Insane ground loop; anybody wanna try?

Overheating conductors in walls is a major cause of house fires so you are correct but then you did prove my point. The grounding scheme in residential construction is designed to protect property not people it’s a well known fact and as I said you just backed that up.

No, I did not back that up. I said that **the breaker** protects against fires in the walls due to overheating conductors, not the grounding scheme. As I said in the last paragraph of my preceding post, "under most circumstances the safety ground conductor has no involvement in the fire protection provided by a breaker."

I also explained how the grounding scheme, specifically the safety ground connection, protects against electrical shock in the event that the AC "hot" wiring within an electrical device shorts to its metallic chassis (if it has a metallic chassis), by causing the breaker to trip before anyone has a chance to come in contact with the "hot" chassis. That is why 3-prong outlets and 3-prong power plugs (on devices that are not "double insulated"), which provide a safety ground connection, replaced 2-prong types (which don't provide a safety ground connection) many decades ago.

I agree, of course, with your statement that "GFCIs will react and kick in milliseconds not seconds and are designed to protect people." However your preceding post, to which I was responding, stated that...

If there is current on the chassis and someone touches it the breaker is probably not going kick in time to save someone. Only a GFCI which is designed to protect people may save someone from a shock or electrocution.

... which as I explained is not a correct statement.

Regards,
-- Al

The ground system in a home is not for people protection it is for fire protection. If there is current on the chassis and someone touches it the breaker is probably not going kick in time to save someone. Only a GFCI which is designed to protect people may save someone from a shock or electrocution.

No, that is not correct. A short between AC "hot" and the metal chassis of an audio component or any other electrical device having a 3-prong power plug is obviously very unlikely to materialize at the same instant someone touches the chassis. So the breaker will almost certainly have tripped some time prior to the chassis being touched, as a result of the excessive current that will be drawn by the short between AC "hot" and the safety ground conductor.

A GFCI can provide **additional** people protection under certain circumstances, such as in bathrooms or other areas where there is water.

And regarding fire protection, the main purpose of a breaker is to protect the AC wiring in the walls from overheating, as a result of an overload. In the event of an overload, the breaker will trip as a result of the excessive current passing through the "hot" conductor, which will normally be returned through the "neutral" conductor, not the safety ground conductor. So under most circumstances the safety ground conductor has no involvement in the fire protection provided by a breaker. That said, however, there are certain scenarios in which defeating the safety ground connection of an audio component could conceivably result in a fire hazard, as well as a shock hazard.

Regards,
-- Al

Can you tell me why PS Audio cords have removable ground pins?

I’m not familiar with their power cords. But I would guess that if those cords allow the AC safety ground connection to be defeated then the PS Audio components that they are presumably intended for use with (or if they are non-removable, that they can only be used with) have internal AC wiring that is "double-insulated" and is therefore very unlikely to ever develop a short to chassis. Assuming the chassis is metallic, that is. If the chassis on any of their components is non-metallic and as a result non-conductive, that would be an additional reason why defeating the safety ground would be a reasonable practice in those particular cases.

Regards,
-- Al

There are amplifiers sold with ground lift switches, so I think this is a common problem and amp manufacturers think it a safe solution.

As a point of information, the ground lift switches that are provided in some components do not isolate AC safety ground from the chassis of the component, as a three-to-two "cheater plug" would do. Rather, they isolate the component’s internal signal ground from the chassis, which does not result in the safety hazard that isolating AC safety ground from the chassis would result in.

Regards,
-- Al

Sounds good!  Glad I was helpful.

Regards,
-- Al
 

Edwyun 2-22-2017
There can be a combination or cumulative effect of issues.

+1. Specifically, it’s possible that there is nothing "wrong" with the AC wiring to the listening room, but it may have different characteristics than the more recent wiring to location 1, which in turn are bringing out an inherent susceptibility of the power amp/preamp combo to ground loop issues.

