You might take a look at the (US) National Electrical Code, Tables
310-16 thru -19.  Basically, in the power world the limitations on a
conductor's amperage are based on max allowable voltage drop, and
maximum allowable temperature to limit damage to the insulation. 2-3
percent is usually the max used in design.  Code allows a max of 5% from
electrical service entrance entrance to final load. Multiple NEC tables
are provided due to the different insulation types available and
environmental installation, such as conduit, direct earth burial or
open-air, all of which affect the cooling rate.  But luckily your only
concern is likely voltage drop, and at milliamp amperage is likely nil.
For example, per Table 8, Conductor Properties (NEC 1996) copper
conductors have the following DC resistance per thousand feet.  From
this you can calculate your voltage drop based on amp draw:

#18 AWG (1620 Circ. Mills)  -  7 strands -  8.45 ohms per k feet
#16 AWG (2580 Circ. Mills)  -  7 strands -  5.29 ohms per k feet
#14 AWG )4110 Circ. Mills)  -  7 strands -  3.26 ohms per k feet

I guess my second life as an "electrical engineer" after switching from
"electronic engineer" may be of *some* help here after all <grin>.

"Eisermann, Phil" wrote:
>
>         [snip circular mil stuff]
>
> > I have not seen "circular" units used before, is this common practice
> > when dealing with wire gauges?
> >
>         hard for me to say if it is "common".  In terms of a mfg.
>         specifying it in catalogs... probably not. Not in any of
>         the ones i have, at any rate. I run into it every now
>         and then. Pressman uses it in his book on SMPS. I have
>         seen it in some sort of training manual for electricians
>         (I think thats what it was) I have a document downloaded
>         from somewhere on the internet that gives wire data,
>         which includes circular mils. It's probably about as common
>         as "slugs" for MechE's. rare, but you still see it every
>         now and then. At least, that's my impression.
>
>         anyone else that's heard of (or uses!) circular mils
>         care to join in?
>
>         it is a nice convenience to use. Since circular mils are
>         already in units of "diameter circles" (not sure how to say
>         this properly), it makes it easy to calculate required wire
>         size. if i need 20 circular mils, i just take the square root
>         to get the diameter. If you're blessed not to have to work
>         in the english system, you could do the same thing with
>         millimeters, to get circular millimeters *grin*  But in the
>         metric system, i think it is more common to use amps per
>         square centimeter (eg there is no equivalent to
>         "circular mils" Not that i know of, anyway).