Shielded wiring

[Fred.Stucklen]

Again, not always true. Again, it depends on what is being powered... I would use the wing as a ground for landing lights (non-HID) and regular incadecent NAV lights. I would use the airframe to ground the incadecent tail light (with a ground wire coming from the tail light to the fuselage).
Yes, "Ground at the source" for shields means ground the shield to the airframe nearest the source of the noise. The HID frame should be connected to the wing structure, but the HID ground wire I would bring back to a ground nearer the battery/firewall IF the HID demonstrated a lot of radiated noise.... The HID power leads (+ & -) would be inside a shielded cable to prevent radiated noise. Some exposed, unshielded wire near the HID shouldn't be much of an issue. If, after wiring in this manner, I still got a lot of HID related noise in the radios, I would then put a power lead filter ( a 'PIE' network) near the HID, using the airframe for the filter ground.

Originally Posted by Dorfie

[I would then ground the shield at the end nearest to the source of potential radiated noise. In this case, the shield helps to minimize the radiated noise only. In the case of a strobe power supplies mounted inside the cabin, the strobe cable shields are generally only grounded near the power supply, not at the bulb end.]

Fred,
Thank you. Question on the above statement.
As I understand, we should never "ground" the negative return wire to the airframe and as such use the airframe for the negative. In stead use a dedicated negative return back to battery negative. (Got that) If you say "ground at the source", do you mean that the shield is contacted with the airframe near the HID? Is it contacted with the HID driver device and not the airframe? Does the HID unit stay insulated from the airframe? Out of necessity the inner wires will run exposed (without shield) for some distance. Does this distance matter?
Thanks.
Johan

[Fred.Stucklen]

Again, not always true. Again, it depends on what is being powered... I would use the wing as a ground for landing lights (non-HID) and regular incadecent NAV lights. I would use the airframe to ground the incadecent tail light (with a ground wire coming from the tail light to the fuselage).
Yes, "Ground at the source" for shields means ground the shield to the airframe nearest the source of the noise. The HID frame should be connected to the wing structure, but the HID ground wire I would bring back to a ground nearer the battery/firewall IF the HID demonstrated a lot of radiated noise.... The HID power leads (+ & -) would be inside a shielded cable to prevent radiated noise. Some exposed, unshielded wire near the HID shouldn't be much of an issue. If, after wiring in this manner, I still got a lot of HID related noise in the radios, I would then put a power lead filter (a 'PIE' network) near the HID, using the airframe for the filter ground.

Originally Posted by Dorfie

[I would then ground the shield at the end nearest to the source of potential radiated noise. In this case, the shield helps to minimize the radiated noise only. In the case of a strobe power supplies mounted inside the cabin, the strobe cable shields are generally only grounded near the power supply, not at the bulb end.]

Fred,
Thank you. Question on the above statement.
As I understand, we should never "ground" the negative return wire to the airframe and as such use the airframe for the negative. In stead use a dedicated negative return back to battery negative. (Got that) If you say "ground at the source", do you mean that the shield is contacted with the airframe near the HID? Is it contacted with the HID driver device and not the airframe? Does the HID unit stay insulated from the airframe? Out of necessity the inner wires will run exposed (without shield) for some distance. Does this distance matter?
Thanks.
Johan

[Dean_aeroleds]

A shield grounded at one end only shields the electric field component of any signals carried in the wires. If the shield is also used to carry the return current then it also helps to null out the magnetic field component since the magnetic field of the outbound and return current are equal and opposite. The only reason that the magnetic field isn't completely canceled is due to the geometry of the wire inside the shield not always producing magnetic fields centered on the same axis. Most shields have more wire strands than then wires they shield so they can handle at least as much current as the wires as long as they are terminated properly at the ends (not handing on by just a few strands).

Dean Wilkinson
CTO, AeroLEDs LLC (and B.S.E.E.)

[roee]

Originally Posted by Dean_aeroleds

A shield grounded at one end only shields the electric field component of any signals carried in the wires. If the shield is also used to carry the return current then it also helps to null out the magnetic field component since the magnetic field of the outbound and return current are equal and opposite. The only reason that the magnetic field isn't completely canceled is due to the geometry of the wire inside the shield not always producing magnetic fields centered on the same axis. Most shields have more wire strands than then wires they shield so they can handle at least as much current as the wires as long as they are terminated properly at the ends (not handing on by just a few strands).

Dean Wilkinson
CTO, AeroLEDs LLC (and B.S.E.E.)

To elaborate a bit on what Dean said:

By using the shield to carry the return current, you're essentially using the shielded wire as a coaxial cable. And yes, as Dean describes, that helps reduce interference associated with the magnetic field component (or completely eliminate it in the ideal case).

However, it should also be noted that by using the shield to carry the return current, you sacrifice some of the suppression of interference associated with the electric field component. No free lunch. And in some cases you can introduce ground loop issues as well, which is a whole 'nother can of worms.

In summary, I just wanted to clarify that there is not one configuration that works well for all possible situations. The best solution depends on the type of source and load, and its particular emissions and susceptibility characteristics. But not everyone has the background to do this kind of analysis. And even for those of us who do, it's not always entirely straight forward because there are various trade-offs involved. Sometimes only testing will reveal the real-world effectiveness of a shielding/grounding scheme.

So what's the best general advice to an RV builder? Whenever possible, simply follow the manufacturer's recommendations when installing a device. They presumably know what they're doing, and presumably have also tested the configurations they recommend. So if you're installing AeroLED lights, your best bet is to follow Dean's recommendations. If you're installing someone else's gizmo, follow their recommendations. Etc.

[ricardoramos]

I have one question on this topic:

When running a RS232 data cable, must I use a 3-wire shielded cable (TxD, RxD and ground + shielding) or can I just run a 2-wire shielded cable (TxD, RxD and shield) ? I wonder if the shield alone can be used as a signal ground connection or do I need to have a dedicated ground wire inside the cable and use the shield with it, tying both together ?

[Raymo]

Originally Posted by ricardoramos

I have one question on this topic:

When running a RS232 data cable, must I use a 3-wire shielded cable (TxD, RxD and ground + shielding) or can I just run a 2-wire shielded cable (TxD, RxD and shield) ? I wonder if the shield alone can be used as a signal ground connection or do I need to have a dedicated ground wire inside the cable and use the shield with it, tying both together ?

You probably do not need to shield this type of connection but you would want to use twisted wires to eliminate cross talk (like CAT-5/6). If you want to use shielded wire to reduce the chance noise interfering with the signals, do not use the shield as a ground. It should be grounded at one end only to dissipate the surrounding noise.

+1 on what Fred said.

[fl-mike]

Originally Posted by ricardoramos

I have one question on this topic:

When running a RS232 data cable, must I use a 3-wire shielded cable (TxD, RxD and ground + shielding) or can I just run a 2-wire shielded cable (TxD, RxD and shield) ? I wonder if the shield alone can be used as a signal ground connection or do I need to have a dedicated ground wire inside the cable and use the shield with it, tying both together ?

You want a dedicated shield. The signal ground is carrying the return currents for the data signals. RS232 is not a transformer coupled differential interface like 10/100Base-T Ethernet. If you use twisted pair, twist each signal with a ground, i.e. TxD/gnd, RxD/gnd with overall shield terminated at both ends. Otherwise, TxD, RxD, gnd, overall shield terminated at both ends.