In article ,
Ignoramus31310 wrote:

On 2010-07-03, Karl Townsend wrote:

"Ignoramus31310" wrote in message
...
I need to crimp a lot of Molex open barrel terminals for those servo
power supplies. I was kind of shocked to find a tool from Molex for
$269. But then I found item HWS16166 from

http://www.phoenixent.com/

It does a great job.

i

I've got one that's like a pair of pliers. I've been disapointed. Most
crimps are great, then I get one where the wire slides back out of the
crimp
after a while. I started pulling on the just crimped connecter to test them
and it mostly solved the issue.

I will use 20 gauge only for all control connections inside the
cabinet. It seems to work well on 20ga. I also tried it with 24 ga and
it seemed good also.

I always pull on every crimp I make.

Most of the trouble is with 22 guage and smaller. Have you tried this tool
with this fine a wire and checked if they will pull out?

It seems OK, but I will definitely check every crimp.

One uses different terminals and crimp dies for the various wire sizes.

The non-ratcheting pliers type crimpers are almost useless.

Joe Gwinn

On 2010-07-03, Joseph Gwinn wrote:
In article ,
Ignoramus31310 wrote:

On 2010-07-03, Karl Townsend wrote:

"Ignoramus31310" wrote in message
...
I need to crimp a lot of Molex open barrel terminals for those servo
power supplies. I was kind of shocked to find a tool from Molex for
$269. But then I found item HWS16166 from

http://www.phoenixent.com/

It does a great job.

i

I've got one that's like a pair of pliers. I've been disapointed. Most
crimps are great, then I get one where the wire slides back out of the
crimp
after a while. I started pulling on the just crimped connecter to test them
and it mostly solved the issue.

I will use 20 gauge only for all control connections inside the
cabinet. It seems to work well on 20ga. I also tried it with 24 ga and
it seemed good also.

I always pull on every crimp I make.

Most of the trouble is with 22 guage and smaller. Have you tried this tool
with this fine a wire and checked if they will pull out?

It seems OK, but I will definitely check every crimp.

One uses different terminals and crimp dies for the various wire sizes.

The non-ratcheting pliers type crimpers are almost useless.

This one is a ratcheting, safety type. It would not release until I
completed the crimp.

i

We use the Molex universal hand crimper (non-ratcheting) for small
production work, on the small MiniFit Jr. pins and the bigger old-style
ones. Molex part number is 638111000, at Mouser for $49 each.
Non-ratcheting but has lots of sizes so you can handle any wire size in a
given pin. You also have to do the insulation crimp as a second op, so not
as fast but very versatile.

-----
Regards,
Carl Ijames

"Ignoramus31310" wrote in message
...
On 2010-07-03, Joseph Gwinn wrote:
In article ,
Ignoramus31310 wrote:

On 2010-07-03, Karl Townsend wrote:

"Ignoramus31310" wrote in
message
...
I need to crimp a lot of Molex open barrel terminals for those servo
power supplies. I was kind of shocked to find a tool from Molex for
$269. But then I found item HWS16166 from

http://www.phoenixent.com/

It does a great job.

i

I've got one that's like a pair of pliers. I've been disapointed. Most
crimps are great, then I get one where the wire slides back out of the
crimp
after a while. I started pulling on the just crimped connecter to test
them
and it mostly solved the issue.

I will use 20 gauge only for all control connections inside the
cabinet. It seems to work well on 20ga. I also tried it with 24 ga and
it seemed good also.

I always pull on every crimp I make.

Most of the trouble is with 22 guage and smaller. Have you tried this
tool
with this fine a wire and checked if they will pull out?

It seems OK, but I will definitely check every crimp.

One uses different terminals and crimp dies for the various wire sizes.

The non-ratcheting pliers type crimpers are almost useless.

This one is a ratcheting, safety type. It would not release until I
completed the crimp.

i

In article ,
Ignoramus31310 wrote:

On 2010-07-03, Joseph Gwinn wrote:
In article ,
Ignoramus31310 wrote:

On 2010-07-03, Karl Townsend wrote:

"Ignoramus31310" wrote in message
...
I need to crimp a lot of Molex open barrel terminals for those servo
power supplies. I was kind of shocked to find a tool from Molex for
$269. But then I found item HWS16166 from

http://www.phoenixent.com/

It does a great job.

i

I've got one that's like a pair of pliers. I've been disapointed. Most
crimps are great, then I get one where the wire slides back out of the
crimp
after a while. I started pulling on the just crimped connecter to test
them
and it mostly solved the issue.

I will use 20 gauge only for all control connections inside the
cabinet. It seems to work well on 20ga. I also tried it with 24 ga and
it seemed good also.

I always pull on every crimp I make.

Most of the trouble is with 22 guage and smaller. Have you tried this
tool
with this fine a wire and checked if they will pull out?

It seems OK, but I will definitely check every crimp.

One uses different terminals and crimp dies for the various wire sizes.

The non-ratcheting pliers type crimpers are almost useless.

This one is a ratcheting, safety type. It would not release until I
completed the crimp.

That's good. But it sounds like you don't have adequate match between wire
size, terminal type (wire size acceptance range), and crimper dies. All must
match.

And, crimping is for stranded wire, not solid wire. That said, in a pinch I
solder the terminal to the solid wire.

I have also soldered stranded wire into terminals. The largest example was when
I was attaching big copper terminals to some AWG #4 wire, to power a lighting
panel for a stage. I was a teenager, and did not have the humongous and
expensive crimper. I needed to use a small terminal, as there was not physical
space for the usual mechanical wire-clamp terminal.

The terminal was made from copper tubing, one end being flattened and punched to
accept the large terminal stud, the other end being an open cup. So, I held the
terminal in a vise and soldered the big wire into the cup with a propane torch
and plumbers' solder and Nocorrode grease flux, just like soldering copper pipe.
The school maintenance folk were surprised at this approach, but soldered
connections were (and are still) acceptable under the National Electrical Code.

Joe Gwinn

On 2010-07-04, Joseph Gwinn wrote:
In article ,
Ignoramus31310 wrote:

[ ... ]

This one is a ratcheting, safety type. It would not release until I
completed the crimp.

That's good. But it sounds like you don't have adequate match between wire
size, terminal type (wire size acceptance range), and crimper dies. All must
match.

Amen! That is why I have a bucket (plastic cat litter bucket)
full of crimpers -- just for terminals -- and not counting the hydraulic
ones for 8 ga and larger.

I've got a similar bucket full of air tools. One of the
benefits of having cats. :-)

And, crimping is for stranded wire, not solid wire. That said, in a pinch I
solder the terminal to the solid wire.

I have also soldered stranded wire into terminals.

With the hazard that if it is subject to vibration, it is far
more likely to fail -- especially if you don't have and use anti-wicking
tweezers.

The largest example was when
I was attaching big copper terminals to some AWG #4 wire, to power a lighting
panel for a stage. I was a teenager, and did not have the humongous and
expensive crimper. I needed to use a small terminal, as there was not physical
space for the usual mechanical wire-clamp terminal.

The terminal was made from copper tubing, one end being flattened and punched to
accept the large terminal stud, the other end being an open cup. So, I held the
terminal in a vise and soldered the big wire into the cup with a propane torch
and plumbers' solder and Nocorrode grease flux, just like soldering copper pipe.

That sounds good -- though a proper eutectic lead/tin solder
would have been better with a rosin flux -- and a solder pot to dip the
wire into to tin it first. Was this AWG #4 wire solid or stranded?

The school maintenance folk were surprised at this approach, but soldered
connections were (and are still) acceptable under the National Electrical Code.

