Ungrounding an Amp--Any Danger?

My integrated amp, a Creek A50i, has a bit of humming. I was about to experiment with removing the ground from my power cord when I saw the sign on the back of the amp that said "This apparatus must be earthed."

Is there any danger (to me or the amp)associated with removing a ground? I will gladly put up with a bit of hum if it means not getting shocked or frying my equipment.
There is DANGER!!! DON'T DO IT!!!
Try using a 3 prong adapter with the ground pin out first.
The idea in "star-grounding" is to leave the component with the highest voltage potential -- the amp -- grounded, and to unground everything (everything) connected to it.
BUT, if you want to determine if the hum is intrinsic to the amp, just disconnect all inputs, so that the integrated has ONLY the speakers still connected. If there's hum through the speakers when you turn the unit on, then you'll know that the problem is not due to a ground loop. If there's no hum, then you may want to try star-grounding.
Ground protects you from being shocked in the event of an internal failure. Cheater plugs are routinely used to locate/cure ground loop issues. Violates electrical safety codes, hence the warning.
Defeating a three prong power cable with a "ground lifter" etc., is not only dangerous, it defeats the UL listing (Underwriter Labs is an independant body which tests electrical devices for safety under extreme conditions), and will invalidate ANY and ALL insurance claims that occur as a result of that equipment. The ground is there for a reason (aside from causing ground loops): in the event of electrical fault within the equipment (caused by internal or external forces), the ground acts as a safety to safely shunt excess current away from the chassis and you. Solve your hum problem by finding out what is causing the hum in the first place.
Your preamp should be the only grounded component to achieve the lowest noise floor. The other components drain to ground through the interconnects, to the preamp, and then through the preamp's grounded power cord.

Danger will then be present if you either

A.) Unplug the preamp from the wall, or
B.) Remove the interconnects between a (ungrounded) component and the (grounded) preamp, or
c.) Both

If you use a cheater plug(s) on a component(s) to eliminate hum (ground loops) just make sure the component has a path to ground -- usually through the ICs and the preamp. If you need to rearrange your stuff, just turn it all off AND unplug it first (which you should do anyway!)

Joe, I'm not sure you're right about grounding thru the amp. Yes it will provide a single ground path, but I understood that "star grounding" is when each component (or each device on a circuit board or chassis) has it's OWN shortest path to the ground point -- thus making the preamp the preferable compoinent to ground, and not the amp.

One thing a lot of people forget to do when using the preamp as the center of the "star", is to make sure ALL directional arrows (if present on the interconnects,) point to the preamp. This insures that all shielding inside the ICs will be connected to ground at the preamp. This isn't 100% true 100% of the time, but MOST cable manufacturers point the arrowheads to the end of the interconnect where they've connected the shield (the other end is not connected). The reason for all the BS about pointing the arrows in the "direction of signal flow" is so the manufacturers can get you to do the right thing without going into an explanation that might be over the head of most consumers.

Thanks for the warnings. I am recalling a story I heard some years ago about one of the former members of the Yard Birds. He electrocuted himself while tuning his electric guitar. I assume his guitar amp wasn't properly grounded.

The hum is definately in my amp since it remains when I disconnect my CDP. Since it's an integrated, there are not separate power plugs for the preamp and amp. Besides, the preamp is passive.

I have seen several products designed to block RFI and EMI, like the Stillpoints ERS cloth ($20 at Music Direct). Would these products be of any help? If not, the hum is minor enough that I can live with it.
Yes, it's dangerous. Assuming you're using unbalanced interconnect cables, it's likely that it's a ground loop that's causing the hum. This is the joy of using unbalanced connections. You could have a poor connector on one of your ICs. You should try to have everything in your system plugged into the same outlet. If you can not isolate and fix the grounding problem you can use an isolation transformer to break the ground loop. Jensen makes such products:


The comment about directional arrows and the shield being only connected at one end does not apply to the majority of unbalanced ICs (like coax) because the shield is used for signal flow and MUST be connected at both ends.
Bob, I'm afraid you are incorrect about "unbalanced" ICs. First of all, the correct terminology is "single-ended" ICs. Some single-ended ICs have a coaxial configuration, where the shield is indeed used for a signal carrier, but they are usually only found today in 75 ohm coaxial FM and TV cable and in some 110 ohm digital interconnect. They are no longer used for audio signal IC, except in those cheap RCA patch cords which are packed with inexpensive gear to get the consumer "up and running" -- and they don't have arrows on them.

The audio ICs I was referring to are known as "shotgun" interconnects, and were the original Bruce Brisson design which put Monster Cable on the map. Bruce left Monster early on to start MIT, which he still owns. The shotgun design is a single-ended interconnect (which means one ground conductor and one hot conductor), but there are TWO identical (double-barreled?) signal conductors surrounded by a ("floating") shield which is only connected at one end and therefore cannot carry any signal or current. The conventional (though not universal) indication for the end at which the shield is connected to the ground ring of the RCA, is an arrow which points to that end. This configuration applies to the majority of decent conventional audio cables today, from the cheapest Audioquest, Straightwire, Monster, XLO, Cardas, etc., to expensive (and even unconventional) designs like Magnan, Purist, Transparent, and Siltech. Two of the well-known brands that represent exceptions to the shotgun design (and only in certain models now) are Nordost (flat conductors) and Kimber (braided conductors). Using the old single-conductor-and-braid/shield design for audio signals anymore is just asking for "hum trouble."

