What type of wire for signal in?

[jneutron]

Originally Posted by VictoriaGuy

This was a new idea for me, so I did a quick bit of 'web research'.
I found this inductance calculator.

That inductance calculator is quite incorrect for all types of wiring one would consider. The return conductor is what defines the total external inductance of a wire. The internal inductance is 15 nH per foot, and completely independent of the conductor diameter.

A thinner conductor will indeed have higher external inductance, this due to the higher current density.

Originally Posted by DF96

I thought 'small' (i.e. thin?) wire has more inductance than thicker wire? But except for long lengths,....

Externally, quite accurate. Internally, invarient.

Originally Posted by gootee

Since it's a low-level signal, you'll probably want to use small wire, to keep the inductance low.

Much more important is the physical configuration. If you can't get shielded twisted pair and ground the shield to the chassis only, at the input end only, then twist two wires tightly together. The idea is to have the smallest possible "loop area" between every natural pair of conductors, everywhere, to prevent any loop from having a time-varying current induced in it by any time-varying electromagnetic fields in the air (and vice versa), avoiding receiving or transmitting hum, for each pair.
Tom

You sound like another geek I know.

Well said..

Cheers, jn

[marce]

Its just a link to his site Andrew, just use back space eh!
home page
Basicly they are the two of the most highly regarded people regarding shielding, EMC and grounding in the world, it only takes a bit of effort to register and you will be able to access a whole host of information covering what is being discussed.
At the end of the day, ALL the experts in the field recomend shielding connected at both ends for rf...

[AndrewT]

Marce.
I asked two questions.
Which tech tip? There are dozens in the list, but none jump out and say "this is me".
Register here? I cannot proceed with your second link until I have a registration code.

[marce]

The best bet with H. Ott is get his book electromagnetic compatability engineering, chapter 3 covers grounding, chapter 6 covers shielding.
You dont require a registration code for the EMC club, just sign up.
Heres one article of interest from H. Ott site.
http://www.hottconsultants.com/pdf_files/ground.pdf

[DF96]

Originally Posted by jneutron

The return conductor is what defines the total external inductance of a wire.

Yes. It was comparatively recently that I realised that the 'inductance' of an isolated piece of wire is not even well-defined, let alone calculable. I suppose it is because we can buy components with inductance labels on them, and there are formulae which give the contribution a piece of wire might make to a closed circuit (given certain assumptions).

[AndrewT]

Marce,
thanks for the OTT link.
I will read through and probably some others.

On reading the conclusion, I see I am one of the converted. I harp on about currents regularly, particularly what route the current returns by.

For every Flow there must be a Return, otherwise circuits cannot exist.

[jneutron]

Originally Posted by DF96

Yes. It was comparatively recently that I realised that the 'inductance' of an isolated piece of wire is not even well-defined, let alone calculable. I suppose it is because we can buy components with inductance labels on them, and there are formulae which give the contribution a piece of wire might make to a closed circuit (given certain assumptions).

About 14 years back, I wrote code to calculate the inductance per foot for any arbitrary cross section of wire. I also wrote code to provide visualization, that's how I created my moniker. It was set to handle 16000 wires so that I could simulate any profile using a large matrix of individual conductors.

When I built the code to calculate the inductance of only 1 conductor with no return in a thin sectional length of wire, no matter how far out in space I went, the integrated field energy kept going up. Only makes sense, as doubling the distance of each thin shell doubled it's length, and the field falls as 1/r so halved. Each shell always had the same total energy..duh..

Originally Posted by AndrewT

On reading the conclusion, I see I am one of the converted. I harp on about currents regularly, particularly what route the current returns by.

For every Flow there must be a Return, otherwise circuits cannot exist.

Yup. Very important to watch all the currents as well as all the current paths.

Cheers, jn