Certainly under the NEC -- but they are not acceptable with
aerospace connections. Crimp terminals are by far preferred there.

Enjoy,
DoN.

--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

In article ,
"DoN. Nichols" wrote:

On 2010-07-04, Joseph Gwinn wrote:
In article ,
Ignoramus31310 wrote:

[ ... ]

This one is a ratcheting, safety type. It would not release until I
completed the crimp.

That's good. But it sounds like you don't have adequate match between wire
size, terminal type (wire size acceptance range), and crimper dies. All
must
match.

Amen! That is why I have a bucket (plastic cat litter bucket)
full of crimpers -- just for terminals -- and not counting the hydraulic
ones for 8 ga and larger.

I've got a similar bucket full of air tools. One of the
benefits of having cats. :-)

And, crimping is for stranded wire, not solid wire. That said, in a pinch
I
solder the terminal to the solid wire.

I have also soldered stranded wire into terminals.

With the hazard that if it is subject to vibration, it is far
more likely to fail -- especially if you don't have and use anti-wicking
tweezers.

All true, but in a pinch ...


The largest example was
when
I was attaching big copper terminals to some AWG #4 wire, to power a
lighting
panel for a stage. I was a teenager, and did not have the humongous and
expensive crimper. I needed to use a small terminal, as there was not
physical
space for the usual mechanical wire-clamp terminal.

The terminal was made from copper tubing, one end being flattened and
punched to
accept the large terminal stud, the other end being an open cup. So, I
held the
terminal in a vise and soldered the big wire into the cup with a propane
torch
and plumbers' solder and Nocorrode grease flux, just like soldering copper
pipe.

That sounds good -- though a proper eutectic lead/tin solder
would have been better with a rosin flux -- and a solder pot to dip the
wire into to tin it first.

I was a teenager doing a one-off job. Well, two-off. So, no reason to get all
that nice kit. I did pre-tin the wire and the cup before sweating wire to cup.


Was this AWG #4 wire solid or stranded?

Stranded.


The school maintenance folk were surprised at this approach, but soldered
connections were (and are still) acceptable under the National Electrical
Code.

Certainly under the NEC -- but they are not acceptable with
aerospace connections. Crimp terminals are by far preferred there.

Absolutely. Crimp is far more reliable, especially under vibration.

Solder-type coax connectors are quite unreliable - the heating-cooling cycle
causes the big nut that clamps the shield to unscrew over time, causing open
shields. I found this out when diagnosing an unreliable Xyplex satistical
multiplexer (connects multiple VT100 terminals to a VAX, in the days before
ethernet became practical). Many of the BNC connectors had open shield
connections. The solution was to cut all the solder BNC connectors off the
RG-58 cable and to install crimped BNC connectors in their place. If I recall,
we used AMP tooling and connectors. Anyway, problem solved, almost overnight.

Joe Gwinn

On 2010-07-04, Joseph Gwinn wrote:
In article ,
"DoN. Nichols" wrote:

On 2010-07-04, Joseph Gwinn wrote:

[ ... ]

And, crimping is for stranded wire, not solid wire. That said, in a pinch
I
solder the terminal to the solid wire.

I have also soldered stranded wire into terminals.

With the hazard that if it is subject to vibration, it is far
more likely to fail -- especially if you don't have and use anti-wicking
tweezers.

All true, but in a pinch ...

Understood. And where you were doing it -- you would not have
much vibration -- unless perhaps for a performance of "Riverdance". :-)

[ ... ]

The terminal was made from copper tubing, one end being flattened and
punched to
accept the large terminal stud, the other end being an open cup. So, I
held the
terminal in a vise and soldered the big wire into the cup with a propane
torch
and plumbers' solder and Nocorrode grease flux, just like soldering copper
pipe.

That sounds good -- though a proper eutectic lead/tin solder
would have been better with a rosin flux -- and a solder pot to dip the
wire into to tin it first.

I was a teenager doing a one-off job. Well, two-off. So, no reason to get all
that nice kit.

No reason -- and probably no resources either. :-)

I did pre-tin the wire and the cup before sweating wire to cup.

Good. Though for that size wire, a solder pot (and a pre-dip in
the rosin flux) makes for a more thorough tinning job.

Was this AWG #4 wire solid or stranded?

Stranded.

So the pre-tinning goes without saying then. :-)

The school maintenance folk were surprised at this approach, but soldered
connections were (and are still) acceptable under the National Electrical
Code.

Certainly under the NEC -- but they are not acceptable with
aerospace connections. Crimp terminals are by far preferred there.

Absolutely. Crimp is far more reliable, especially under vibration.

Solder-type coax connectors are quite unreliable - the heating-cooling cycle
causes the big nut that clamps the shield to unscrew over time, causing open
shields. I found this out when diagnosing an unreliable Xyplex satistical
multiplexer (connects multiple VT100 terminals to a VAX, in the days before
ethernet became practical). Many of the BNC connectors had open shield
connections. The solution was to cut all the solder BNC connectors off the
RG-58 cable and to install crimped BNC connectors in their place. If I recall,
we used AMP tooling and connectors. Anyway, problem solved, almost overnight.

Hmm ... various styles available. I've got two which could do
the job.

The first one does a separate crimp for the shield termination
and for the center conductor pin. This is more likely to be used for
BNC style connectors.

The other crimps both the shield and the center conductor at the
same time. There are two small windows which the crimper reaches in
through to crimp the center conductor. But these are normally for
insert coax pins to go in block terminals -- and D-series connectors
like the 13W3 used by Sun for monitor connections. (I have been looking
for the connector inserts for a long time -- just to have a few which
work with that crimper. :-)

Have you ever worked with the shield termination ferrules used
for daisy-chaining a bunch of shields together and ending with a
standard insulated wire to get a crimp-on pin for going into a block
where the shields are not truly coax, but rather things like shielded
twisted pair for low level signals but not RF frequencies? The crimper
is the 59000 IIRC, with a whole series of interchangeable dies for
different shield diameters. When you find them on eBay they may or may
not have dies in them -- but the vendor almost never tells you which
dies are installed. (Identified by color stripes to match the color of
the ferrule which fits.) Purple for RG-174 (skinny coax cable) and
similar size single conductor shielded (and perhaps some very small
shielded twisted pair), and other colors for larger sizes, with brown
and green being the ones I used most often.

Enjoy,
DoN.

--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

On Sun, 04 Jul 2010 11:05:14 -0400, Joseph Gwinn
wrote:

And, crimping is for stranded wire, not solid wire. That said, in a pinch I
solder the terminal to the solid wire.

Which means that the wire will, sooner or later, break right at the
terminal.

Gunner

One could not be a successful Leftwinger without realizing that,
in contrast to the popular conception supported by newspapers
and mothers of Leftwingers, a goodly number of Leftwingers are
not only narrow-minded and dull, but also just stupid.
Gunner Asch

On 2010-07-05, Joseph Gwinn wrote:
In article ,
"DoN. Nichols" wrote:

On 2010-07-04, Joseph Gwinn wrote:

[ ... ]

we used AMP tooling and connectors. Anyway, problem solved, almost
overnight.

Hmm ... various styles available. I've got two which could do
the job.

The first one does a separate crimp for the shield termination
and for the center conductor pin. This is more likely to be used for
BNC style connectors.

The other crimps both the shield and the center conductor at the
same time. There are two small windows which the crimper reaches in
through to crimp the center conductor. But these are normally for
insert coax pins to go in block terminals -- and D-series connectors
like the 13W3 used by Sun for monitor connections. (I have been looking
for the connector inserts for a long time -- just to have a few which
work with that crimper. :-)

I used what the department had available, and any AMP crimp system was going to
solve the problem.

