who can settle this argument
 

A friend is doing his own network wiring. He told me what he did and I
have serious doubts as to whether it will work at all. This is what
he's done: he ran an underground shielded cable between metal junction
boxes mounted on the outside of two buildings. The shield is grounded
at one end, not connected to anything just wrapped and taped at the
other, (this was my suggestion). So far everything sounds OK. now
here's the interesting part. He's apparently taken a twenty foot
network cable with RJ45's on either end, chopped it in half, and wire
nutted this to the underground color for color at each end. He plans
to connect this between the new computer which he hasn't received yet
and the network. I told him that he should have used a termination
block and offered that worst case this arrangement won't work at all.
Best case he'll suffer a significant loss of speed. He doesn't believe
that there will be any problem with this at all though, and if it is
it would be minor. So who's right? Can this type of cob job work, and
if so would performance be significantly compromised, and how much?
Thanks, Lenny

who can settle this argument
 

klem kedidelhopper wrote:

A friend is doing his own network wiring. He told me what he did and I
have serious doubts as to whether it will work at all. This is what
he's done: he ran an underground shielded cable between metal junction
boxes mounted on the outside of two buildings. The shield is grounded
at one end, not connected to anything just wrapped and taped at the
other, (this was my suggestion). So far everything sounds OK. now
here's the interesting part. He's apparently taken a twenty foot
network cable with RJ45's on either end, chopped it in half, and wire
nutted this to the underground color for color at each end. He plans
to connect this between the new computer which he hasn't received yet
and the network. I told him that he should have used a termination
block and offered that worst case this arrangement won't work at all.
Best case he'll suffer a significant loss of speed. He doesn't believe
that there will be any problem with this at all though, and if it is
it would be minor. So who's right? Can this type of cob job work, and
if so would performance be significantly compromised, and how much?
Thanks, Lenny


if it was that simple, people wouldn't bother with the proper
connectors. It may work, at low data rates. It may have problems with
stray RF, or other problems with reliability.

who can settle this argument
 

On Sat, 13 Oct 2012 16:35:08 -0700 (PDT), klem kedidelhopper
wrote:

A friend is doing his own network wiring.

Find a new friend that doesn't do their own network wiring.
I'll resist the temptation to crosspost this to comp.dcom.wiring.

he ran an underground shielded cable between metal junction
boxes mounted on the outside of two buildings.

How far and what type of cable? If it's shielded CAT5E or CAT6 (ScTP,
FTP, STP, S/UTP, etc)
http://en.wikipedia.org/wiki/Foiled_twisted_pair#Cable_shielding
then there's a chance it will work. If the cable is some kind of
microphone or signal cable, forget it.

The shield is grounded
at one end, not connected to anything just wrapped and taped at the
other, (this was my suggestion).

Ummm.... He's suppose to use shielded RJ45 connectors.
https://www.google.com/search?hl=en&q=shielded+rj45&tbm=isch

So far everything sounds OK.

Nope. No numbers for cable, length, or connectors.

He's apparently taken a twenty foot
network cable with RJ45's on either end, chopped it in half, and wire
nutted this to the underground color for color at each end.

Barf. That might work for slow 10BaseT/HDX (half duplex), but will
not fly for FDX (full duplex), 100Mbits/sec or Gigabit.

He plans
to connect this between the new computer which he hasn't received yet
and the network.

What's on the other end of the cable?

I told him that he should have used a termination
block and offered that worst case this arrangement won't work at all.
Best case he'll suffer a significant loss of speed. He doesn't believe
that there will be any problem with this at all though, and if it is
it would be minor. So who's right? Can this type of cob job work, and
if so would performance be significantly compromised, and how much?

If (and only if) the original cable is CAT5 or CAT6, then the easiest
fix is to put RJ45 plugs or jacks at the end points. If he needs to
add 10ft of CAT5 or CAT6 on each end, then do it with a proper patch
cable. Lose the shielded (un-necessary) and the wire nuts.

Type 66 punch down blocks sorta work at 10mbits/sec, but screw up
badly at 100mbits/sec (my experience). Type 110 punch down blocks do
better, especially when they are specifically designed for networking.
However, nothing is better than an RJ45 connector.

I don't care to speculate on exactly where it will screw up. I just
know it will fails somewhere.

Incidentally, if this is a ham operator, and the shielding is to
eliminate RFI, shielding is not going to help much. I know several
super stations that are using 2.4 and 5.7GHz Wi-Fi in place of
ethernet cable solely to eliminated RFI.

