Grounding Shielded Wire - One or Both Ends?

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

antero

Well-Known Member
HBA Supporter
Joined
Dec 8, 2009
Messages
96
Location
Centennial, CO
Hanging out with lots of airplane people I have heard anecdotally time and time again that shielded wire should be used on sensitive equipment, especially things like audio wires, and that the shield should be grounded at ONE END ONLY. Usually this means grounding the shield at the source, i.e. grounding to the radio chassis for headphone connections and not at the headphone jack end. I have seen many homebuilts wired up this way. But I started to question this method after reading through my Garmin SL-30 radio installation manual. There, it shows in MOST cases that the shield should be grounded at BOTH ENDS - examples: the headphone connections (one to the radio chassis and one to aircraft ground) and radio/remote CDI signal connections (grounded to both chassis). However, it also shows that the radio / audio panel interconnections should be grounded at ONE END only, being the audio panel chassis.

Now, quoting scripture:

AC 43.13-1B, 11-89

With the increase in number of highly sensitive electronic devices found on modern aircraft, it has become very important to ensure proper shielding for many electric circuits. Shielding is the process of applying a metallic covering to wiring and equipment to eliminate interference caused by stray electromagnetic energy. Shielded wire or cable is typically connected to the aircraft’s ground at both ends of the wire, or at connectors in the cable. Electromagnetic Interference (EMI) is caused when electromagnetic fields (radio waves) induce high-frequency (HF) voltages in a wire or component. The induced voltage can cause system inaccuracies or even failure, therefore putting the aircraft and passengers at risk. Shielding helps to eliminate EMI by protecting the primary conductor with an outer conductor. Refer to MIL-DTL-27500, Cable, Power, Electrical and Cable Special Purpose, Electrical Shielded and Unshielded General Specifications.
and

AC 43.13-1B, 11-106

Wiring of sensitive circuits that may be affected by EMI must be routed away from other wiring interference, or provided with sufficient shielding to avoid system malfunctions under operating conditions. EMI between susceptible wiring and wiring which is a source of EMI increases in proportion to the length of parallel runs and decreases with greater separation. EMI should be limited to negligible levels in wiring related to critical systems, that is, the function of the critical system should not be affected by the EMI generated by the adjacent wire. Use of shielding with 85 percent coverage or greater is recommended. Coaxial, triaxial, twinaxial, or quadraxial cables should be used, wherever appropriate, with their shields connected to ground at a single point or multiple points, depending upon the purpose of the shielding. The airframe grounded structure may also be used as an EMI shield.
I am confused. For anyone who knows better than I, what is the proper method for grounding shielding - is it one end or both ends, or does it depend? Does it matter if it is an EMI source vs an EMI sink? What other considerations should I be thinking about?

Thanks in advance!
 

Himat

Well-Known Member
Joined
May 5, 2011
Messages
2,868
Location
Norway
The short answer, it depends.
The practical answer, try both and see what that do work best. Just remember, changing the earthing on one devise might influence all others.

Why it depends? It's a about the wavelengt of the "noise" signal and the lenght of the wires. And what kind of coupling there is between the noise and the wires. In some cases there are two screens, the outer terminated in both ends and the inner in one end. A detail, "pigtails" terminating the screen might draw noise into the circuit if done the wrong way. In marine applications it is common to use cable glands that also terminate the screen to the enclosure.
 

Dan Thomas

Well-Known Member
Joined
Sep 17, 2008
Messages
6,589
Our avionics tech tells me that audio wire shields are generally shielded at one end, the radio end, with the jack end ungrounded. Antenna cables have their shields grounded at both ends.

I can't figure out exactly why, but I have had luck with eliminating some alternator whine in the headsets by isolating the audio jacks from ground.

Like this:

Jack_Insulation_Washers.jpg

Dan
 

Joe Fisher

Well-Known Member
Joined
Feb 10, 2007
Messages
1,379
Location
Galesburg, KS South east Kansas
With antennas they are a push pull circuit ( think A/C). So the ground plane needs to be exactly the opposite polarity to the antenna at all times. The shielding is the same length as the antenna wire so you should have a match.
 

Himat

Well-Known Member
Joined
May 5, 2011
Messages
2,868
Location
Norway
Our avionics tech tells me that audio wire shields are generally shielded at one end, the radio end, with the jack end ungrounded. Antenna cables have their shields grounded at both ends.

I can't figure out exactly why, but I have had luck with eliminating some alternator whine in the headsets by isolating the audio jacks from ground.

