is in the wall and has little to do with how well safety gound is bonded to earth. The issue is not conducting a short to ground but small eddy currents that are the understandable result of 2 or 3 phase power conductors. These stay voltages may be just a few milivots but that is hardly enough to be conducted to saftey ground but certainly enough to create eddy currents or hum. The only way to locate the offending party is to pull the system apart cable by cable. Often, despite the price, the fastest method of solving the issue is to install an isolation transformer and go home. If the equipment we sold used balanced audio and video schemes we would have far fewer of these issues. But it costs more and even when Blue states ruled it was thought ( I suppose correctly ) that the additonal money for a balanced circuit would result in fewer sales in a consumer market where gound issues happen but not all of the time.

Anyone who has dealt with an electrician with respect to lighting control systems will tell you that most do not understand loads let alone AC vs DC electrial theory. Unfortunately, Larry is so atypical of what I run into with respect to electricans with respect to the amount and type of knowledge he possesses. Most electricans are simply not that smart. In fact, most are down right stupid. If they needed to know the sorts of knowledge that Larry has there would be a dramatic shortage of electrical contactors and workers. They know how to install madison clips, electrical fixtures, and wire a panel ( most days ). They do not need to understand dc theroy nor do they need to deal with ground loops as they don't encounter them often--save when they have anoying A/V guys on the job.

Alan