RE: on topic this time :)

From: Donald Becker (becker@cesdis1.gsfc.nasa.gov)
Date: Fri Jan 07 2000 - 18:26:24 EST


On Fri, 7 Jan 2000, Chris Knipe wrote:

> On Thu, 6 Jan 2000, John LeMay wrote:
>
> >What if I tried to use RG6 or RG59 (~72 ohm impedence as opposed to 50 ohm
> >in RG8 and RG58) This makes little difference in the RF world. Is it
> >different in the network world?
>
> Apparently after just looking at the IEEE standard, and a little (well
> actually rather big) book about networking, THINNET (by the IEEE standard)
> supports the following...
>
> Cable Media that is supported (Thinnet):
>
> Cable Impedance Description
> RG-58/U 50 Ohm Solid Copper Center
> RG-58 A/U 50 Ohm Wire Strand Center
> RG-58 C/U 50 Ohm Military Version Of RG-58 A/U
> RG-59 75 Ohm Cable TV Wire
> RG-62 93 Ohm ARCnet Specification

I'll squash this before it goes too far.

Coax Ethernet only runs with 50ohm termination. Otherwise the collision
levels won't be right and you'll either get no collisions or a "collision"
with every transmission.

Ethernet is only reliable if you use 50ohm cable as well. A short length of
random coax cable might work for very small networks, but the longer the
segment the less likely it is to work. There is a Natsemi App Note on using
other cable impedences, but that involves modifying the transceiver hardware
and no one does it.

10base2 "thinnet" transceivers are electrically exactly the same as 10base5
transceivers. Only the connectors are different. The difference in
permitted length is because 10base5 cable has lower DC resistance (but the
same 50ohm AC impedance) and much better shielding. (10base5 is *really*
overkill on the shielding -- the only danger from a nuclear blast is melting
the cable ;-)

You can probably push 10base5 to 700meters if you have very few connectors.

> THINNET can also RELIABLY only transmit for 185 meters (about 610 feet).

Further with better cable and fewer attachments. Not even that far with
corrosion on the connectors.

Donald Becker
Scyld Computing Corporation, and
USRA-CESDIS, becker@cesdis.gsfc.nasa.gov

-
To unsubscribe from this list: send the line "unsubscribe linux-net" in
the body of a message to majordomo@vger.rutgers.edu



This archive was generated by hypermail 2b29 : Sat Jan 15 2000 - 21:00:28 EST