Dan
And you ar not the first one that can't figure out exactly why. It's quite common and it is because the problems in practice is complex. The basic theroy is well known, but to apply it the theory is another matter. EMI, electromagnetic interference, and EMC, electromagnetic compability, have been somewhat the "black magic" part of electronics.
 

kent Ashton

Well-Known Member
Joined
Aug 15, 2008
Messages
799
Location
Concord, NC
Our avionics tech tells me that audio wire shields are generally shielded at one end, the radio end. . . . I can't figure out exactly why
Ground loops, my friend. If two grounds for the same device have slightly different potential, a small current will flow that will affect the small current from the sensor. In my Cozy, I had a fuel pressure sender that was grounded at the gauge and at the sender. It caused me no end of problems with false and spurious readings. Readings would change at different flight conditions and altitudes. I might taxi-out with a good FP reading and it would go to zero during climb; or it might be low on the ground and good in the air. I finally grounded all shields and negatives on the engine block and my problems went away.
 

antero

Well-Known Member
HBA Supporter
Joined
Dec 8, 2009
Messages
96
Location
Centennial, CO
I finally grounded all shields and negatives on the engine block and my problems went away.
When you say this, do you mean you grounded all shields on both ends or one end?

Yes, I have heard the reason that grounding both ends of the same shield is more prone to ground loops many many times, which is why i was so surprised to see the wording in AC 43-13 suggesting the opposite, and was surprised to see Garmin recommending it be done in their installation manual.
 

Himat

Well-Known Member
Joined
May 5, 2011
Messages
2,868
Location
Norway
A long time ago I worked as a service engineer on maritime navigation systems. As part of my job I had to learn about EMI and EMC. As far as I remember, there is a difference in “earthing” of the screen and “terminating” the screen. If the purpose is to encapsulate the electronics within a Faraday cage, you want to terminate the screen to the enclosures at both ends. The outside of the screen might only be connected to the structure at one point. Same with the power supply to the electronics, this should only be connected to the structure in one single point. Actually the ground loop problem on a ship can prove “interesting”. Pull a cable between two points a hundred meters apart and there might be a potential difference by tens of volts. This with a steel hulled ship.

I guess there are several books about EMI and EMC out there. At the time I found that “EMC for Product Designers, Second Edition by Tim Williams”, did give a good introduction to the subject. As the book in a ten year span evolved to the fourth edition it might be a reasonable source of information.
 

Joe Fisher

Well-Known Member
Joined
Feb 10, 2007
Messages
1,379
Location
Galesburg, KS South east Kansas
The Hiperbipe has steel tube frame with the battery and master solenoid behind the seats. I put two heavy ground straps from the steel frame to the crank case. I wanted to make sure that the throttle cable didn't glow red hot during start. I used Westteck electric instruments. When I tested the engine getting ready for flight the instrument readings went crazy. Oil pressure, Fuel pressure,oil temp were unusable. What fixed it was to run a separate ground from each sender directly to the instrument case. The airframe ground was not good enough.
 

Dan Thomas

Well-Known Member
Joined
Sep 17, 2008
Messages
6,589
The Hiperbipe has steel tube frame with the battery and master solenoid behind the seats. I put two heavy ground straps from the steel frame to the crank case. I wanted to make sure that the throttle cable didn't glow red hot during start. I used Westteck electric instruments. When I tested the engine getting ready for flight the instrument readings went crazy. Oil pressure, Fuel pressure,oil temp were unusable. What fixed it was to run a separate ground from each sender directly to the instrument case. The airframe ground was not good enough.
That was a common problem with Cessna's electric oil temp gauges. As the ground cable connections between the engine case and airframe got old and dirty and a little corroded, the oil temp indication would spike. A separate wire between the crankcase, near the sender, and the instrument, solved the problem.

Dan
 

Dan Thomas

Well-Known Member
Joined
Sep 17, 2008
Messages
6,589
Ground loops, my friend. If two grounds for the same device have slightly different potential, a small current will flow that will affect the small current from the sensor. In my Cozy, I had a fuel pressure sender that was grounded at the gauge and at the sender. It caused me no end of problems with false and spurious readings. Readings would change at different flight conditions and altitudes. I might taxi-out with a good FP reading and it would go to zero during climb; or it might be low on the ground and good in the air. I finally grounded all shields and negatives on the engine block and my problems went away.
I know about ground loops and the hassles they cause. They're often a factor in alternator or strobe noise, as the current flow through the airframe finds itself hampered by rotten ground connections or oxidized joints in the metal, and looks for other routes offering less resistance. It sometimes finds those via antenna or audio cable shields or instrument senders.