Also, a word about "shielded" power cords. Manufacturer supplied power cords, if they are shielded (and most aren't) must have the shield connected at both the AC plug end and the equipment (or IEC plug) end. That is a UL underwriting requirement. Good for your power saw, bad for your audio -- instant ground loop!! So, if you feel your equipment might benefit from a shielded power cord, get an aftermarket one with a floating shield and no UL rating.

One place I absolutely recommend using a (floating) shield power cord is the DAC, CDP, or DVD. These devices tend to broadcast digital RFI from their power cords unless they are adequately shielded. They also pump digital RFI into the electric lines, so put them on their own circuit, or use a power conditioner which has a couple of outlets with digital filters specifically for this purpose.
Yes, the story is that Keith Relf died from electrocution while playing his electric guitar at home. A common version has him playing it in the bathtub, but his family has said this is a myth and he was found dead in his basement recording studio. Supposedly his amp developed an electrical fault, and it's also said that he was probably standing on an exposed gasline pipe that had something to do with it. What's interesting about this to me (other than that I am a Yardbirds fan) is that I play electric guitar, and can vouch for the fact that the vast majority of vintage amps do not have 3-pronged plugs in the first place.
Would somebody please explain to me what the Signal Ground Lift switch on the back of my amp is for? IN LAYMAN'S TERMS. I don't really understand electricity, you see. I believe the switch is currently lifted, but I've no clear idea what that did for me (again, in layman's terms). "Well, it needs to be either lifted or not," I thought to myself, and "lifted" seemed somehow the sonically appropriate term. Maybe I should have read the owners' manual..

My amp has a hum that you can only hear when your ear is right up to it; i.e. a pretty quiet amp? Is this the sort of (very subtle) hum people are complaining about? All my stuff is plugged into a power conditioner, which of course has three-prong receptacles, and the power conditioner is plugged into a (grounded) wall outlet. All of these important tasks -- so vital to turning the stereo ON, so you can, you know, LISTEN TO IT -- were accomplished on Day 1. I'd have thought that was the last thought I'd need to give to the matter of electricity, but maybe I'm just hopelessly naive.
Nsgarch: High quality coaxial analog interconnects, if not prevalent, are indeed made, by van den Hul for instance (which I happen to own)...there are others...
By all means, contact the factory or authorized dealer/tech about the problem. Until then, just put up with the a little hum. No doubt this can be repaired or corrected. Just my 2 cents.
Zaikesman, re the van den Hul "screened" (shielded) coaxial products: There are only two unbalanced cables which are spec'd. for audio signal transfer, and both of them use vdH's proprietary non-metallic "Linear Structured Carbon" as the shield. Since it's carbon (non metallic) it's not affected by EMI (hum) and does a pretty good job soaking up RFI.

But as I said in my earlier post (I should have listed it I suppose) this product is definitely a hybrid. Most of vdH's audio cables do use balanced (shotgun) conductors in fact. You can see them all at:


07-13-06: Skippack
My amp has a hum that you can only hear when your ear is right up to it; i.e. a pretty quiet amp? Is this the sort of (very subtle) hum people are complaining about?
No. What you describe is normal. A ground loop hum can be heard from several feet away, or more.
Nsgarch, as you describe "single-ended" it sounds like an unbalanced connection to me -- two conductors and one is tied to ground. I'll agree with you that ICs that have two conductors (a hot and a ground as you say) and a shield may not have the shield connected to the ground conductor at both ends. The point I was alluding to is that the signal return flows in the ground conductor in an unbalanced cable.

Please explain why using a shielded power cord produces and "instant ground loop."

