Is shielded wire necessary for chipamp inputs?

[ransom peek]

Is it important to use shielded wire for the signal inputs on a chipamP? If so, then will twisting the two wires be equally effective, or is it worth the effort to secure shielded cable?

At these low audio frequencies, it seems that shielding may not be necessary.

[peranders]

If you use a shielded cable you don't have to think about it but it will probably do with a twisted wire.

[PMA]

Originally posted by ransom peek


At these low audio frequencies, it seems that shielding may not be necessary.

Unfortunately these low audio frequencies love to be influenced by high frequencies like EMI and RFI, especially in non-linear amplifying components and circuits.

[poobah]

The need for shielding depends on what you plan to use for an enclosure. If you have a full metal enclosure, you can probably get by without shielding.

The purpose of a shield is to block electrostatic fields from contaminating your signal (this energy is capacitively coupled into your circuit). The need for a shield is determined by the strength of your signal, the strength of the garbage fields floating around, and the level of signal quality you wish to maintain. Generally, the smaller the signal, like a microphone or a phono cartridge, the more attention you should pay to shielding. Most would agree that without a full metal enclosure your inputs should be shielded. There are also those that will insist your wires float in free space, are insulated with calendared rhino horn, and are surrounded by tons of exotic metals... shields are cheap.

While capacitive coupling does increase with frequency, you should not dismiss the need for shielding just because audio is low frequency.

Twisting wires is whole 'nother story. Moving (changing) magnetic fields will induce voltage in a wire. Twisting in done to mitigate the effects of stray magnetic fields. Were you to X-Ray a twisted pair of wires, you would see many small "eyes" or loops. The magnetic field "cutting" these loops induces voltage. Without any twisting, a magnetic field would see one large loop, and it will induce a voltage. When the wires are twisted the fields see many small loops... the trick is that half of these loops are reversed. So a whole series of voltages will be induced, but; half of them are of opposite polarity and they tend cancel out. Is it perfect? No. Does it work? You betcha! There are many exceptions of course.

This street go both ways... wires are also twisted so that they don't radiate a large coherent magnetic filed. This is done in AC wiring throughout most things.

Should you twist your wires? Sometimes, looks cool too!



EDIT: Why is it so hard to type in little boxes???????

[davidlzimmer]

Here is my 2 cents:

If you build the power supply in the same enclosure, I would use sheilding. (Groung the shield on one end)

If in separate enclosure, very small slightly twisted conducters work fine.

[ransom peek]

OK, using shielding, how can you ground at only one end. Each of the 2 inputs comes in on an rca jack, and need the center signal and the ground (outer case of rca jack) to the briangt lm3875 amp board.

The amp board has two connection points to solder for each of the two lines, one is signal, one is ground.

I understand that if you ground the shield at both ends you cna have ground loops, but how do you make the two lead connections for each channel if the shield is not used as the ground?

[davidlzimmer]

You can ground at both ends. I just meant to ground at least one end.

[davidlzimmer]

The circuit ground should be to the star ground. Do not depend on the sheilding to ground the PCB.

[quasi]

Without exception I have always achieved better results with shielded cable for line level inputs. But it's easy to try both and make up your own mind.

[davidlzimmer]

There ya go