[mythtv-users] Slightly OT: receiver input buzzing/humming

[Steve Adeff]

On Wednesday 18 January 2006 21:38, Steve Briggs wrote:
> Steve Adeff said:
> > phasing shouldn't matter as long as the grounds are shared and
> > proper sizing, etc. Ground loops in proper working equipment are
> > formed when the impedence to ground for the two components are
> > different. Connecting them together then causes one of the
> > components to enter into the ground return of the other component.
>
> Technically, when the potential (voltage) to ground isn't the same
> between the two components; then there's a potential difference between
> the two components and currents can flow from one to the other.

right, but we're not talking the hot side of the power line here, we're 
talking about GROUND, and unless your using two different grounds for your 
equipment, or something is drastically wrong, ground is ground is ground. The 
only time your situation will occur in home theatre equipment is if there is 
a bad cap/etc causing the AC voltage to leak into ground.


> In addition, when the 2 components are on different outlets, the
> ground wires follow different physical paths and stray 60 Hz magnetic
> fields will induce currents on the loops -- if the components are
> plugged into the same outlet, the area enclosed by the ground loop
> is much less and the corresponding currents are less.

first, if both loops ground independently of each other and you combine the 
two grounds at one end you have no real opportunity for magneticly inferred 
currents in the ground wire, and if you are getting one, you have an immensly 
strong alternating magnetic field where it shouldn't be (ie in your home) , 
and I'd be careful not to let small animals near it.

Now, this seperate grounds issue is why I advocate 3-prong components, even if 
you use two outlets as long as their grounds are common (earth, which it 
should be) you'll get no magnetic ground hum because you'll be grounding 
everything everywhere in either a giant loop or if both ground plugs have 
independent grounds then both sides to ground.

> True, but in practice, it's almost impossible to actually
> isolate anything, particularly with cheap power supplies grounded
> to the case and RCA cables (also generally grounded to the case)
> used for interconnects.

its not impossible, like you said, just don't go with a cheap design. Two 
prong components should also not be using the case for power supply ground, 
only electronics "common".

> IMHO, RCA cables are a horrible design; when you have ground currents,
> they run on the shield, which turns the shield (supposed to shield the
> inner conductor from stray signals) into a radiator.  Bonding the
> components together with a separate ground strap (admittedly a
> band-aid, but often the easiest solution, particularly for those of us
> who can't put everything on the same outlet) provides a lower-impedance
> path between the components and keeps the currents off the RCA
> cable shield.

agreed, which is why G-d invented XLR cabling ;-)

> I also note in industrial environments, millivolt sensor wires are
> always shielded, but the shield is only connected at one end, never
> both (to prevent currents on the shields).  I've often wondered
> (but never done a good test) what would happen with RCA cables if
> you disconnected the shield at one end of the cables to prevent
> currents on the shields.

its not so much to prevent currents on the shield as to prevent linking stray 
grounds between equipment. In fact, having one end open invites currents on 
the sheild because the potential at the ungrounded end has to pass through 
the resistance of the wire to get to ground, causing a voltage. Now, with 
milivolt wires this isn't a problem because the impedence and voltages are so 
low you shouldn't even be inducing a voltage, but with higher power lines 
this is a huge deal, 120V home wiring should not be affected by this except 
in the presence of a rediculously strong magnetic field.

-- 
another Steve