Indeed -- if you had both the right crimp tools and terminals.
I tend to have crimpers for terminals which I've never found, and
vice-versa. :-)


Have you ever worked with the shield termination ferrules used
for daisy-chaining a bunch of shields together and ending with a
standard insulated wire to get a crimp-on pin for going into a block
where the shields are not truly coax, but rather things like shielded
twisted pair for low level signals but not RF frequencies?
is the 59000 IIRC, with a whole series of interchangeable dies for

Never had the pleasure.

Particularly beats having to unweave the braid, twist it into
tails, slide on insulation, and the solder (or crimp) to pins in the
connector. Particularly when there are a half-dozen shields to be
terminated in one pin. :-) Each ferrule has two bulges (pre-crimping) to
accept two wires -- either out the back along the jacketed wire or out
the front to the connector pins.

But I do recall a problem in the late 1970s when a coax-cable computer link
between buildings stopped working, even though there was continuity. Turned out
that an electrician had cut the cable (don't recall if accidental or not) and
spliced it back together with a pair of small wire nuts. The data signals
bounced right off that impedance step. The poor man did not understand why we
were laughing so hard, but sensed that he was somehow involved.


*Big* smile!

Yes, I can imagine him not understanding it at all. He works
with 60 Hz, and no runs long enough to be near a quarter wavelength or
anything else significant. :-)

The ohmmeter says it is good -- so it must be good. :-)

Thanks,
DoN.

--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

In article ,
"DoN. Nichols" wrote:

On 2010-07-05, Joseph Gwinn wrote:
In article ,
"DoN. Nichols" wrote:

On 2010-07-04, Joseph Gwinn wrote:

[ ... ]

we used AMP tooling and connectors. Anyway, problem solved, almost
overnight.

Hmm ... various styles available. I've got two which could do
the job.

The first one does a separate crimp for the shield termination
and for the center conductor pin. This is more likely to be used for
BNC style connectors.

The other crimps both the shield and the center conductor at the
same time. There are two small windows which the crimper reaches in
through to crimp the center conductor. But these are normally for
insert coax pins to go in block terminals -- and D-series connectors
like the 13W3 used by Sun for monitor connections. (I have been looking
for the connector inserts for a long time -- just to have a few which
work with that crimper. :-)

I used what the department had available, and any AMP crimp system was
going to
solve the problem.

Indeed -- if you had both the right crimp tools and terminals.
I tend to have crimpers for terminals which I've never found, and
vice-versa. :-)

This was in a big company, so we by chance did have a matching set for RG-58.


Have you ever worked with the shield termination ferrules used
for daisy-chaining a bunch of shields together and ending with a
standard insulated wire to get a crimp-on pin for going into a block
where the shields are not truly coax, but rather things like shielded
twisted pair for low level signals but not RF frequencies?
is the 59000 IIRC, with a whole series of interchangeable dies for

Never had the pleasure.

Particularly beats having to unweave the braid, twist it into
tails, slide on insulation, and the solder (or crimp) to pins in the
connector. Particularly when there are a half-dozen shields to be
terminated in one pin. :-) Each ferrule has two bulges (pre-crimping) to
accept two wires -- either out the back along the jacketed wire or out
the front to the connector pins.

But I do recall a problem in the late 1970s when a coax-cable computer link
between buildings stopped working, even though there was continuity.
Turned out
that an electrician had cut the cable (don't recall if accidental or not)
and
spliced it back together with a pair of small wire nuts. The data signals
bounced right off that impedance step. The poor man did not understand why
we
were laughing so hard, but sensed that he was somehow involved.


*Big* smile!

Yes, I can imagine him not understanding it at all. He works
with 60 Hz, and no runs long enough to be near a quarter wavelength or
anything else significant. :-)

The ohmmeter says it is good -- so it must be good. :-)

We did not try to explain it to him. It would be a magic show at best/

I have had the need to explain such things to other electricians, and gotten a
big fight for my trouble. A person who does not know what he does not know, or
even that there are such things.

Joe Gwinn

On 7/5/2010 9:47 PM, Joseph Gwinn wrote:
In ,
"DoN. wrote:

On 2010-07-05, Joseph wrote:
In ,
"DoN. wrote:

On 2010-07-04, Joseph wrote:

[ ... ]

we used AMP tooling and connectors. Anyway, problem solved, almost
overnight.

Hmm ... various styles available. I've got two which could do
the job.

The first one does a separate crimp for the shield termination
and for the center conductor pin. This is more likely to be used for
BNC style connectors.

The other crimps both the shield and the center conductor at the
same time. There are two small windows which the crimper reaches in
through to crimp the center conductor. But these are normally for
insert coax pins to go in block terminals -- and D-series connectors
like the 13W3 used by Sun for monitor connections. (I have been looking
for the connector inserts for a long time -- just to have a few which
work with that crimper. :-)

I used what the department had available, and any AMP crimp system was
going to
solve the problem.

Indeed -- if you had both the right crimp tools and terminals.
I tend to have crimpers for terminals which I've never found, and
vice-versa. :-)

This was in a big company, so we by chance did have a matching set for RG-58.


Have you ever worked with the shield termination ferrules used
for daisy-chaining a bunch of shields together and ending with a
standard insulated wire to get a crimp-on pin for going into a block
where the shields are not truly coax, but rather things like shielded
twisted pair for low level signals but not RF frequencies?
is the 59000 IIRC, with a whole series of interchangeable dies for

Never had the pleasure.

Particularly beats having to unweave the braid, twist it into
tails, slide on insulation, and the solder (or crimp) to pins in the
connector. Particularly when there are a half-dozen shields to be
terminated in one pin. :-) Each ferrule has two bulges (pre-crimping) to
accept two wires -- either out the back along the jacketed wire or out
the front to the connector pins.

But I do recall a problem in the late 1970s when a coax-cable computer link
between buildings stopped working, even though there was continuity.
Turned out
that an electrician had cut the cable (don't recall if accidental or not)
and
spliced it back together with a pair of small wire nuts. The data signals
bounced right off that impedance step. The poor man did not understand why
we
were laughing so hard, but sensed that he was somehow involved.


*Big* smile!

Yes, I can imagine him not understanding it at all. He works
with 60 Hz, and no runs long enough to be near a quarter wavelength or
anything else significant. :-)

The ohmmeter says it is good -- so it must be good. :-)

We did not try to explain it to him. It would be a magic show at best/

I have had the need to explain such things to other electricians, and gotten a
big fight for my trouble. A person who does not know what he does not know, or
even that there are such things.

That's one thing that bugs me about the State of Connecticut, one _must_
have an electrician's license to work on network cables. Never seen an
electrician with said license yet who could identify a cable scanner,
let alone who owned one,.

In article ,
"J. Clarke" wrote:

On 7/5/2010 9:47 PM, Joseph Gwinn wrote:
In ,
"DoN. wrote:

On 2010-07-05, Joseph wrote:
In ,
"DoN. wrote:

On 2010-07-04, Joseph wrote:

[ ... ]

we used AMP tooling and connectors. Anyway, problem solved, almost
overnight.

Hmm ... various styles available. I've got two which could do
the job.

The first one does a separate crimp for the shield termination
and for the center conductor pin. This is more likely to be used for
BNC style connectors.

The other crimps both the shield and the center conductor at the
same time. There are two small windows which the crimper reaches in
through to crimp the center conductor. But these are normally for
insert coax pins to go in block terminals -- and D-series connectors
like the 13W3 used by Sun for monitor connections. (I have been looking
for the connector inserts for a long time -- just to have a few which
work with that crimper. :-)

I used what the department had available, and any AMP crimp system was
going to
solve the problem.