And if I understand your other comments correctly you are recommending that people circumvent the protection of a grounded power cord on all but one component and thus rely on the ground of the interconnection cables to ground the ungrounded component. Do you really think that is wise?
Apart from what's already been said, I have found that not grounding my amplifier yields better dynamic range.
Nsgarch: I own vdH The First Ultimate and The First Metal Screen coaxial IC's. Metal-free TFU uses a carbon fiber screen/return, which far from being especially immune to EMI/RFI, is actually not recommended for use in situations where hum is likely to be picked up, including connections to tube amps -- a drawback in an otherwise outstandingly high fidelity design that I can personally attest to. TFMS was introduced to help in just those applications, by virtue of (as the name implies) an additional metal-mesh screen layer. vdH also makes several other models employing coaxial construction for analog use, both all-metal and metal/carbon hybrids (The Bay, The Name, D-300, The Source, The Well, D-310, The Combination, and MC Silver are all coaxial designs intended for audioband signals). WireWorld is another company that makes coaxial analog IC's, using a variation where the inner and outer conductors are both tubular. (It might be worth adding that, although coaxial cables by definition have their screens connected at both ends, WireWorld still claims these models are audibly directional due to grain orientation.)
I have to agree with Bob above. Manufacturers connect one side of a single ended interconnect to ground (except in a truly balanced system-a discussion for another day.) In a 2 wire cable with a shield, one wire and the shield will be connected (the shield being at one end for this discussion but it could just as well be connected at both ends) to ground in the unit. The other wire is considered hot. Since the audio signal is AC (alternating current) it has to be referenced to ground. If you lift the chasis ground (and this is what you're doing), it will leave any overload (short, etc.) to travel through the interconnect. You have to look at it as house wiring. On 115v, one line is hot, one is neutral(the neutral carries current but has no potential-hence the name neutral) and the other is ground (shield.)This is exactly like an interconnect. One wire is hot and the other is neutral, BUT, the white (neutral) also goes back to the grounding buss!(This is why a two wire cable w/a shield is no different than one with a wire and a shield and both work. The added wire gives a better balanced path for current.) One side of the interconnect has to be neutral at some point or it want flow current. (Current is what pushes voltage BTW-can't have one without the other) That doesn't mean the signal doesn't transverse both wires, it does, but one is always going to be referenced to ground(neutral) somewhere.
I would always leave the source of largest current draw grounded (the amp.) If something happened to the transformer(like a short) it's maximum amperage will have to be pulled through the interconnect of another component that is tied to ground since you no longer would have a separate chasis ground to carry it.
I use a Theta Dreadnaught with a 2.2 KVA transformer (2200 watts.) Theta specifically says do not float the ground. I can see why.
Is floating grounds safe, yea, for the most part and long as you don't have a power supply failure. Smaller components will draw much less current if a problem occurs.
I have never understood why people make audio out to be something so special in electrical principles. It's not. It conforms to basic electrical fundamentals like anything else. Go to a book store and read up on the subject. Interconnects are nothing but wire to transfer a small AC signal from one point to another. It works EXACTLY like household wiring which is AC but at a much lower voltage for audio. (Household AC is usually around 115v-120v @60hz. If you could find a driver to handles the current it would make for a loud 60hz note!!!!!)
Another point is when current is unbalanced in your electrical box, it balances itself by sending a differential current down the neutral to create a balance. It's always good to ground most things! Better it travel the ground than through you if a disaster strikes and you touch the appliance. BTW, just for general information, even though ground and neutral end up at the same buss, they are NOT the same. The 3rd wire ground is for your protection. Neutral is to complete the circuit.
Bob and bigtee, we are wandering far from Mr. Socprof's thread, so let's not do that.

I'd be happy to continue this discussion either privately (you can email me) or, if either one of you would like to start a thread on the subject of balanced versus single-ended interconnect design, and their susceptibility to EMI/RFI contamination. That is not the subject of Mr. Socprof's inquiry.


Neil, no, I really don't think I wondered from the original point. My post was to explain, in simple terms, why lifting a ground can be a bad thing. Wasn't that what the original poster wanted to know?
You brought up in a follow up to Bob's post about shotgun interconnects. Call it what you want but it is still an unbalanced single ended interconnect. It really doesn't matter what type of single ended cable you use. If you ground one end of a shield only and you have 2 wires left to carry the signal, one of the wires must be considered a neutral because one will still be connected to ground. Neutral ultimately goes back to ground and since the shield is not capable of conducting a flow, one of the other wires will carry the load if a power supply failure occurs. I just wanted to clarify that the type of cable really didn't make a difference when lifting a unit ground. Current during any failure will be required to travel the interconnect if the ground is lifted. That was my point. I also stated, "Balanced was for another day." I think the differences have been discussed and at length here and other places.
Bigtee: So??
Nsgarch, So What?
the problem you're describing may be caused by cable tv. Is it near by? If so you may want to try and disconnect the cable at its first point of entry. then check for hum in your system. if it's gone, all you need is a ground loop eliminator, such as Mondial Magic Box.
Skippack, let me answer your (very reasonable) question which is in the spirit of this thread: The signal ground lift switch is used to lift (disconnect or separate) the signal ground return leg in the amp's audio circuit(s) from the metal chassis (which still remains connected to the ground prong of the AC plug. Bryston has done this for years (do you have a Bryston?) This means that your non-shotgun coaxial single-ended interconnect is less likely to become part of a ground loop and pick up hum, while at the same time you are still personally pretty well protected from most shock hazards, even if you're dumb enough to remove your interconnects without turning off or unplugging your amp (just don't touch anything inside because under those conditions, it will take literally hours for the power capacitors to discharge!!)

When the manufacturer provides them, you always see these switches on amps but never on preamps or other components. That's because manufacturers expect you'll always want to ground the preamp to make it the center of a proper "star grounding" layout.

If you are using balanced interconnects between your amp and preamp, meaning the kind with an XLR connector and three signal conductors (plus, minus and ground) with a (optional) shield, then it's not necessary to "lift" the signal ground at the amp or use a cheater on its AC plug because any noise generated in, or picked up by the signal conductors will be self-cancelling.
Nsgarch: Thanks! Yes, it's a Bryston amp. I use single-ended ICs, so I guess I'll leave the switch on lift. There is no discernible difference in sound or noise-level, however (to my ears).