/Question. Which end of a shielded wire should be run to
/ground?
Shielding of wires adds a modicum of protection for
two kinds of propogation into or out of a wire:
The physical positioning of two wires together
in the same bundle causes them to couple to
each other in two ways - electrostatic and
electromagnetic. A wire which carries high
frequency noise in less than an idealized
"trasmission line" mode will radiate components
of that noise just like an antenna.
Electromagnetic coupling occurs as a result of
a magnetic field which MUST exist about any wire
carrying a flow of electrons. Any modulation of
intensity (some signal component other than pure
d.c.) in a wire makes it the primary winding of
a transformer. Adjacent wires are potential
"secondaries".
Electrostatic coupling occurs any time there
is a dynamic (rapidly changing) voltage difference
between adjacent conductors. The insulation between
wires (including the air space) acts just like the
dielectric of a capacitor and some amount of the
antagonist's signal can couple into parallel routed
victims.
Radiated noises travel about just as the signals
into and out of the antenna on your hand held radio.
Shielding will attenuate electro-static coupling
a great deal. Electro-magnetic coupling is only
slightly attenuated by common shielding techniques.
Attenuation of radiated coupling with shields falls
somewhere between the values for electrostatic and
electromagnetic coupling.
As a general rule, I ground the shield on one
end only and to the device for which we're building
a noise-wall.
What are the criteria for requiring shielded wires?
Before you can decide which end to ground, you must
first decide if shielding is a good idea. You have
to decide which devices in the airplane are potential
victims and which are potential antagonists. You also
need to decide what propogation modes exists between
V/A pairs and how the propogation can be attenuated
to acceptable levels or eliminated.
Obvious antagonsist are ignition wires. Spark plug
wires seldom run in bundles with other wires but they
are strong potential antagonists -AND- radiators.
So, we use resitance wire to de-qualify the conductors
and good carriers of high frequency energy -AND-
shield them to keep all the noises bottled up inside.
Obvious victims are audio systems, signal lines between
radios and VOR indicators, small signal lines between
encoders and transponders, etc.
All three propogation modes can prevail but the strongest
is electrostatic. The best defence against electrostatic
coupling is distance . . don't put vicitm wires in same
bundles with antagonist wires . . . same thing works well
for electro-magnetic coupling . . . but electromagnetic
coupling is pretty weak unless the antagonist wire carries
lots of current . . . 10 amp or more.
The strongest defence against electromagnetic coupling is the
twisted pair of wires . . will touch on that again later. Shielding
both victim and antagonist wires will help with attenuating
the electrostatic modes. Radiated coupling is very rare.
Shielding of victim wires is sometimes helpful but you
generally don't have much problem with radiated victimization
unless the victim is badly designed.
/I saw a friends professionally wired Lancair IVP
/(P for Pile of money), and it has shields on almost
/all the radio wires, grounded at the radio/electronic
/goodie. I didn't see where the alternator output was
/grounded, (actually he has 2).
It's not uncommon for an amateur builder to consider the
things I've just written above and throw his hands up.
There's about a million possibilities and he's right.
So, you can either shield everything in hopes that you've
covered all the potential problems but I assure you.
It's heavy, time consuming and for the most part, totally
unnecessary.
/Specificially where should you ground the:
/Radio power leads
Everything behind the panel should get power ground from
a single point if possible. I developed the firewall/panel
ground bus that B&C builds especially for this purpose.
Single point grounding eliminates the possiblity for noises
getting into systems via other paths (i.e. ground loops -
QUITE common in composite airplanes).
/Headset stuff
Headsets are quite happy wired with twisted pairs and no
shielding. Insulate the headset jack from local ground
where it mounts and ground via one lead of the twisted
pair to the "headset ground" or "headset LO" pin on the
audio system.
Microphones too will work quite nicely in most systems as
a twisted trio of wires . . . again, insulate jack at point
of attachment to airframe and gound to appropriate pin on
audio system.
I've been stocking a 5-conductor, shielded cable that I
used on my installations. Two conductors for headset,
three for microphone/push-to-talk and a separate shield
ground over the bundle. The shield and microphone LO tie
together at the audio system.
/Alternator output wire
Shielding this wire is totally unncessary . . . I know,
thousands of Cessnas do it and for the life of me, I don't
know why. I was working at Cessna when some of the ADF
noise problems were being worked on the "new" line
of ARC 300 series radios. A capacitor on the alternator
-AND- shielding of the alternator wires were added too
but I don't recall any quantitive testing done to justify
the shielding . . . the filter took most of the noise out
of the ADF . . . shielding MIGHT have helped some VHF
comm systems but I kinda doubt it.
/Alternator field wire
Same as above.
/What others should use shielded wire?
Anything recommended by the manufacturer of a product.
For example, the strobe folks usually supply a shielded
trio for their strobe heads. Audio/Intercomm System
manufacturers should be very explicit as to where
shielded wire is necessary or recommended. They should
also tell you EXACTLY where to ground it. Same thing
goes for any other black boxes in your avionics suite.
For the most part, I find that 95% of builder problems
with noise are a result of architecture (too many ground
behind the panel, poor wire routing considerations) or
failure to follow manufacturer's instructions.
Shielding is some help on some noise problems. However,
you cannot predict that shielding is going to be the fix
of choice without knowing the antagonist/vicitm/propogation
mode for a particular noise problem, . . . most of the time,
shielding won't help.
This is why it's so important to pay attention to ALL noise
control techniques. Let me wire your airplane and I'll bet
I can get it to run noise-free with NO shielding on any
wires behind the panel.
About 10 years ago, when the 'Connection was just getting
started, one of my readers called and told me about all
the shielding and filters he'd installed on his airplane.
He then asked if there were any other filters I'd recommend.
I asked if he was having a noise problem and he told me that
he'd not yet flown the airplane! This guy had invested
POUNDS and HOURS in a noise reduction system that may well
have produced no positive results if other considerations
were not a part of his overall noise control program.
I recommended that he take it all out and get rid of the
weight and complexity. IF and WHEN he had a noise problem,
we'd work out the VAP (victim/antagonist/propogation) loop
and fix it.
Regards,
Bob . . .
AeroElectric Connection
////
(o o)
| |
| Go ahead, make my day . . . |
| Show me where I'm wrong. |
72770.552@compuserve.com
http://www.aeroelectric.com
|