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

who can settle this argument
 

klem kedidelhopper wrote:
A friend is doing his own network wiring. He told me what he did and I
have serious doubts as to whether it will work at all. This is what
he's done: he ran an underground shielded cable between metal junction
boxes mounted on the outside of two buildings. The shield is grounded
at one end, not connected to anything just wrapped and taped at the
other, (this was my suggestion). So far everything sounds OK. now
here's the interesting part. He's apparently taken a twenty foot
network cable with RJ45's on either end, chopped it in half, and wire
nutted this to the underground color for color at each end. He plans
to connect this between the new computer which he hasn't received yet
and the network. I told him that he should have used a termination
block and offered that worst case this arrangement won't work at all.
Best case he'll suffer a significant loss of speed. He doesn't believe
that there will be any problem with this at all though, and if it is
it would be minor. So who's right? Can this type of cob job work, and
if so would performance be significantly compromised, and how much?
Thanks, Lenny

This is probably OK, if by "network" he's shooting for connecting two 9600
baud lease line modems together.

who can settle this argument
 

On Sun, 14 Oct 2012 03:22:59 +0000 (UTC), Cydrome Leader
wrote:

This is probably OK, if by "network" he's shooting for connecting two 9600
baud lease line modems together.

Chuckle. I think I know where the friend got the wire nut idea:
http://imageshack.us/photo/my-images/341/p1060034f.jpg/
https://www.google.com/search?tbm=isch&q=wiring+mess
http://www.vibrant.com/cable-messes.php
Suddenly, I have this craving for spaghetti.

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

who can settle this argument
 

Assuming this is an Ethernet * connection...

Ethernet cables (even Cat6) aren't that expensive. Why does he need to make
up his own cable?

This seems to be another example of a complex "solution" to a non-existant
problem.

who can settle this argument
 

On Sun, 14 Oct 2012 02:39:12 -0700, "William Sommerwerck"
wrote:

This seems to be another example of a complex "solution" to a non-existant
problem.

Yep, except it's not a solution because it probably will not work.
Sigh, they're everywhe
http://cheezburger.com/5816464384

Another reason why it might not work:
http://cheezburger.com/979454720
http://cheezburger.com/235715840

More problems disguised as solutions:
http://failblog.cheezburger.com/thereifixedit

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

who can settle this argument
 

Jeff Liebermann wrote:

On Sun, 14 Oct 2012 02:39:12 -0700, "William Sommerwerck"
wrote:

This seems to be another example of a complex "solution" to a non-existant
problem.

Yep, except it's not a solution because it probably will not work.
Sigh, they're everywhe
http://cheezburger.com/5816464384

Another reason why it might not work:
http://cheezburger.com/979454720
http://cheezburger.com/235715840


You've exposed the secret of why you never coil up excess network
cable. Shame on you! ;-)


More problems disguised as solutions:
http://failblog.cheezburger.com/thereifixedit


Every US tech needs at least one of these:

http://cheezburger.com/6610534656

who can settle this argument
 

On Sun, 14 Oct 2012 16:47:03 -0400, "Michael A. Terrell"
wrote:

Every US tech needs at least one of these:
http://cheezburger.com/6610534656

Hmmm.... It won't fit:
http://i380.photobucket.com/albums/oo246/gblentz/SuicidePlug.jpg

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

who can settle this argument
 

Jeff Liebermann wrote:

On Sun, 14 Oct 2012 16:47:03 -0400, "Michael A. Terrell"
wrote:

Every US tech needs at least one of these:
http://cheezburger.com/6610534656

Hmmm.... It won't fit:
http://i380.photobucket.com/albums/oo246/gblentz/SuicidePlug.jpg


You know those are only installed in lawyer's homes.

who can settle this argument
 

Jeff Liebermann wrote:
On Sun, 14 Oct 2012 03:22:59 +0000 (UTC), Cydrome Leader
wrote:

This is probably OK, if by "network" he's shooting for connecting two 9600
baud lease line modems together.

Chuckle. I think I know where the friend got the wire nut idea:
http://imageshack.us/photo/my-images/341/p1060034f.jpg/
https://www.google.com/search?tbm=isch&q=wiring+mess
http://www.vibrant.com/cable-messes.php
Suddenly, I have this craving for spaghetti.

I'm sort of numb to photos like that these days.

I really wish I was allowed to share photos of the bad stuff I've come
across.

who can settle this argument
 

Michael A. Terrell wrote:

Jeff Liebermann wrote:

On Sun, 14 Oct 2012 16:47:03 -0400, "Michael A. Terrell"
wrote:

Every US tech needs at least one of these:
http://cheezburger.com/6610534656

Hmmm.... It won't fit:
http://i380.photobucket.com/albums/oo246/gblentz/SuicidePlug.jpg


You know those are only installed in lawyer's homes.

There was actually a device called "the moving cable" at a 1990s ISP I
worked at that had a normal power plug connected to a male IEC connector.
It had the obligatory electrical tape "lump" of a splice as you'd expect
to see in something like that.

The end was shrouded so no pins were completely exposed, but the idea was
you could move computers between racks, live by supplying power into the
monitor power out socket that old power supplies had while unplugging the
normal power connector.

I never got to witness it's use, but am still surprised everybody that did
always fed power it power from the correct phase.