Grounding cable shields

Help Support GroupDIY:

JMan

Well-known member
Joined
Oct 26, 2019
Messages
295
Location
Chicago
As I mentioned before pin 1 of the XLR should be tied to the ground immediately where it is mounted onto the chassis. The scheme JMAN has given in his diagram is already a bad idea in terms of pin 1 earth connections.

If he is using a plastic enclosure (which is even a worse idea) then he may need to do that, and yes keeping the earth wire as short as possible would be a good design practice but not hugely important if it is a few inches longer.
I knew if I put that drawing up that everyone was going to fixate on that aspect. 😅 It was just a simple quick illustration that was partially borrowed (I just changed a few details), and since my question wasn’t about the grounding of those wires, I was hoping it wouldn’t become the focal point of the discussion. FWIW, I am grounding my xlrs at the connector, and I’m not using a plastic enclosure.

As for the screens for the signal wires used inside the case the situation is slightly different. Good practice would be to tie the screens on one end, and at a single point at the chassis/earth point near the IEC, so that all noise currents would run towards that point. The mistake would be to earth them on varying chassis points within the case. Now you have varying ground paths and that is an invitation to a problem.

This relates more directly to my original question (I don’t mind that the discussion has gone elsewhere, but I like getting opinions on what I was actually asking too).

I feel like I’m getting mixed messages in this thread, though. Based on what has been said, it seems that screened signal wires are useful for capacitive noise but not magnetic (which honestly was news to me...why have I been wearing this tinfoil hat, then? 😜 Joking!). And my AC lines — my main noise consideration — are going to be on the magnetic induction side of that, so screened wires won’t provide much useful protection. Instead, tightly twisted pairs (along with, I’m assuming, good spatial arrangement) will be the better option.

Does this mean that twisted pair shielded wire is a bad idea, though? I have plenty of that, and I figured it would provide extra (if, I guess, unnecessary) insurance. I can go either way, though.
 

Bo Deadly

Well-known member
Joined
Dec 22, 2015
Messages
2,520
Location
New Jersey, USA
But the inductance and capacitance of a wire (conductor) is another story as these will have effect on the signal, but not in the sense of reducing or increasing the wavelength, but as effect on signal attenuation.

It's not attenuation you need to worry about. The wire will actually amplify the RF at it's resonant frequency. All wires have resistance, inductance and capacitance. Below about 1kHz, resistance is dominant. But above that, the inductance is actually not completely insignificant and at RF the impedance of a short wire could be hundreds of ohms. This makes an RLC circuit that of course has a resonant frequency and as I'm sure you understand the inductive and capacitive reactance can resonate and amplify the signal by potentially quite a lot. Enough to radiate noise into the enclosure. So, as ccaudle already said, the wire is effectively an antenna that can transmit RF inside the enclosure at that resonant frequency.

There are more RF sources around us today then there ever have been so keeping those sources out is not really pedantic theory.
 
Last edited:

Bo Deadly

Well-known member
Joined
Dec 22, 2015
Messages
2,520
Location
New Jersey, USA
Possibly there was a potential for electrostatic interference, possibly the equipment designer did not understand physics very well.

Shielding of very low level signals over vaguely high impedance nets is important. It's not uncommon to have some pretty high impedance nets in old audio designs like running a small signal out to a 100K pot and back. These nets are vulnerable to imperfections in the enclosure or because RF is getting in through pin 1 of connectors as described in this thread or maybe some kind of switching noise from the supply or maybe so that the unit could be tested with the enclosure open or ...
 
Last edited:

Bo Deadly

Well-known member
Joined
Dec 22, 2015
Messages
2,520
Location
New Jersey, USA
This relates more directly to my original question (I don’t mind that the discussion has gone elsewhere, but I like getting opinions on what I was actually asking too).

My fault. Sorry. I brought up the pin 1 to ground thing. Still a good discussion to have.

I feel like I’m getting mixed messages in this thread, though. Based on what has been said, it seems that screened signal wires are useful for capacitive noise but not magnetic (which honestly was news to me...why have I been wearing this tinfoil hat, then? 😜 Joking!). And my AC lines — my main noise consideration — are going to be on the magnetic induction side of that, so screened wires won’t provide much useful protection. Instead, tightly twisted pairs (along with, I’m assuming, good spatial arrangement) will be the better option.

