Current rating of an RJ45 connection

[Joerg]

I beg to differ. Although most of the stuff I work with is rated from
0 to 70C, reality has a way of exceeding the specifications. For
example, realistic automotive temperature spec is -40C to about 100C
(or more). Engine compartments get very hot.

And some automotive engineers should have their head examined. On one
car a module that controlled some stuff in the engine was literally
mounted on top of the engine. Sometiems on hot days the van would refuse
to start ...

The starter on my wife's Corolla just died. Where did they put it?
Between intake manifold and engine! Unreachable, and gets hot there.

<http://en.wikipedia.org/wiki/Operating_temperature>
Recently, we attached a disposable temperature data logger into an
aluminum radio case and let it run for most of salmon season. It went
well over 100C for extended periods when the sun was shining directly
on the case. The vessel owner also complained that his iPod blew up
when left sitting on the top of the same radio. He blamed RF from the
radio, but it was more likely fried by excessive heating.
Incidentally, when I did an autopsy on the radio, most of the screws
had rattled loose. I should have also attached a vibration and
accelerometer data logger.

I was taught very early on that eveything on a boat must be nicely
strapped in and will corrode away in no time. And that stainless doesn't
really mean stainless.

Clever design. Wrap the handle with duct tape. Most such things are
designed for a specific useful life. I suspect (but can't prove) that
research has been performed into reducing the life of plastics under
the guise of having the plastic break down in the landfill. While it
seems a worthy area of research, I suspect that such plastics would
also deteriorate before they reach the landfill. A modest example is
the simulated rubber coating used to simulate a real rubber handle on
some hand held tools (e.g. Sears IR thermometer) and your shovel. It
is designed to turn to sticky mush after about 3-4 years, which is
just about right to insure that none are returned under warranty. I
found that ordinary rubbing alcohol does a tolerable job of removing
the sticky goo without affecting the underlying plastic (usually ABS).

This one is soft goo stuff all through :-(

I've seen similar controlled failure designs in the selection of
electrolytic capacitor voltage ratings and power transistor thermal
cycle calculations. It's fairly easy to predict the onset of such
component failures.


What design review? I don't want to go into excessive detail, but
basically the design was purchased from a consultant in India and
rushed into manufacture without adequate testing. ...

Ah, a design from Outsourcia

... First to market and
cost were the overriding concerns. There were plenty of other things
wrong with it. I was sweating blood because the independent
certification lab was complaining that they couldn't make it pass Part
15 Class A without major modifications. My job was to make these
major modifications, but not change anything, since the product was
being manufactured and stockpiled pending approval. Other engineers
working on the device used this as an opportunity to fix a few things,
which required my approval since it affected the type certification.
That made me effectively in charge of coordinating everything and
guaranteed that I would be blamed if anything went wrong. The BFC fix
was added under my watch. Things went amazingly smoothly. The
certification lab was able to generate the required FCC paperwork.
Changes were made that didn't involve a PCB revision. There was some
grumbling over the necessary rework. No real problems until the RJ45
arcing problem appeared.


That's one mistake I haven't made since. It's an open invitation for
abuse as I learned the hard way. However, I created my own problem. I
could have said "it's not my job" and let someone else handle it. I
was thinking that I could play hero and impress management
sufficiently to get more work from the company. That didn't happen.

If someone insists on fixed but I either insist on cast-in-concrete
specs plus change-order procedure, or decline. Since neither is desired
by clients it's all by the hour for me.

 

[[email protected]]

And some automotive engineers should have their head examined. On one
car a module that controlled some stuff in the engine was literally
mounted on top of the engine. Sometiems on hot days the van would refuse
to start ...

The starter on my wife's Corolla just died. Where did they put it?
Between intake manifold and engine! Unreachable, and gets hot there.

I've fixed the starter on few corollas and getting that thing
unbolted
down behind the engine can drive you nuts

the fix is simple, there's even a cheap kit for it, the thing that
break
is the "switch" inside, basically two pieces of copper angle that
gets
shorted via a copper disc when the started gear is engaged
one of them gets eaten away, buy the kit or bend a piece of copper
and
it will run for another 10 years

-Lasse

 

[Joerg]

I've fixed the starter on few corollas and getting that thing
unbolted
down behind the engine can drive you nuts

With my arm size I don't see the slightest chance to reach in there
unless I take the whole intake manifold off instead of just the air
filter. Looks like I have to pay for a tow and have it fixed at a shop.
Because it's an automatic there is no way to get the car started. All
there is is that telltale loud click.

the fix is simple, there's even a cheap kit for it, the thing that
break
is the "switch" inside, basically two pieces of copper angle that
gets
shorted via a copper disc when the started gear is engaged
one of them gets eaten away, buy the kit or bend a piece of copper
and
it will run for another 10 years

Yup, seen it in a Youtube video. Most likely one of the copper pieces
has sailed off because no matter how often I try it will not establish
contact and crank.

Hurumph ... grumble ... grumble ...

 

[[email protected]]

With my arm size I don't see the slightest chance to reach in there
unless I take the whole intake manifold off instead of just the air
filter. Looks like I have to pay for a tow and have it fixed at a shop.
Because it's an automatic there is no way to get the car started. All
there is is that telltale loud click.

yep that's the symptoms, some times a great big wack with a piece of
wood
will make it work a few times

it's only held on with two 8mm bolts from the end, afair the problem
is that
you can't see what you are doing

Yup, seen it in a Youtube video. Most likely one of the copper pieces
has sailed off because no matter how often I try it will not establish
contact and crank.

one of them gets eroded by the sparking I guess, eventually it is so
thin
the disc can't reach it

-Lasse

 

[Nico Coesel]

Not sure if the difference is that stark in real life. I own a white car
and a couple of weeks ago during really hot summer days I had a black
Volkswagen rental. I didn't feel that much of a difference when coming
back to the parked car after severak hours.

