[sdiy] Magnetic Bug squashed !

[Magnus Danielson]

From: harry <harrybissell at prodigy.net>
Subject: Re: [sdiy] Magnetic Bug squashed !
Date: Sat, 26 May 2001 23:54:44 -0400

> Hi Magnus:

Hi Harry,

> Thanks for the comments.

Anytime!

> I'm using the nylon shoulder washer trick almost all the time now... The idea
> is that I control whan and where a ground occurs... not the random chance
> of the jacks themselves.

Certainly a good and valid point - in audio!
I've seen people trying the same idea in RF environment (say signals
of several 100 MHz) - not a good idea!

> They make nylon 1/4 jacks but I hate them. They strip out way too
> easy. So I'm using a flat and a shoulder washer, and the steel hardware.
> Only a problem on a really thick panel...

Yeah, right. It takes some time to find a solution which is both
sturdy as well as electrically functional.

> The chassis actually gets its ground now from the rack itself... I've found
> that eliminating all but one ground in the system is usually the lowest noise
> solution...

Why do you think this is the prefered way of doing things in say
telephone stations? ;o)

Well, actually, I think that you should not fiddle with the ground
connections on the power side at all. But, you should not rely on
those grounds for providing a quite ground for your audio signal.

> I prefer the audio cables for that point... since you usually want those
> grounds anyway. But IF you are using AC mains powered equipment its
> a real good idea to have a resistive or capacitive ground for
> safety.

Indeed! Fiddeling with the way the safety wireing in gear works just
shows that you are not looking at it with a professionals eyes. You
avoid the problem but create new ones. There should be a DC path and
there should be means of prohibiting buildup of large voltages and
currents. All this is really a field of its own. There are good books
written about it.

> I know that a major fault will weld the audio cables in place (I
> don't care... they will conduct enough current to
> trip the breaker... eventually).

This only shows another missconception people have, it is not the
overvoltage which kills many things, its the overcurrent which kills
them! In the end you really must consider the amount of energy
exposure (leading to overtemperatures) at points.

You don't want to learn how ridiculously small currents it takes to
make the heart give up.

My point, as well as Harrys, is that don't compromise the safety of
gear just to reduce noise, there are usually much better ways of doing
things.

> The shielded transformer is a good thing too... I run my AC mains through a
> shielded
> Toroid isolation transformer... a 120V input / 120VCT output. The center tap
> goes
> to a separate ground rod for the studio. VERY quiet... BUT you must not run
> equipment on the normal 120V (hot-neutral) and connect THAT to the system...
> or
> you BLOW it (literally)

Indeed. Indeed. When you mess with such things, you are out on a very
dangerous path.

What you can do is to use an isolational transformer to restore a
sound relationship between the ground and the live/zero. You don't
actually have to replace the ground since you now have isolated away
the potential diffrances between them. Thus, the ground-to-zero
voltage becomes zero again (as it should be from your main fuse).

Unluckilly there exist no real standard for how to do all this and all
kinds of solutions exists. That people tend to run unbalanced signals
all over is also not much of a help.

I wrote a little thing on grounding over at SAS, maybe it's time to
translate it to english and drop it on this list...

Cheers,
Magnus - connecting grounds