Indeed -- if you had both the right crimp tools and terminals.
I tend to have crimpers for terminals which I've never found, and
vice-versa. :-)

This was in a big company, so we by chance did have a matching set for
RG-58.


Have you ever worked with the shield termination ferrules used
for daisy-chaining a bunch of shields together and ending with a
standard insulated wire to get a crimp-on pin for going into a block
where the shields are not truly coax, but rather things like shielded
twisted pair for low level signals but not RF frequencies?
is the 59000 IIRC, with a whole series of interchangeable dies for

Never had the pleasure.

Particularly beats having to unweave the braid, twist it into
tails, slide on insulation, and the solder (or crimp) to pins in the
connector. Particularly when there are a half-dozen shields to be
terminated in one pin. :-) Each ferrule has two bulges (pre-crimping) to
accept two wires -- either out the back along the jacketed wire or out
the front to the connector pins.

But I do recall a problem in the late 1970s when a coax-cable computer
link
between buildings stopped working, even though there was continuity.
Turned out
that an electrician had cut the cable (don't recall if accidental or not)
and
spliced it back together with a pair of small wire nuts. The data
signals
bounced right off that impedance step. The poor man did not understand
why
we
were laughing so hard, but sensed that he was somehow involved.


*Big* smile!

Yes, I can imagine him not understanding it at all. He works
with 60 Hz, and no runs long enough to be near a quarter wavelength or
anything else significant. :-)

The ohmmeter says it is good -- so it must be good. :-)

We did not try to explain it to him. It would be a magic show at best.

I have had the need to explain such things to other electricians, and
gotten a
big fight for my trouble. A person who does not know what he does not
know, or
even that there are such things.

That's one thing that bugs me about the State of Connecticut, one _must_
have an electrician's license to work on network cables. Never seen an
electrician with said license yet who could identify a cable scanner,
let alone who owned one,.

Yes. An iron rice bowl: http://en.wikipedia.org/wiki/Iron_rice_bowl.

Joe Gwinn

On 2010-07-07, Joseph Gwinn wrote:
In article ,
"J. Clarke" wrote:

On 7/5/2010 9:47 PM, Joseph Gwinn wrote:
In ,
"DoN. wrote:

[ ... ]

The ohmmeter says it is good -- so it must be good. :-)

We did not try to explain it to him. It would be a magic show at best.

I have had the need to explain such things to other electricians, and
gotten a
big fight for my trouble. A person who does not know what he does not
know, or
even that there are such things.

That's one thing that bugs me about the State of Connecticut, one _must_
have an electrician's license to work on network cables. Never seen an
electrician with said license yet who could identify a cable scanner,
let alone who owned one,.

Yes. An iron rice bowl: http://en.wikipedia.org/wiki/Iron_rice_bowl.

Does this apply to:

1) Network cabling outdoors?

2) Network cabling in commercial or apartment buildings?

3) Or even home wiring?

When our house was being expanded, I ran some thicknet ethernet
cable (picked up big spools of it at a hamfest) between several places
where networked computers are sometimes set up. Had I known about
100BaseT and faster coming down the pike, I would have put in Cat-5
cable instead.

Hmm ... also -- does it apply when running ethernet through
fiber optics? No wires there at all! No excuse for needing someone who
is accustomed to wiring voltages around. :-)

And how about today's WiFi?

Granted, I have a friend with an old house where it does not
work well at all. The walls are not drywall, but rather real plaster
over metal mesh. But I can't imagine an electrician knowing enough
about that to even diagnose the problem. :-)

Enjoy,
DoN.

--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

On Tue, 06 Jul 2010 22:05:30 -0400, the renowned Joseph Gwinn
wrote:

In article ,
"J. Clarke" wrote:

On 7/5/2010 9:47 PM, Joseph Gwinn wrote:
In ,
"DoN. wrote:

On 2010-07-05, Joseph wrote:
In ,
"DoN. wrote:

On 2010-07-04, Joseph wrote:

[ ... ]

we used AMP tooling and connectors. Anyway, problem solved, almost
overnight.

Hmm ... various styles available. I've got two which could do
the job.

The first one does a separate crimp for the shield termination
and for the center conductor pin. This is more likely to be used for
BNC style connectors.

The other crimps both the shield and the center conductor at the
same time. There are two small windows which the crimper reaches in
through to crimp the center conductor. But these are normally for
insert coax pins to go in block terminals -- and D-series connectors
like the 13W3 used by Sun for monitor connections. (I have been looking
for the connector inserts for a long time -- just to have a few which
work with that crimper. :-)

I used what the department had available, and any AMP crimp system was
going to
solve the problem.

Indeed -- if you had both the right crimp tools and terminals.
I tend to have crimpers for terminals which I've never found, and
vice-versa. :-)

This was in a big company, so we by chance did have a matching set for
RG-58.


Have you ever worked with the shield termination ferrules used
for daisy-chaining a bunch of shields together and ending with a
standard insulated wire to get a crimp-on pin for going into a block
where the shields are not truly coax, but rather things like shielded
twisted pair for low level signals but not RF frequencies?
is the 59000 IIRC, with a whole series of interchangeable dies for

Never had the pleasure.

Particularly beats having to unweave the braid, twist it into
tails, slide on insulation, and the solder (or crimp) to pins in the
connector. Particularly when there are a half-dozen shields to be
terminated in one pin. :-) Each ferrule has two bulges (pre-crimping) to
accept two wires -- either out the back along the jacketed wire or out
the front to the connector pins.

But I do recall a problem in the late 1970s when a coax-cable computer
link
between buildings stopped working, even though there was continuity.
Turned out
that an electrician had cut the cable (don't recall if accidental or not)
and
spliced it back together with a pair of small wire nuts. The data
signals
bounced right off that impedance step. The poor man did not understand
why
we
were laughing so hard, but sensed that he was somehow involved.


*Big* smile!

Yes, I can imagine him not understanding it at all. He works
with 60 Hz, and no runs long enough to be near a quarter wavelength or
anything else significant. :-)

The ohmmeter says it is good -- so it must be good. :-)

We did not try to explain it to him. It would be a magic show at best.

I have had the need to explain such things to other electricians, and
gotten a
big fight for my trouble. A person who does not know what he does not
know, or
even that there are such things.

That's one thing that bugs me about the State of Connecticut, one _must_
have an electrician's license to work on network cables. Never seen an
electrician with said license yet who could identify a cable scanner,
let alone who owned one,.

Yes. An iron rice bowl: http://en.wikipedia.org/wiki/Iron_rice_bowl.

Joe Gwinn

Gotta "protect the public", you know.


Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com

In article ,
"DoN. Nichols" wrote:

On 2010-07-07, Joseph Gwinn wrote:
In article ,
"J. Clarke" wrote:

On 7/5/2010 9:47 PM, Joseph Gwinn wrote:
In ,
"DoN. wrote:

[ ... ]

The ohmmeter says it is good -- so it must be good. :-)

We did not try to explain it to him. It would be a magic show at best.

I have had the need to explain such things to other electricians, and
gotten a
big fight for my trouble. A person who does not know what he does not
know, or
even that there are such things.

That's one thing that bugs me about the State of Connecticut, one _must_
have an electrician's license to work on network cables. Never seen an
electrician with said license yet who could identify a cable scanner,
let alone who owned one,.

Yes. An iron rice bowl: http://en.wikipedia.org/wiki/Iron_rice_bowl.