More specifically, see pages 31 to 35 of this paper, written by a renowned expert on such matters, in which he describes "what drives 99% of all ground loops." As you’ll see, Romex (which perhaps was used in the relatively recent wiring of location 1) is much less likely to bring out the susceptibility to ground loop issues that is inherent in some designs than wires randomly positioned in conduit (which perhaps was used in the older wiring to the listening room).

If that is what is occurring, and it seems like a reasonable bet at this point, a solution may be to have an electrician install new wiring to the listening room, using Romex. Alternatively, inserting a suitably chosen Jensen transformer between the preamp and each of the amps would probably resolve the problem, while also perhaps reducing other forms of noise, at a cost of around $300 plus some additional cabling.

Reports that have been provided here regarding the transparency of Jensen transformers have generally been very positive, although I can recall a couple of people who have provided comments that are a little bit mixed.

A suitable Jensen model would be the PI-RR, which would accept an RCA cable at its input and provide an RCA output, or the PI-XR, which would accept an XLR cable at its input and provide an RCA output. You would need two of them, one connected via a short RCA cable to the input of each amp. The cable from the preamp to the Jensen transformer could be whatever length is necessary. Here is a good supplier, although they can also be ordered directly from Jensen.

Also, although it is apparently unrelated to the problem I would still strongly advise that at some point soon you eliminate the XLR-to-RCA adapters, that are shorting the signal on pin 3 of the preamp’s XLR outputs to ground. If the sub does not provide XLR inputs you could either special order adapters from Cardas which leave pin 3 open, or have a cable supplier custom make an XLR to RCA cable which leaves pin 3 open, or purchase a Jensen PI2-XR transformer which would provide two channels of balanced to unbalanced conversion. Or you could use the preamp’s XLR outputs for the main signal path, connecting them to the amps via the Jensen PI-XR transformers, and connect RCA cables from the preamp to the sub.

Good luck. Regards,
-- Al

An experiment I'd suggest at this point that might prove to be informative would be to see if either Quicksilver amp produces hum when it and the preamp are connected and powered up as usual, but with the other Quicksilver amp disconnected from the preamp and unplugged from the AC.

Regards,
-- Al
 

the sub is connected to the XLRs via adapter and then by RCA cables.

I wouldn’t be surprised if that turns out to be the key to the problem. While the post I made earlier today in this thread addresses a somewhat different situation, it provides what may very well be relevant background.

The following paragraph in that post may be particularly relevant:

Most XLR-to-RCA (and RCA-to-XLR) adapters short the inverted signal on XLR pin 3 to ground (XLR pin 1), although I believe Cardas can supply adapters on special order that leave pin 3 unconnected. Shorting XLR pin 3 to ground is appropriate when adapting RCA outputs to XLR inputs, and may be appropriate when adapting a transformer-coupled XLR output to an RCA input. However, at best it is poor practice when adapting an actively driven XLR output to an RCA input. And while many and perhaps most components would be able to tolerate that, with some designs shorting that signal to ground may result in major performance issues (see this thread, for example), or conceivably even damage, eventually if not sooner.

The likelihood of a problem arising as a result of shorting an output signal on XLR pin 3 to ground with an adapter figures to generally be greatest when the output impedance of the circuit providing the signal is low.  That is the case with your Capri preamp (80 ohms balanced output impedance, presumably corresponding to 40 ohms for each of the two signals in the balanced signal pair).  In the case of the thread linked to in the paragraph I quoted above the result was buzzing, but hum seems very conceivable as well.

Regards,
-- Al

Unplugging the sub has little effect ....

Is the sub connected at speaker level (to the amps) or at line level (to the preamp)?  And if the latter, is it connected to the preamp via XLRs, while the power amps are connected to the preamp via RCAs as you stated?

In any case, if you already haven't, try disconnecting the sub from whichever component(s) it is connected to, in addition to unplugging it from the AC.

Regards,
-- Al