A properly balanced connection has a twisted pair of strands. The reason is that now magnetic interference is canceled out by the twisting.
The shield has to go directly to the chassis and the chassis and signal ground should be connected in 1 place only.
If you do not meat these criteria, you simply have an unbalanced connection.
One important thing to remember is that a balanced connection needs to have the same impedance for both positive and negative signals or else CMRR will suffer. That means that the dimentions of the cable are far more critical than with unbalanced cables.
The shield has to go directly to the chassis and the chassis and signal ground should be connected in 1 place only.
If you do not meat these criteria, you simply have an unbalanced connection.
One important thing to remember is that a balanced connection needs to have the same impedance for both positive and negative signals or else CMRR will suffer. That means that the dimentions of the cable are far more critical than with unbalanced cables.
Last edited:
Last edited:
While it is far preferable that the + and - signal wires be twisted together inside of a single shield, as far as I know, it is not required to meet the definition of "balanced". All it means, as far as I know, is that the + and - signal wires have equal impedance, equal voltage and exactly opposite phase.
I have the copper data and the silver data.
The 3 references given do not show the silver oxide data.
The 3 references given do not show the silver oxide data.
That seems to contradict with Rane's techincal notes, there's a link earlier in the thread.
That's what I thought too... But there's a point here having done some wiki reading...
http://en.wikipedia.org/wiki/Balanced_audio said:
This suggests that by having each line much further apart than in the normal twisted pair configuration, each line could accept different interference. And that's something that a balanced signal can't compensate for. Hmm...
Lets not go too OT about this. It doesn't really matter to me how conductive silver oxide is. So long as the silver is clean clean and oxide free on both the plug and socket when they are connected together, the only silver that will oxidise is the area exposed to air. The tiny metal to metal contact areas where the plug and socket touch, and the actual electrical transfer takes place are not exposed to air and will remain oxide-free, so long as the connectors aren't moved around.
A coax with the screen disconnected at one end is merely a screened single wire - it won't act like real coax because there is no return current.
A balanced connection which is not twisted will not cancel out magnetically induced interference. You could end up worse off than a simple unbalanced connection!
A balanced connection which is not twisted will not cancel out magnetically induced interference. You could end up worse off than a simple unbalanced connection!
One other note that might help a little: Twisting of the conductor pairs doesn't, in and of itself, result in interference cancellation, which only requires that the same average interference be applied to both conductors (and that the receiving end be differential). But it does help greatly to encourage interference to be more same. But so does coax, and as DF96 has mentioned already, what's being proposed is *not* coax.
BTW, the conductivity of silver is only slightly different from copper, less than 10%.
All good fortune,
Chris
BTW, the conductivity of silver is only slightly different from copper, less than 10%.
All good fortune,
Chris
Perhaps I need to approach this a different way.
The thing I don't like about coax is that the return path, and hence some part of the signal; must travel along the screen. The screen is a poor path for signal to use, as it is usually made up of a matrix of wires that are sometimes plated, and the shortest path for the current to follow is probably a combination of through the strands of the braid as well as going across thousands the juction of one strand touching another. Theory tells me that can't be as good as one straight continuous conduction contained witin a signal-independent screen.
A balanced signal overcomes this issue, but I can't find silver XLR connectors that are affordable (>1000$ for a few M). Solid silver RCA on the other hand, are very affordable.
If I can find an affordable silver twisted pair coax at least, maybe I could connect silver 2 RCA plugs either end. One core to each centre pin for signal and ground. I'm thinking if I strip say a good 4 inches off the jacket at the end, split the shield into two, twist it round each core as they split off, then heatshrink the two; I can connect 2 RCA's. Then connect the shield, chassis and signal ground only to the source. That way the only direct connection between the source and amplifier would be through 2 solid silver cores, as the screen is disconnected at the receiving end.
The thing I don't like about coax is that the return path, and hence some part of the signal; must travel along the screen. The screen is a poor path for signal to use, as it is usually made up of a matrix of wires that are sometimes plated, and the shortest path for the current to follow is probably a combination of through the strands of the braid as well as going across thousands the juction of one strand touching another. Theory tells me that can't be as good as one straight continuous conduction contained witin a signal-independent screen.