Does this apply to:

1) Network cabling outdoors?

2) Network cabling in commercial or apartment buildings?

3) Or even home wiring?

Depends on the local laws, I suppose. One should not be expecting a real
technical rationale.


When our house was being expanded, I ran some thicknet ethernet
cable (picked up big spools of it at a hamfest) between several places
where networked computers are sometimes set up. Had I known about
100BaseT and faster coming down the pike, I would have put in Cat-5
cable instead.

Hmm ... also -- does it apply when running ethernet through
fiber optics? No wires there at all! No excuse for needing someone who
is accustomed to wiring voltages around. :-)

Hmm. This could be fun! Well, it looks like a wire....


And how about today's WiFi?

Clearly, does not apply.


Granted, I have a friend with an old house where it does not
work well at all. The walls are not drywall, but rather real plaster
over metal mesh. But I can't imagine an electrician knowing enough
about that to even diagnose the problem. :-)

Eight or nine years ago, I paid an electrician to install the CAT5e cable from
basement to my wife's office on the second floor, and I will say that the
electrician and his helper earned their pay on that one. It's an old house,
with wood lath plaster and strategically-placed bricks as fire-stops. It took
hours, and they were working all the while.

I then terminated and connected the installed cable. It works just fine.

Installing that wire probably cost what a full WiFi setup would have cost, but
once installed the wire just works, needs no sysadmin effort, has no security
drama, and has far greater data capacity than the WiFi of the day. Only
recently have wireless LANs even approached the theoretical datarate of a CAT5e
wire. In practice, wireless LANs rarely achieve anything like the capacity of a
cable.

If I ran cable today, it would be CAT7, which supports gigabit datarates, far
exceeding any likely wireless LAN technology. The CAT7 wire is expensive to be
sure, but installation costs will swamp the wire cost.

But the current 100 megabits per second is more than sufficient.


Actually, the best thing to install is plastic conduit, if one has the
opportunity to do it reasonably easily, like if the walls are already open for
some other reason.

I recall lots of ads from various wire manufacturers touting their latest
(fastest, expensive) kinds of cable, saying that companies should buy the better
stuff even if not strictly needed today, as "future-proofing". But, conduit is
the true "future-proofing", as one can always upgrade the wire within, and
easily. Nor does one need to guess which direction the technology will go and
which wire it will need. For some reason, the ads didn't mention the conduit
option.


Joe Gwinn

On Jul 5, 11:00*pm, "J. Clarke" wrote:
On 7/5/2010 9:47 PM, Joseph Gwinn wrote:





In ,
* "DoN. *wrote:

On 2010-07-05, Joseph *wrote:
In ,
* "DoN. *wrote:

On 2010-07-04, Joseph *wrote:

* * * *[ ... ]

we used AMP tooling and connectors. *Anyway, problem solved, almost
overnight.

* * * *Hmm ... various styles available. *I've got two which could do
the job.

* * * *The first one does a separate crimp for the shield termination
and for the center conductor pin. *This is more likely to be used for
BNC style connectors.

* * * *The other crimps both the shield and the center conductor at the
same time. *There are two small windows which the crimper reaches in
through to crimp the center conductor. *But these are normally for
insert coax pins to go in block terminals -- and D-series connectors
like the 13W3 used by Sun for monitor connections. *(I have been looking
for the connector inserts for a long time -- just to have a few which
work with that crimper. :-)

I used what the department had available, and any AMP crimp system was
going to
solve the problem.

* * * *Indeed -- if you had both the right crimp tools and terminals.
I tend to have crimpers for terminals which I've never found, and
vice-versa. :-)

This was in a big company, so we by chance did have a matching set for RG-58.

* * * *Have you ever worked with the shield termination ferrules used
for daisy-chaining a bunch of shields together and ending with a
standard insulated wire to get a crimp-on pin for going into a block
where the shields are not truly coax, but rather things like shielded
twisted pair for low level signals but not RF frequencies?
is the 59000 IIRC, with a whole series of interchangeable dies for

Never had the pleasure.

* * * *Particularly beats having to unweave the braid, twist it into
tails, slide on insulation, and the solder (or crimp) to pins in the
connector. *Particularly when there are a half-dozen shields to be
terminated in one pin. :-) Each ferrule has two bulges (pre-crimping) to
accept two wires -- either out the back along the jacketed wire or out
the front to the connector pins.

But I do recall a problem in the late 1970s when a coax-cable computer link
between buildings stopped working, even though there was continuity.
Turned out
that an electrician had cut the cable (don't recall if accidental or not)
and
spliced it back together with a pair of small wire nuts. *The data signals
bounced right off that impedance step. The poor man did not understand why
we
were laughing so hard, but sensed that he was somehow involved.

* * * **Big* smile!

* * * *Yes, I can imagine him not understanding it at all. *He works
with 60 Hz, and no runs long enough to be near a quarter wavelength or
anything else significant. :-)

* * * *The ohmmeter says it is good -- so it must be good. :-)

We did not try to explain it to him. *It would be a magic show at best/

I have had the need to explain such things to other electricians, and gotten a
big fight for my trouble. * A person who does not know what he does not know, or
even that there are such things.

That's one thing that bugs me about the State of Connecticut, one _must_
have an electrician's license to work on network cables. *Never seen an
electrician with said license yet who could identify a cable scanner,
let alone who owned one,.- Hide quoted text -

- Show quoted text -

NJ has an extortion scheme where you can "apply" for an exemption for
low-voltage network & telecom wiring. This self-certification costs
$120. In over 30 years of doing this work, nobody has ever asked me
for a certificate. Maybe I'll get one someday.

Regarding Iggy's original post about the crimper - there is simply no
excuse for not doing it right. The tools are expensive, but having the
right crimper for the right terminal for the right wire is the ONLY
way to ensure a reliable termination. "Pulling" on the terminal after
crimping is pretty subjective. When I was in production engineering,
we spent a lot of time and money measuring crimps with specialized
micrometers and pulling on them with calibrated scales.

OTOH, when I was at AT&T, I was impressed by the union electricians I
worked with on DSLAM installations. They had the right tools for the
job and knew how to use them, in contrast to some of the equipment
they were installing, some of which had crimps that looked like they
were done with vise-grips.

On 7/7/2010 8:30 AM, Joseph Gwinn wrote:
In ,
"DoN. wrote:

On 2010-07-07, Joseph wrote:
In ,
"J. wrote:

On 7/5/2010 9:47 PM, Joseph Gwinn wrote:
In ,
"DoN. wrote:

[ ... ]

The ohmmeter says it is good -- so it must be good. :-)

We did not try to explain it to him. It would be a magic show at best.

I have had the need to explain such things to other electricians, and
gotten a
big fight for my trouble. A person who does not know what he does not
know, or
even that there are such things.

That's one thing that bugs me about the State of Connecticut, one _must_
have an electrician's license to work on network cables. Never seen an
electrician with said license yet who could identify a cable scanner,
let alone who owned one,.

Yes. An iron rice bowl:http://en.wikipedia.org/wiki/Iron_rice_bowl.

Does this apply to:

1) Network cabling outdoors?

2) Network cabling in commercial or apartment buildings?

3) Or even home wiring?

Depends on the local laws, I suppose. One should not be expecting a real
technical rationale.

In Connecticut if you get paid to touch premise wiring you need the
license unless you work for the phone company, which has a specific
exemption. That includes network cabling outdoors, network cabling in
commercial or apartment buildings, and home wiring. Homeowners in most
areas can do their own wiring subject to permitting and inspection.

