Cable shield as a Faraday cage

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I take the approach that a floating screen makes things unpredictable / chaotic. Actual effect depends on numerous other factors - electrical balance / impedances / mechanical orientation etc...etc...
In fact any floating conductor can be problematic in a system of any complexity and likely cost you time / money in any lab testing .
But a faraday cage can be floating, or does it work better if it's connected to the circuit's reference so the whole system floats at the reference point of the cage?
 
But a faraday cage can be floating, or does it work better if it's connected to the circuit's reference so the whole system floats at the reference point of the cage?

In practical terms - tie it to the low impedance fixed potential source - usually the reference.
The idea of a floating cage is interesting but bear in mind that nothing is really "free floating" at rf due to parasitic / unintended / leakage etc - basically all that stuff that you don't draw on the schematic :)
 
Member
Joined 2014
Paid Member
Yes the cage can be floating (think aircraft, car, battery powered audio recorder) with respect to 'ground' but as rmaudio says its better that there is some reference it's all tied to. Once I have finished the current build phase I will only have one mains powered unit (power amp) in the entire chain, so only a single PE connection. In theory I could have everything else ungrounded.

Think modern mass market audio equipment that is double insulated so only has 2 core mains. No 'ground' in the traditional sense.
 
Tony and Neils paper is entirely worth the time and is very much on the money.

Note that a Faraday cage is very much a skin depth dependent thing, they only really become effective once the thickness of the metal exceeds a few skin depths, so good for RF, not so much for 50Hz, except that at 50Hz you are typically so electrically small that you are thinking electric or magnetic fields rather then electromagnetic fields.

A Faraday cage also by definition must ENTIRELY surround whatever is being screened, bringing a pair of wires into and out of it via a pair of holes makes it do something, but it is no longer as simple as a Faraday cage.

Low frequency E fields are easy to screen, while low frequency H fields are best addressed with twisting the pair, faraday cages are effective once you get fast enough that the effects become electromagnetic rather then easily separable over the length of your cable.

Regards, Dan.
 
Tony and Neils paper is entirely worth the time and is very much on the money.

Note that a Faraday cage is very much a skin depth dependent thing, they only really become effective once the thickness of the metal exceeds a few skin depths, so good for RF, not so much for 50Hz, except that at 50Hz you are typically so electrically small that you are thinking electric or magnetic fields rather then electromagnetic fields.

A Faraday cage also by definition must ENTIRELY surround whatever is being screened, bringing a pair of wires into and out of it via a pair of holes makes it do something, but it is no longer as simple as a Faraday cage.

Low frequency E fields are easy to screen, while low frequency H fields are best addressed with twisting the pair, faraday cages are effective once you get fast enough that the effects become electromagnetic rather then easily separable over the length of your cable.

Regards, Dan.

Neil ? Do you mean Keith or is Neil another of his names ?

Yeah - I'm thinking 'rf' wrt this. Mains and low frequency is something of a different ball game as you point out.

Take the point that Faraday Cage is continuous. But I think we realise that we are talking something imperfect but approaching it once we need to take wires in/out. Same for ventilation / mesh etc.
 
Take the point that Faraday Cage is continuous. But I think we realise that we are talking something imperfect but approaching it once we need to take wires in/out. Same for ventilation / mesh etc.
Yes and no, what is meant by continuous? A Faraday cage can be made from mesh. So, surely if there were holes it wouldn't be a Faraday shield but still a cage, the bandwidth of the shielding being determined by the hole size?
 
And take a look at this - not the quickest read but worth it:

Bonding Cable Shields at Both Ends to Reduce Noise
Yes, very interesting, thank you.

Regards single ended use I found this interesting:

"Most single-ended voltage-driven unbalanced signals in cables have a source impedance of £100W, so the capacitive coupling is insignificant compared with the inductive noise coupling, which in turn is insignificant when compared with the end-to-end shield voltage caused by shield current (remember, the shield is in the signal’s return path, so series noise in the shield = series noise in the signal). A 30m long cable with 1A in the shield will suffer around 10mV of inductively coupled noise at 50Hz, but with a typical shield resistance of 0.5W the voltage generated in the shield will be 0.5V."

In my initial post I was wondering about using STP cable, and billsherv mentioned it could be better than coax. It seems this would be a way to separate the shield current from the signal return path (assuming there is one of course ;))
 
That article was about Pro Audio
The last one is by Bill Whitlock and explains a number of things well, including hybrid shielding. Remember whilst reading this that balanced does not mean differential drive, just that the impedance is the same on both legs.
Yes, I read those articles some time ago but it was good to read them again and understand a little bit more. In the last one connecting the screen at the sending end only is advocated, this setup is also tested in the pro audio article mentioned above and is only slightly worse than connecting it at both ends. This would keep currents out of the screen but would also mean that the Faraday cage connection between enclosures would be broken unless there was a hybrid connection that end I presume?
It's a shame they didn't test with it floating..............
 
Last edited:
Member
Joined 2014
Paid Member
The hybrid connection is connected at RF both ends, but DC at one end. Now domestic audio works fine with some seriously bad interconnects (say Kimber) so there is a strong argument that, unless you are running 50ft leads you need not worry. But if you sweat the small stuff it's worth experimenting.

As Bruno points outs, much of the benefit can be had by just putting a second braid on the cable and reducing its DCR resistance.

And there are always transformers, line transformers being seriously marmite in the home audio world :)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.