A balanced signal overcomes this issue, but I can't find silver XLR connectors that are affordable (>1000$ for a few M). Solid silver RCA on the other hand, are very affordable.
If I can find an affordable silver twisted pair coax at least, maybe I could connect silver 2 RCA plugs either end. One core to each centre pin for signal and ground. I'm thinking if I strip say a good 4 inches off the jacket at the end, split the shield into two, twist it round each core as they split off, then heatshrink the two; I can connect 2 RCA's. Then connect the shield, chassis and signal ground only to the source. That way the only direct connection between the source and amplifier would be through 2 solid silver cores, as the screen is disconnected at the receiving end.
Not entirely correct: A balanced connection only needs to have equal impedance. But equal voltage and opposite phase is not needed, because the receiving end is a difference amplifier. So it reacts only to the difference of the positive and negative sides. That means that the positive can have 0 voltages on it and the negative side can have the audio signal and it will work perfectly (if the impedances are matched witch is not that eazy).
But the purpose of a balanced connection is to isolate the audio signal from its surroundings. Iaw nothing must be able to interfere with the signal. Only a twisted pair of wires can eliminate magnetic interference and the shield eliminates electrical interference. If you don´t want that just use unbalanced connections.
An externally hosted image should be here but it was not working when we last tested it.
Not according to this picture.
How many XLR connectors do you need? Does that "M" mean a million or meter? The term "connector" means only the connector and not the wire so your statement is confusing.
Do you mean solid silver twisted pair wire? I've never looked for that so don't know what's out there. I like Canare Star Quad microphone wire for interconnects.
You can check out this:
Shielded Twisted Pair Controlled Impedance Wire
It's silver plated though.
The best way that I know of to use twisted pair with single ended is to use the wires as the + and ground for the RCA connector, and connect the shield to ground at one end only.
Yes it does. It reduces the interference by roughly ensuring that the tiny loops formed by the send and return wires are of both polarities. The aim is to get no net loop area so nothing for magnetic flux to link to. The two wires can't be in the same place, because of insulation thickness, so two wires lying next to each other but untwisted would still have a small but non-zero area between them. Twisting is the simplest way to solve this. Next simplest is coaxial. After that is four wires, in opposite pairs.Chris Hornbeck said:
This thread has piqued my interest in silver oxide. It is interesting that there is next to nothing about its electrical properties on the web. So far this is all I've found that seems to directly mention it: Physical properties of silver oxide thin films by pulsed laser deposition: effect of oxygen pressure during growth and http://www.physics.iitm.ac.in/research_files/synopsis/Ravichandraraju.pdf
Last edited:
If you are worried about using the screen as signal return then use a screened twisted pair. Use the pair for signal, and the screen as a screen. A twisted pair does not have to be used balanced. If screens were as bad as you fear, then they would not work too well for RF.
Last edited:
Did you look in the CRC Handbook of Chemistry and Physics? That's pretty much the standard reference for this sort of thing.
Dirk95100 said:
Perhaps I've misinterpreted the picture. To me that looks like at the balanced outputs and inputs, consisiting of 3 connections (+, -, and G): G, at both ends is connected to chassis ground and shield.
Does it in fact show (as the only reference to signal ground is on the input side) that the signal ground for the output is deliberately omitted and hence left completely unconnected?
Sorry about that, I made a couple of confusing mistakes in that post. I meant to say connectors and wire but I had found some affordable solid silver twisted pair, so now the only issue is the lack of affordable solid silver XLR connectors and sockets.
And yes, I meant solid silver twisted pair wire.
I assume you're suggesting that screens aren't as bad as I fear, because screens do in fact work very well for RF. Do you have any reference to show how RF over coax compares to over RF over twisted pair shielded to substantiate this?
Regarding silver oxide, if it's really a concern that the metals on both connector and socket will oxidise over time and drop in conductivity, you could coat the whole lot in silicone grease before connecting to prevent oxidisation over most if not all of the contact surfaces.
Last edited:
There are several silver oxides, so which one are we talking about? Some are conductive and others are not.
- Status
- Not open for further replies.