When our house was being expanded, I ran some thicknet ethernet
cable (picked up big spools of it at a hamfest) between several places
where networked computers are sometimes set up. Had I known about
100BaseT and faster coming down the pike, I would have put in Cat-5
cable instead.

Hmm ... also -- does it apply when running ethernet through
fiber optics? No wires there at all! No excuse for needing someone who
is accustomed to wiring voltages around. :-)

Hmm. This could be fun! Well, it looks like a wire....

The law is specific about the definition of "electrical work" and it
looks like fiber is exempted. So is wiring under 24v used for
controlling lawn sprinklers--if you have a network-aware law sprinkler
attached to your LAN does that mean that you're exempt?

And how about today's WiFi?

Clearly, does not apply.

The definition is: (2) "Electrical work" means the installation,
erection, maintenance, alteration or repair of any wire, cable, conduit,
busway, raceway, support, insulator, conductor, appliance, apparatus,
fixture or equipment that generates, transforms, transmits or uses
electrical energy for light, heat, power or other purposes, but does not
include low voltage wiring, not exceeding twenty-four volts, used within
a lawn sprinkler system;"

One could argue wifi either way.

Granted, I have a friend with an old house where it does not
work well at all. The walls are not drywall, but rather real plaster
over metal mesh. But I can't imagine an electrician knowing enough
about that to even diagnose the problem. :-)

Eight or nine years ago, I paid an electrician to install the CAT5e cable from
basement to my wife's office on the second floor, and I will say that the
electrician and his helper earned their pay on that one. It's an old house,
with wood lath plaster and strategically-placed bricks as fire-stops. It took
hours, and they were working all the while.

I then terminated and connected the installed cable. It works just fine.

Installing that wire probably cost what a full WiFi setup would have cost, but
once installed the wire just works, needs no sysadmin effort, has no security
drama, and has far greater data capacity than the WiFi of the day. Only
recently have wireless LANs even approached the theoretical datarate of a CAT5e
wire. In practice, wireless LANs rarely achieve anything like the capacity of a
cable.

Never. 802.11N has a theoretical maximum throughput of around 450
Mb/sec, CAT5E handles 1000.

If I ran cable today, it would be CAT7, which supports gigabit datarates,

CAT 5e supports gigabit just fine. You don't need CAT7 for, well,
anything. 10 gig Ethernet runs on 6A. 7 is something that the cable
manufacturers want you to think you need, but you don't unless you have
unusual circumstances.

far
exceeding any likely wireless LAN technology. The CAT7 wire is expensive to be
sure, but installation costs will swamp the wire cost.

But the current 100 megabits per second is more than sufficient.

If it's not, you can get gigabit NICs for 20 bucks.

Actually, the best thing to install is plastic conduit, if one has the
opportunity to do it reasonably easily, like if the walls are already open for
some other reason.

I recall lots of ads from various wire manufacturers touting their latest
(fastest, expensive) kinds of cable, saying that companies should buy the better
stuff even if not strictly needed today, as "future-proofing".

And it never was. CAT5 needed to be recertified to 5E to run gigabit
(most CAT5 installations pass 5E, but they weren't tested to it until it
became part of the requirement). CAT6 didn't meet the 10G spec, leading
to 6A. 7 is a solution in search of a problem.

But, conduit is
the true "future-proofing", as one can always upgrade the wire within, and
easily. Nor does one need to guess which direction the technology will go and
which wire it will need. For some reason, the ads didn't mention the conduit
option.


Joe Gwinn

On 2010-07-07, Joseph Gwinn wrote:
In article ,
"DoN. Nichols" wrote:

On 2010-07-07, Joseph Gwinn wrote:
In article ,
"J. Clarke" wrote:

[ ... ]

That's one thing that bugs me about the State of Connecticut, one _must_
have an electrician's license to work on network cables. Never seen an
electrician with said license yet who could identify a cable scanner,
let alone who owned one,.

[ ... ]

Does this apply to:

1) Network cabling outdoors?

2) Network cabling in commercial or apartment buildings?

3) Or even home wiring?

Depends on the local laws, I suppose. One should not be expecting a real
technical rationale.

I was asking in particular about the Connecticut situation which
J. Clarke mentioned.

[ ... ]

Hmm ... also -- does it apply when running ethernet through
fiber optics? No wires there at all! No excuse for needing someone who
is accustomed to wiring voltages around. :-)

Hmm. This could be fun! Well, it looks like a wire....

Might be fun to take it to court. :-)

[ ... ]

Granted, I have a friend with an old house where it does not
work well at all. The walls are not drywall, but rather real plaster
over metal mesh. But I can't imagine an electrician knowing enough
about that to even diagnose the problem. :-)

Eight or nine years ago, I paid an electrician to install the CAT5e cable from
basement to my wife's office on the second floor, and I will say that the
electrician and his helper earned their pay on that one. It's an old house,
with wood lath plaster and strategically-placed bricks as fire-stops. It took
hours, and they were working all the while.

Sounds like you are thankful that *you* did not try to do it
yourself. :-)

I then terminated and connected the installed cable. It works just fine.

Of course.

Installing that wire probably cost what a full WiFi setup would have cost, but
once installed the wire just works, needs no sysadmin effort, has no security
drama, and has far greater data capacity than the WiFi of the day. Only
recently have wireless LANs even approached the theoretical datarate of a CAT5e
wire. In practice, wireless LANs rarely achieve anything like the capacity of a
cable.

Of course. I've got a wireless bridge running to a friend's
house across the street and down one house -- using circular waveguide
antennas which I machined up. Encryption enabled, good aggressive
firewall on each end, and all logins from one side to the other via ssh.

If I ran cable today, it would be CAT7, which supports gigabit datarates, far
exceeding any likely wireless LAN technology. The CAT7 wire is expensive to be
sure, but installation costs will swamp the wire cost.

But the current 100 megabits per second is more than sufficient.

Certainly. I'm running my home directories from a server on
another machine, and the speed is quite adequate for most things. If I
want to run find(1) on the home directory, I do that on the server
itself, of course.

Actually, the best thing to install is plastic conduit, if one has the
opportunity to do it reasonably easily, like if the walls are already open for
some other reason.

Hmm ... specifically plastic? Is there a problem running
twisted-pair ethernet in close proximity to a metal surround? I know
that some of the network wiring we installed at work was shielded
four twisted pair cable. Or is plastic conduit simply less expensive?

I recall lots of ads from various wire manufacturers touting their latest
(fastest, expensive) kinds of cable, saying that companies should buy the better
stuff even if not strictly needed today, as "future-proofing". But, conduit is
the true "future-proofing", as one can always upgrade the wire within, and
easily. Nor does one need to guess which direction the technology will go and
which wire it will need. For some reason, the ads didn't mention the conduit
option.

:-)

Enjoy,
DoN.

--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

On Wed, 07 Jul 2010 07:10:03 -0400, Spehro Pefhany
wrote:


Gotta "protect the public", you know.


Best regards,
Spehro Pefhany
Translate to "protect the union"
Gerry :-)}
London, Canada

On 7/7/2010 9:29 PM, Gerald Miller wrote:
On Wed, 07 Jul 2010 07:10:03 -0400, Spehro Pefhany
wrote:


Gotta "protect the public", you know.


Best regards,
Spehro Pefhany
Translate to "protect the union"

That's pretty well the way it works. You can't get in the club unless
you do a two year apprenticeship or show 6 years "equivalent
experience", which would presumably be military or telco.