Old cars often had mechanics pulling their hair out trying to figure out why the right taillight went dim every time the left turn signal flashed. The flashing bulb's ground was bad, so the left turn signal current would find a path through the taillight filament (in the same bulb) and over to the right taillight, which was wired in parallel, backwards through the tail filament (dimming it) and thence to ground. The mechanic would fool with the dimming bulb instead of the one that appeared to be OK.

And yes, I know that electron flow is the other way.

Dan
 

kent Ashton

Well-Known Member
Joined
Aug 15, 2008
Messages
799
Location
Concord, NC
When you say this, do you mean you grounded all shields on both ends or one end?
Ground only one end of the shield for a particular wire bundle and ground that shield-end where the negative(s) for the gauge is(are) grounded. I.E., don't set up a sensor circuit with two ground points.
 

Kristoffon

Well-Known Member
Joined
Sep 4, 2009
Messages
348
Location
Brazil
As said, the crux of the problem is ground loops.

The rule becomes that if the device you're connecting to has its on ground, connect one end only. If however the device for whatever reason doesn't have its ground then you connect both ends and that becomes the way the ground gets to the device.

For example, I'm building a wooden aircraft. The antennas have to get their ground through the antenna cable from the radio so both ends of the mesh get connected. If it were a metal aircraft then of course the better ground path would be through the airframe skin and then only one side of the wire shield would be connected to ground to avoid creating a loop. A headphone will also have to get its ground through the cable (people don't conduct electricity very well after all).
 

DaveD

Well-Known Member
Joined
May 28, 2010
Messages
340
Location
Perth, Western Australia
+1 on what Kent & Kristoffon just said, but:

(people don't conduct electricity very well after all).
Actually people are pretty good conductors! Hand to foot resistance is typically around ~200 Ohm, and much less with sweaty hands and bare feet! If we were insualtors, electrocution wouldn't be a problem!

If you have hum in you audio equipment which disappears when you touch the metal chassis, it's because you have a dodgy ground which you are fixing by
using your own body...
 

Himat

Well-Known Member
Joined
May 5, 2011
Messages
2,868
Location
Norway
Ground loops in aircraft? I thought that taildragers was most prone to that.:)

More serious, it is possible to have a ground loop in a circuit at radio frequency even with one end “floating” when measuring with a DC ohm meter.

Another important thing to think about is “earth” connectors.
A long thin wire, and even a 4 square millimetre wire is in this case “thin” is no good as a radio frequency “earth” connector. It will act as an inductor at higher frequencies, blocking the high frequency current. A wide braided strap, like the ones often found in cars connecting the negative battery pole to the body, is better.
 

cloudsurfer

Member
Joined
Nov 27, 2013
Messages
13
Location
Tallahassee FL USA
Hello,
This looked like a good place to throw out my first post. Need a bit of help identifying a part and finding out where to get more, always knew it as a ground disc but Im sure theres an offical name for it. It's where your raychem or other mfgr shield termination would tie (solder) to on a panel. Have a few but they are about to be consumed on a project and I just hate not knowing where to get more of a part I used up. Size wise theyre about 0.75" in diameter.
DSC_0113.jpg
 

N804RV

Well-Known Member
Joined
Jun 9, 2013
Messages
377
Location
Mount Vernon, WA
...For anyone who knows better than I, what is the proper method for grounding shielding - is it one end or both ends, or does it depend? Does it matter if it is an EMI source vs an EMI sink? What other considerations should I be thinking about?

Thanks in advance!
There's more than one way to skin that cat. I'd say follow the wiring diagram and instructions provided by the manufacturer of the gear you are installing.
 

Aerowerx

Well-Known Member
Joined
Dec 1, 2011
Messages
5,782
Location
Marion, Ohio
As other have said, it depends.

One thing to check for is if the shield is the return signal path, as in a coaxial cable. Then it must be connected at both ends. In general, the return path should be physically close to the 'signal' path. A lot of electronics newbies make this mistake, particularly with digital signals, but it also applies to audio.

The idea mentioned of isolating a headphone jack from a panel is good, particularly if the shield carries the audio signal.

The best method I can think of is a twisted pair with two isolated shields. Connect the inner shield at both ends, and the outer shield to chassis at the source end. Of course, the device must be designed to use twisted pair.
 

Aerowerx

Well-Known Member
Joined
Dec 1, 2011
Messages
5,782
Location
Marion, Ohio
Actually people are pretty good conductors! Hand to foot resistance is typically around ~200 Ohm, and much less with sweaty hands and bare feet!
Maybe if you have a cut on your finger and foot at the same time. You can kill yourself with a flashlight battery if you know how. Blood is a good conductor. Typical skin resistance is in the millions of ohms, but it doesn't take much to break down the skin insulation and get to the blood. (Don't ask me how I know!:shock:)[/QUOTE]
 
Top