1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

Getting rid of ground loops: What is the best option?

Discussion in 'Sound Science' started by chewy4, Nov 2, 2012.
1 2 3
5 6
  1. oloendithas
    As I were saying "on all power plugs except the PC one", so all other devices are grounded trough the PC.
  2. Strangelove424
    I'm not sure what you mean by "grounded through the PC". Because I see this as such a big safety issue, I am not comfortable waking away with such an ambiguity here, saying to myself "well, he'll figure it out." So I will explain myself more fully, to communicate the relevant dangers better.

    The only components grounded through a PC (chassis) are the motherboard, cards, drives, etc. The PC power supply is itself grounded by its ground connector plug and the wall socket's ground. The ground of your house's wiring can become "noisy" as different devices compete for the same ground potential and create a "ground loop". Anytime a ground connector is isolated with black tape or cheat plug, any device downstream of that plug is in danger of electrocuting you if it short circuits. As soon as you turn it on or plug it in, you will become the ground and complete the circuit. Circuit boards made with cheap solder are common even in simple household products. One failure could be life threatening. As detailed in the safety section of the Wiki article I linked, a professor at Cleveland State electrocuted himself with a faulty fluorescent lamp.
  3. pinnahertz
    In an effort to keep a balanced perspective...

    1. While theoretically possible, if the house ground is good, or even fair, it is a low impedance path to earth ground. For that to become "noisy" a couple of things have to happen. First, there must be significant current carried through the building ground to earth. That's an abnormal condition that should not exist, and would generally be caused by a serious fault. Second, the impedance of the building ground would have to be high enough for the voltage drop across it to be a significant. If not, it's just a low impedance path to ground. The idea of a noisy ground is often over-blown. I've only actually seen the problem once, and that was in a high-rise office building, and it wasn't so much that the ground was noisy as it was just higher impedance that was usable for audio. We had to install a separate audio ground system as a result.

    To be fair, many older home and building grounds are not very good, but by that I'm more referring to the ground system's ability to sink a surge from a power line or lighting strike, a very different problem. This can be tested by an electrician, or just install a new ground rod, which are cheap and you can work out your own frustrations pounding it 6' into the earth if you like.

    2. The statement is technically correct, but alarmist in nature. Again, something has to go seriously wrong, on the order of the chassis having enough voltage on it to be lethal. That condition, ground or not, would cause many other problems, and is therefore designed out as a probable fault condition. Technically possible, highly unlikely.

    3. This is actually incorrect. Assuming the worst fault condition of the hot wire shorted to a metal case (again, something that is carefully engineered to be as much an impossibility as practical), two other conditions must be present for the situation to become lethal. First, you have to make good electrical contact with the energized part. That means no insulating paint, and moist skin. Dry hands are very, very poor conductors, like in the meg-ohms, which makes it pretty hard to get lethal current flowing through them. Wet or sweaty, that's a different story. And the second critical part: you have to provide a path through your body to neural or ground, ideally one that passes current through a critical organ like the heart or brain. Standing on a carpeted floor in shoes and touching the hot wire may tickle you, but you certainly won't be electrocuted to death. Standing barefoot on a basement floor...that would do it, but who does that? When you combine the un-likelihood of hot chassis failure with touching with a body with a good low resistance path to neural or ground, you come up with a very, very low chance of this happening.

    4. Um...cheap solder? Sorry, no such thing. If it's going to work as an electrical connection you don't have a lot of choices for the solder blend, basically either the old 60/40 lead-based type or the new lead-free type. If you don't use that, you won't have a reliable connection, hence manufacturers don't use "cheap solder", especially on circuit boards which are usually flow-soldered or wave-soldered. You have to use the good stuff or the process doesn't work at all.

    5. The Wiki article cites a professor who was electrocuted pugging in a florescent lamp. "Electrocuted" does not mean "killed", though it can. The article also cites 9 deaths from audio equipment, presumably shock. Lightning strikes kill 40-50 per year. You stand a better chance of being killed by lightning than you do being killed by electrical shock from an ungrounded and faulty appliance.

    Don't get me wrong here, I'm not advocating cheater plugs for anything but diagnostics. I'm not suggesting these kind of faults can't happen. But let's also be realistic. For something like 1 in 10M, where a ground fault coincides with a human with low resistance skin and a good contact to neutral/ground, there may be an issue. Lets not go nuts. And ground noise is pretty rare. You can make a much bigger difference in your life safety just by choosing not to drive when it rains, and fastening your seat belt and making sure your air bags work. Much bigger. Like several orders of magnitude.

    Ground loops have nothing to do with an electrical fault, and ground loop current is microscopic compared to a ground fault. They aren't the same thing, and must be handled completely differently. Ground loops are caused by leakage of voltage to something that's not supposed to have it, usually the result of poor design, though occasionally a component failure. Since unbalanced audio uses ground as a reference, If the audio ground has a high impedance, or isn't even really a true ground, thus having some form of voltage variation on it, that voltage becomes the new reference, acting as a bias on the desired audio signal. If the shield of a cable connecting two devices is carrying a leakage current, that can be coupled to the signal conductor too. Sometimes improving a ground connection, or using a separate ground wire can help. Other times it makes more sense to break the loop with some form of isolator, like a transformer or USB isolator. There are few rules to curing ground loops other than to try the obvious, then work towards the less obvious/likely. I keep a couple of cheap audio transformers in the kit for a quick solution to buzzing computers and subwoofers.
    NorCal likes this.
  4. oloendithas
    And again every data cable (DisplayPort, DVI, USB, RCA) has a ground wire, so they all are connected to the ground of PC and thus grounded through it's power plug.
  5. Strangelove424
    1. Tell that to the people with noisy grounds. I am not one of them.