The screens actually do attenuate magnetic noise. Just not very well at low frequencies. If you're enclosure is all metal (and you don't have RF radiating inside because of the pin 1 thing) then you shouldn't have any RF for the screens to screen. The most likely source of noise within the enclosure is going to be magnetic in nature from rapidly switching currents from power supplies which are going to be in the range of 50-60 Hz for linear supplies 100kHz-200kHz for SMPS (which is still considered low frequency and certainly not RF). But a bad SMPS could be the variety where it switches off an inductor to generate a high voltage spike which could be electrostatic I suppose (not sure about that - I don't use or care for bad SMPS).

Does this mean that twisted pair shielded wire is a bad idea, though? I have plenty of that, and I figured it would provide extra (if, I guess, unnecessary) insurance. I can go either way, though.

It's not a bad idea. It's just not going to help because you don't really need both to twist and shield.

The twist is for cancelling equal and opposite AC generated magnetic fields emitted by the wires. But this implies that some decent current is being sloshed around. So it's largely important for supply lines but you might have a headphone amp or output transformer drive lines that could benefit from being twisted.

The shield is for shunting electromagnetic interference like hum from power cables in the walls and RF that is all around us.
 
Last edited:

ccaudle

Active member
Joined
Mar 18, 2010
Messages
44
Location
Houston
Shielding of very low level signals over vaguely high impedance nets is important

Yes, unfortunately I did not take time to write out my assumptions, which is that the question was regarding solid state electronics with relatively low driving and receiving impedance.
You can look at it as a voltage divider, the driving impedance and voltage level, the parasitic capacitance between the interference source voltage and the receiver, and the receiving impedance, and work out what the capacitively coupled interference at the receiver would be.

Not directly about analog electronics, but "High-Speed Digital Design: A Handbook of Black Magic" by Johnson and Graham has a lot of good information on crosstalk in digital circuits that can be generalized to any type of circuit once you grasp the fundamentals being presented.
Also good is "Grounding and Shielding Techniques" by Ralph Morrison and "Noise Reduction Techniques in Electronic Systems" by Henry Ott, both considered classic texts on the subject. The reprint of the June 1995 special issue of the Journal of the Audio Engineering Society just dealt with grounding and shielding concerns in audio equipment, no longer available as hard copy reprint but still available as a PDF download, definitely one to have on the bookshelf.


it seems that screened signal wires are useful for capacitive noise but not magnetic

For power line noise, which is what the original post asked about specifically. The considerations become different once the frequency is high enough to be considered RF, wherever you want to draw that line in your particular situation, usually anywhere from a few MHz upward, with a transition region between LF and RF where multiple effects are in play so you either have to cover all your bases, or decide what is most important and just concentrate on that.

Also, I was assuming typical copper type shield braid, the effectiveness of shielding against magnetic interference depends on thickness and magnetic permeability, so a thick steel conduit can be an effective magnetic shield, but a copper braid is not. Probably thick steel conduit to route your control wiring is not practical.

People have literally written entire books on the subject, so it is a little difficult to make sure to cover all the important points in a 100 word forum post.
 

JohnRoberts

Well-known member
Staff member
Moderator
Joined
Nov 30, 2006
Messages
20,482
Location
Hickory, MS
I already listed the Morrison and Ott books in the reference book post above.

One minor caveat, those books written back in the 70s were dealing with a different world for RF noise, that said the laws of physics remain unchanged.

JR
 

MisterCMRR

Well-known member
Joined
Jul 1, 2019
Messages
86
Location
Ventura, CA
Saying "wavelength of the rf depends on the wire length" is just wrong. The wavelength of RF depends only on its frequency - and the velocity of light, but that's not an issue here ;). As previously stated, wire length determines how efficient it will be as either a radiator of RF or a receiver of it.

While I'm posting, I'd like to say that one should be careful to dismiss a non-switching power supply as a source of RF interference. Many ordinary rectifier diodes will have substantial "reverse recovery time" (when the voltage across it reverses, it continues to conduct briefly - the "recovery" time). And, worse yet, many will rather instantaneously "snap" off, generating harmonics into the hundreds of kHz or higher. The wiring attached to the rectifiers will radiate both an electric and magnetic field. If you look closely at many linear power supply designs, you'll see capacitors across the diodes (10 nF is typical) - they are there to slow the turn-off transient of the diodes. I've heard some nasty-sounding "buzz" in audio gear that was completely cured by installing such capacitors. It's always best to stop this kind of stuff right where it's created - in this case the diodes. There are many more "parasitic" components and processes in electronic systems than most engineers appreciate!
 

Latest posts

Top