In article ,
"DoN. Nichols" wrote:

On 2010-07-07, Joseph Gwinn wrote:
In article ,
"DoN. Nichols" wrote:

On 2010-07-07, Joseph Gwinn wrote:
In article ,
"J. Clarke" wrote:

[ ... ]

[snip]

Granted, I have a friend with an old house where it does not
work well at all. The walls are not drywall, but rather real plaster
over metal mesh. But I can't imagine an electrician knowing enough
about that to even diagnose the problem. :-)

Eight or nine years ago, I paid an electrician to install the CAT5e cable
from
basement to my wife's office on the second floor, and I will say that the
electrician and his helper earned their pay on that one. It's an old
house,
with wood lath plaster and strategically-placed bricks as fire-stops. It
took
hours, and they were working all the while.

Sounds like you are thankful that *you* did not try to do it
yourself. :-)

Absolutely. I've pulled my share of wire over the years, and old houses are the
worst.


I then terminated and connected the installed cable. It works just fine.

Of course.

Installing that wire probably cost what a full WiFi setup would have cost,
but
once installed the wire just works, needs no sysadmin effort, has no
security
drama, and has far greater data capacity than the WiFi of the day. Only
recently have wireless LANs even approached the theoretical datarate of a
CAT5e
wire. In practice, wireless LANs rarely achieve anything like the capacity
of a
cable.

Of course. I've got a wireless bridge running to a friend's
house across the street and down one house -- using circular waveguide
antennas which I machined up. Encryption enabled, good aggressive
firewall on each end, and all logins from one side to the other via ssh.

I'm assuming that you added the waveguide antennas to standard wireless LAN
equipment.


If I ran cable today, it would be CAT7, which supports gigabit datarates,
far
exceeding any likely wireless LAN technology. The CAT7 wire is expensive
to be
sure, but installation costs will swamp the wire cost.

But the current 100 megabits per second is more than sufficient.

Certainly. I'm running my home directories from a server on
another machine, and the speed is quite adequate for most things. If I
want to run find(1) on the home directory, I do that on the server
itself, of course.

Actually, the best thing to install is plastic conduit, if one has the
opportunity to do it reasonably easily, like if the walls are already open
for
some other reason.

Hmm ... specifically plastic? Is there a problem running
twisted-pair ethernet in close proximity to a metal surround? I know
that some of the network wiring we installed at work was shielded
four twisted pair cable. Or is plastic conduit simply less expensive?

CAT5 and CAT5e are not shielded, so running in steel conduit may cause
unexpected losses and reflections. But I have not tried this. CAT7 is
shielded, so running in steel conduit would have no effect. I don't recall if
CAT6 is shielded, but I think it is unshielded. J.Clarke may know more.

But my reason for plastic was economic, not technical.

Joe Gwinn

In article , "J. Clarke"
wrote:

On 7/7/2010 8:30 AM, Joseph Gwinn wrote:
In ,
"DoN. wrote:

On 2010-07-07, Joseph wrote:
In ,
"J. wrote:

On 7/5/2010 9:47 PM, Joseph Gwinn wrote:
In ,
"DoN. wrote:

[ ... ]

The ohmmeter says it is good -- so it must be good. :-)

We did not try to explain it to him. It would be a magic show at best.

I have had the need to explain such things to other electricians, and
gotten a
big fight for my trouble. A person who does not know what he does not
know, or
even that there are such things.

That's one thing that bugs me about the State of Connecticut, one _must_
have an electrician's license to work on network cables. Never seen an
electrician with said license yet who could identify a cable scanner,
let alone who owned one,.

Yes. An iron rice bowl:http://en.wikipedia.org/wiki/Iron_rice_bowl.

Does this apply to:

1) Network cabling outdoors?

2) Network cabling in commercial or apartment buildings?

3) Or even home wiring?

Depends on the local laws, I suppose. One should not be expecting a real
technical rationale.

In Connecticut if you get paid to touch premise wiring you need the
license unless you work for the phone company, which has a specific
exemption. That includes network cabling outdoors, network cabling in
commercial or apartment buildings, and home wiring. Homeowners in most
areas can do their own wiring subject to permitting and inspection.

When our house was being expanded, I ran some thicknet ethernet
cable (picked up big spools of it at a hamfest) between several places
where networked computers are sometimes set up. Had I known about
100BaseT and faster coming down the pike, I would have put in Cat-5
cable instead.

Hmm ... also -- does it apply when running ethernet through
fiber optics? No wires there at all! No excuse for needing someone who
is accustomed to wiring voltages around. :-)

Hmm. This could be fun! Well, it looks like a wire....

The law is specific about the definition of "electrical work" and it
looks like fiber is exempted. So is wiring under 24v used for
controlling lawn sprinklers--if you have a network-aware law sprinkler
attached to your LAN does that mean that you're exempt?

Definite iron ricebowl issue.


And how about today's WiFi?

Clearly, does not apply.

The definition is: (2) "Electrical work" means the installation,
erection, maintenance, alteration or repair of any wire, cable, conduit,
busway, raceway, support, insulator, conductor, appliance, apparatus,
fixture or equipment that generates, transforms, transmits or uses
electrical energy for light, heat, power or other purposes, but does not
include low voltage wiring, not exceeding twenty-four volts, used within
a lawn sprinkler system;"

One could argue wifi either way.

Given that radio communications is a major use, and yet isn't mentioned, it
ought to be an easy argument.

And the backup argument would be that obviously the electricians need to obtain
radio transmitter licenses from the FCC before touching any such equipment.


Granted, I have a friend with an old house where it does not
work well at all. The walls are not drywall, but rather real plaster
over metal mesh. But I can't imagine an electrician knowing enough
about that to even diagnose the problem. :-)

Eight or nine years ago, I paid an electrician to install the CAT5e cable
from
basement to my wife's office on the second floor, and I will say that the
electrician and his helper earned their pay on that one. It's an old
house,
with wood lath plaster and strategically-placed bricks as fire-stops. It
took
hours, and they were working all the while.

I then terminated and connected the installed cable. It works just fine.

Installing that wire probably cost what a full WiFi setup would have cost,
but
once installed the wire just works, needs no sysadmin effort, has no
security
drama, and has far greater data capacity than the WiFi of the day. Only
recently have wireless LANs even approached the theoretical datarate of a
CAT5e
wire. In practice, wireless LANs rarely achieve anything like the capacity
of a
cable.

Never. 802.11N has a theoretical maximum throughput of around 450
Mb/sec, CAT5E handles 1000.

I was being nice. And hardwire links achieve a far greater fraction of their
theoretical throughput than wireless links.


If I ran cable today, it would be CAT7, which supports gigabit datarates,

CAT 5e supports gigabit just fine. You don't need CAT7 for, well,
anything. 10 gig Ethernet runs on 6A. 7 is something that the cable
manufacturers want you to think you need, but you don't unless you have
unusual circumstances.

If I recall, the rationale for CAT7 is to carry 40/100 Gigabit ethernet short
distances within datacenters, the theory being that optical was too expensive
and not needed for such short links. Don't know if the market agreed.

In many systems I have worked on, we must use shielded twisted pair such as CAT7
to ensure that the ethernet works despite heavy EMI. The alternative is fiber,
and cable is far easier for links between adjacent cabinets and/or to devices
not supporting fiber.

And CAT7 is just dandy for carrying LVDS signals. This may not be what the wire
companies dream of - volume far too low.



far
exceeding any likely wireless LAN technology. The CAT7 wire is expensive
to be
sure, but installation costs will swamp the wire cost.

But the current 100 megabits per second is more than sufficient.

If it's not, you can get gigabit NICs for 20 bucks.

Yep.


Actually, the best thing to install is plastic conduit, if one has the
opportunity to do it reasonably easily, like if the walls are already open
for some other reason.

