On Oct 21, 2024, karl@??? wrote:
> Dan Purgert:
> ...
> > > (When I look at a pinout diagram on ethernet cable data is basically
> > > run on a pair so that gives me access to data trans and receive + a
> > > power cable and at 18 ga its far larger than most data cables which
> > > are at 24 ga iirc.)
> >
> > Not really sure what you're trying to convey here -- "ethernet(tm)"
> > pinout isn't going to come into play, unless you're planning to hook
> > this up to a 10/100/1000BASE-T computer network.
>
> You need to know where the pairs are, so you don't end up with signal
> on e.g. blue+green/white, which gives you worse signal integrety. You
> want to keep the pairs intact and the pinout gives you that.

Getting the right layout at either end is certainly "important" -- but
following the TIA-568B (or -568A) layout used for 10/100/1000BASE-T
isn't necessarily the *correct* answer for the application at hand,
assuming one is attempting to use some form of existing breakout rather
than rolling their own.

For example, I've come across some that have an RS485 pinout to the
effect of either

    1,4 -- Tx+/-
    5,8 -- Rx+/- 
    2,3,6,7, -- No Connection

OR

1,2 - VCC (e.g. 3 or 5 volts)
3,4 - TX+/-
5,6 - Rx+/-
7,8 - GND

The first one was (IIRC) one of several alternatives (incl an RS232 with
hardware flow-control) from some vendor of I want to say PLCs. The
second one was an abomination from something like a wannabe arduino or
some such (as in, they were going for vendor lockin with their cabling).


--
|_|O|_|
|_|_|O| Github: https://github.com/dpurgert
|O|O|O| PGP: DDAB 23FB 19FA 7D85 1CC1 E067 6D65 70E5 4CE7 2860