    2. Electrocution is alarmist by nature. I can’t think of anything more alarming. Even alarms themselves must take a back seat to the alarming nature of electrocution. In fact, if you get tired of simply alarming crooks you electrify your fence or the chair they are sitting in, therefore it is more alarming than alarms.

    3. Well, you gave me the slot machine let me fill in the reels... A dirty, clammy-handed barefoot hipster, in his overpriced studio loft with designer concrete flooring walks over to his equally overpriced boutique (but secretly failing) Chinese amp. Not that being Chinese has anything to do with it, but it does. This amp has no paint, and a fancy bare chassis to match said hipster’s modern sense of style. Since this hipster had so many problems with “ground loops” on his poorly isolated low quality amp, he used cheater plugs to lower background noise. This hipster is… dun dun... seconds from disaster.

    4. I’ve heard of bad solder or soldering applications creating larger than usual fail rates in some products or runs.

    5. Electrocution causes 300 deaths and 2,500 injuries annually. So your chances are much better than a lightning strike. And since your chances of winning the lotto is worse than a lightning strike, then you are sure to die in your lifetime to electrical bafoonery before winning the lotto.

    6. Alarmism is much better for educational purposes because it instills fear. Look how well alarmism works on the rest of Head-Fi. But I’ll be serious for a moment. There is no reason to take the risk. Simple as that. Yeah, it’s rare, but it still exists and there’s no reason to be illogical and risky when the world has made things logical and safe for you. Those cheater plugs, for instance, come little hangy eyelet things below for screwing them into a grounded screw on the plug. So they are effectively not cheater plugs, they are alternative grounds, but people even cheat the cheater plug.

    7. Safe and logical. Now you’re talking balanced like.

    8. It’s more fun to post after I’ve had a couple drinks.

    9. I should do it more often.

    10. Goodnight.
  6. Strangelove424
    NM, I meant things other than computer components grounded through the chassis. A lamp. A blender. An amp. A fan. A dryer. A toaster, etc. etc. etc.
  7. Arpiben
    For information purposes only.
    Even if electrocution injuries are around 200 cases per year in France, I prefer to have an alarmist attitude taken into account:

    1. Human body impedance:
    Depends mainly on:
    - entry and exit of voltage/current
    - skin depth
    - skin dryness
    - organs
    - voltage intensity and duration

    Skin impedance: varies from 10 kOhms to 500 Ohms
    impedance is around 900 Ohms
    Using clothings, gloves, shoes increase the total impedance.

    Peau means Skin
    Sèche means Dry
    Humide means Humid
    Mouilée means with Water
    Immergée means Under Water


    Body_Impedance 02.png

    2.Effects of current on Human body:

    The 500ms duration and 30mA value is the norm for ground current differential protection in my country.

    Body_Impedance 03.jpg
  8. pinnahertz
    1. I don't know any.
    3. Hypothetical cherry-picking. A loft with a concrete floor would not be grounded. And the floor would probably be wood. Are we going to keep this up?
    4. Bad solder application, yes. A failure of a high voltage, large solder joint, no.
    5. The figure I quoted is for audio gear. If your figure is global, there's still no worry.
    6. I'm not suggesting taking a risk.
    7 - 10. Good Grief.
  9. pinnahertz
    I disagree with the skin impedance figures. Try measuring skin with an ohm meter. 5K would be low, 1M would be dry, average.
  10. Arpiben
    You will have more details with the following document in english.
    Values don't seem to differ vs my country electricity safety data.
    Please note that skin resitance is higher when DC voltage is applied vs AC (50Hz/60Hz).

    FIG. 4 Total body impedance ranges for hand-to-hand or hand-to-foot contacts
  11. bigshot
    This thread is funny. I want one of you guys to run an experiment and stick a fork in the wall socket to determine the severity of the shock. Make sure you write down your expected results before you do it in case you aren't around to submit it for peer review yourself!
    NorCal and Arpiben like this.
  12. Strangelove424

    I've said my peace, and am ready to go Darwinian on this subject.
  13. Speedskater
    Some other forums would not permit posts that suggested dangerous practices like removing Safety Ground/Protective Earth wires or connections!
    A data cable or an audio interconnect cable is not a substitute for a Safety Ground/Protective Earth wire.
  14. bigshot
    Folks should be aware that insurance can refuse to pay after a house fire if the source of the fire was a non-UL listed piece of non-tested equipment or jury rigged wiring. It's better to fix things properly. Grounding everything properly is the solution.
  15. amirm
    There is no such exclusion in any homeowner's insurance policy. Indeed there is no reference to UL anything.

    You buy insurance for accidents that may very well be your fault. You can set your house no fire by leaving a hot pad next to your gas stove. Totally unsafe. Totally your fault. Yet insurance will absolutely pay. Ditto for smoking in bed, etc.

    What UL does is provide some level of protection for your the company selling the product. If their device catches on fire but they have UL certification and testing, their nose is much cleaner than if they did not. Neither has anything to do with you as the homeowner.

    Heck you or your child can short out an outlet, start a fire and the insurance company would still pay. There is no such exclusion in the policy.

    I bet every person here has non-UL tested equipment anyway. A ton of chinese made power equipment has either no UL listing or fake ones.

    The insurance that we buy has the very low percentage of risks for all it already built into its pricing.

    If you disagree otherwise, let's see the policy statement for your insurance.
1 2 3
5 6

Share This Page