I recall lots of ads from various wire manufacturers touting their latest
(fastest, expensive) kinds of cable, saying that companies should buy the
better stuff even if not strictly needed today, as "future-proofing".

And it never was. CAT5 needed to be recertified to 5E to run gigabit
(most CAT5 installations pass 5E, but they weren't tested to it until it
became part of the requirement). CAT6 didn't meet the 10G spec, leading
to 6A.

Yep.

CAT7 is a solution in search of a problem.

As discussed above.


Joe Gwinn

On 7/7/2010 10:30 PM, Joseph Gwinn wrote:
In ,
"DoN. wrote:

On 2010-07-07, Joseph wrote:
In ,
"DoN. wrote:

On 2010-07-07, Joseph wrote:
In ,
"J. wrote:

[ ... ]

[snip]

Granted, I have a friend with an old house where it does not
work well at all. The walls are not drywall, but rather real plaster
over metal mesh. But I can't imagine an electrician knowing enough
about that to even diagnose the problem. :-)

Eight or nine years ago, I paid an electrician to install the CAT5e cable
from
basement to my wife's office on the second floor, and I will say that the
electrician and his helper earned their pay on that one. It's an old
house,
with wood lath plaster and strategically-placed bricks as fire-stops. It
took
hours, and they were working all the while.

Sounds like you are thankful that *you* did not try to do it
yourself. :-)

Absolutely. I've pulled my share of wire over the years, and old houses are the
worst.


I then terminated and connected the installed cable. It works just fine.

Of course.

Installing that wire probably cost what a full WiFi setup would have cost,
but
once installed the wire just works, needs no sysadmin effort, has no
security
drama, and has far greater data capacity than the WiFi of the day. Only
recently have wireless LANs even approached the theoretical datarate of a
CAT5e
wire. In practice, wireless LANs rarely achieve anything like the capacity
of a
cable.

Of course. I've got a wireless bridge running to a friend's
house across the street and down one house -- using circular waveguide
antennas which I machined up. Encryption enabled, good aggressive
firewall on each end, and all logins from one side to the other via ssh.

I'm assuming that you added the waveguide antennas to standard wireless LAN
equipment.


If I ran cable today, it would be CAT7, which supports gigabit datarates,
far
exceeding any likely wireless LAN technology. The CAT7 wire is expensive
to be
sure, but installation costs will swamp the wire cost.

But the current 100 megabits per second is more than sufficient.

Certainly. I'm running my home directories from a server on
another machine, and the speed is quite adequate for most things. If I
want to run find(1) on the home directory, I do that on the server
itself, of course.

Actually, the best thing to install is plastic conduit, if one has the
opportunity to do it reasonably easily, like if the walls are already open
for
some other reason.

Hmm ... specifically plastic? Is there a problem running
twisted-pair ethernet in close proximity to a metal surround? I know
that some of the network wiring we installed at work was shielded
four twisted pair cable. Or is plastic conduit simply less expensive?

CAT5 and CAT5e are not shielded, so running in steel conduit may cause
unexpected losses and reflections. But I have not tried this. CAT7 is
shielded, so running in steel conduit would have no effect. I don't recall if
CAT6 is shielded, but I think it is unshielded. J.Clarke may know more.

Cat 6 & 6A are unshielded. So far all of the successful X-base-T
cabling has been unshielded.

But my reason for plastic was economic, not technical.

Joe Gwinn

On 2010-07-08, Joseph Gwinn wrote:
In article ,
"DoN. Nichols" wrote:

On 2010-07-07, Joseph Gwinn wrote:
In article ,
"DoN. Nichols" wrote:

On 2010-07-07, Joseph Gwinn wrote:
In article ,
"J. Clarke" wrote:

[ ... ]

Sounds like you are thankful that *you* did not try to do it
yourself. :-)

Absolutely. I've pulled my share of wire over the years, and old houses are the
worst.

I can believe that.

[ ... ]

Of course. I've got a wireless bridge running to a friend's
house across the street and down one house -- using circular waveguide
antennas which I machined up. Encryption enabled, good aggressive
firewall on each end, and all logins from one side to the other via ssh.

I'm assuming that you added the waveguide antennas to standard wireless LAN
equipment.

Right. A bit more directionality and gain -- to improve our
signal during nasty weather, and to somewhat reduce people in a position
to snoop on the signals -- even with all the protection in place.

[ ... ]

Actually, the best thing to install is plastic conduit, if one has the
opportunity to do it reasonably easily, like if the walls are already open
for
some other reason.

Hmm ... specifically plastic? Is there a problem running
twisted-pair ethernet in close proximity to a metal surround? I know
that some of the network wiring we installed at work was shielded
four twisted pair cable. Or is plastic conduit simply less expensive?

CAT5 and CAT5e are not shielded, so running in steel conduit may cause
unexpected losses and reflections.

O.K. That I can accept.

But I have not tried this. CAT7 is
shielded, so running in steel conduit would have no effect. I don't recall if
CAT6 is shielded, but I think it is unshielded. J.Clarke may know more.

But my reason for plastic was economic, not technical.

O.K. Out of curiosity -- how long does it take a motivated
mouse to gnaw through the wall of a plastic conduit? (Motivation could
be fingerprints from the sandwich the installer was eating. :-)

Enjoy,
DoN.

--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

In article ,
"DoN. Nichols" wrote:

On 2010-07-08, Joseph Gwinn wrote:
In article ,
"DoN. Nichols" wrote:

On 2010-07-07, Joseph Gwinn wrote:
In article ,
"DoN. Nichols" wrote:

On 2010-07-07, Joseph Gwinn wrote:
In article ,
"J. Clarke" wrote:

[ ... ]

Sounds like you are thankful that *you* did not try to do it
yourself. :-)

Absolutely. I've pulled my share of wire over the years, and old houses
are the
worst.

I can believe that.

[ ... ]

Of course. I've got a wireless bridge running to a friend's
house across the street and down one house -- using circular waveguide
antennas which I machined up. Encryption enabled, good aggressive
firewall on each end, and all logins from one side to the other via ssh.

I'm assuming that you added the waveguide antennas to standard wireless LAN
equipment.

Right. A bit more directionality and gain -- to improve our
signal during nasty weather, and to somewhat reduce people in a position
to snoop on the signals -- even with all the protection in place.

[ ... ]

Actually, the best thing to install is plastic conduit, if one has the
opportunity to do it reasonably easily, like if the walls are already
open for some other reason.

Hmm ... specifically plastic? Is there a problem running
twisted-pair ethernet in close proximity to a metal surround? I know
that some of the network wiring we installed at work was shielded
four twisted pair cable. Or is plastic conduit simply less expensive?

CAT5 and CAT5e are not shielded, so running in steel conduit may cause
unexpected losses and reflections.

O.K. That I can accept.

But I have not tried this. CAT7 is
shielded, so running in steel conduit would have no effect. I don't recall
if CAT6 is shielded, but I think it is unshielded. J.Clarke may know more.

Turns out that CAT6 is unshielded, so CAT7 it would be.


But my reason for plastic was economic, not technical.

O.K. Out of curiosity -- how long does it take a motivated
mouse to gnaw through the wall of a plastic conduit? (Motivation could
be fingerprints from the sandwich the installer was eating. :-)

Hmm. I've never heard of this happening, and lots of plastic conduit and pipe
is used. Mice are lazy too. They would lick the conduit clean before trying to
gnaw through.

I assume that part of the conduit development process was to pick a plastic
formulation that does not taste good to rodents. It's certainly a big deal for
wire and conduit intended for direct burial.

Joe Gwinn