Interestingly the color we see doesn't matter. Paint can have totally
different properties at different wavelengths. There is black paint
which is 'white' if you look at it with an infrared camera.

 

[Nico Coesel]

All the Gold utilized has changed all that.

Cat 5 and 6 say for the last 15 years, has been just fine as far as
contact issues go.

You obviously never had to run an IT department. Connector problems
cause 99% of the failures. And a network often had quite a few
connectors between the switch and the device. In most cases one cable
from the switch to the patch panel and then one cable from the outlet
to the PC. That is 4 RJ45 connectors between the switch and a PC!

 

[Joerg]

yep that's the symptoms, some times a great big wack with a piece of
wood
will make it work a few times

it's only held on with two 8mm bolts from the end, afair the problem
is that
you can't see what you are doing

They looked more like 6mm to me. Problem is that at least one needs to
be loosened from underneath the car and the wire from the start relay
also connects from below. There is all sorts of structurally important
stuff between that and the bottom, no way I can reach that. What a pain.

one of them gets eroded by the sparking I guess, eventually it is so
thin
the disc can't reach it

This was all so easy on my old Citroen 2CV. You could reach just about
anything directly.

 

[Joerg]

You obviously never had to run an IT department. Connector problems
cause 99% of the failures. ...

Not really. People do. At least that's what IT pros told me.

Best one was in the late 80's, the days of coax Ethernet. "But I needed
a 50ohm terminator real quick!"

[...]

 

[[email protected]]

Interestingly the color we see doesn't matter. Paint can have totally
different properties at different wavelengths. There is black paint
which is 'white' if you look at it with an infrared camera.

It does matter, however it's not all that matters. A black (optically) car
will absorb more energy in the visible range than a white car. As you point
out, what it does in the IR range can't be deduced by the "color" in the
visible range.

 

[[email protected]]

They looked more like 6mm to me. Problem is that at least one needs to
be loosened from underneath the car and the wire from the start relay
also connects from below. There is all sorts of structurally important
stuff between that and the bottom, no way I can reach that. What a pain.

I did all from the top standing next to the car, problem was turning
bolt one turn at a time with a spanner feeling the way, can't remember
if a socket wouldn't fit or I just didn't have one

once you have two bolts off you can get out to where it is easy to get
the rest off

-Lasse

 

[Joerg]

I did all from the top standing next to the car, problem was turning
bolt one turn at a time with a spanner feeling the way, can't remember
if a socket wouldn't fit or I just didn't have one

once you have two bolts off you can get out to where it is easy to get
the rest off

I just called someone who recommended a local shop so I'll do the
decadent thing and just have it done. The American Corollas have all
sorts of emissions control stuff hovering over the engine. It looks like
an aircraft engine in there. Without taking a lot of that off I couldn't
even reach in to feel a socket wrench into position.

 

[Joerg]

And some automotive engineers should have their head examined. On one
car a module that controlled some stuff in the engine was literally
mounted on top of the engine. Sometiems on hot days the van would refuse
to start ...

It probably works ok when the vehicle is moving and there's plenty of
air flow. The thermostatically controlled fan on the radiator will
help with the air flow when it's not moving. There was a program to
reduce the amount of copper (wire) used in automobiles a few years
ago. I suspect mounting the black box on the engine is one of those
copper saving measures[1]. Part of this is my nightmare come true:
<http://en.wikipedia.org/wiki/42-volt_electrical_system>

I think the 42V stuff has fizzeld, as expected. But mounting electronics
on top of a motor is real stupidity, one just does not do that.

Ever get the feeling that some vehicles were never designed to be
repaired by mere mortals? There are only so many places on an engine
where one can locate a starter motor along the circumference of the
flywheel.

In this case there would have been space towards the radiator, but ...

I was taught very early on that eveything on a boat must be nicely
strapped in and will corrode away in no time. And that stainless doesn't
really mean stainless.

Stainless on stainless has the amazing ability to be self corrosive.
<http://yarchive.net/electr/galvanic_corrosion.html>
In fact, stainless steel can be both anodic and cathodic
to ITSELF. This means that a single piece of stainless
can act as a galvanic battery and promote its own corrosion.
Corrosion in a marine environment is tricky but not impossible. Marine
radios tend to be fairly well protected. However, in the distant
past, I designed parts of the AN/SRD-21 and 22 direction finders for
the US Coast Guard. Aluminum box and stainless hardware initially did
a great job of corroding everything together. The big surprise was
that nobody ever told us that the radios were going to mounted on the
exposed bridge of a 42 footer. Standard procedure is to hose down
everything with a "clean" water hose. A 60 PSI water jet can do an
amazing amount of damage to the front of a radio, such as cave in the
displays and shred the loudspeaker.

Ouch!

[...]

Ah, a design from Outsourcia

Nope, Elbonia:
<http://www70.homepage.villanova.edu/matthew.liberatore/dilbert/elbonia.htm>



One of the other engineers wore the hat and beard to a meeting. The
design was quite good for the amount of time allowed. However, the
company also had the PCB and packaging done in India. While the
design company was sufficiently experienced in circuit design to do a
decent job, the packaging and PCB layout were a mess, and may have
been done by a sub-contractor. Both eventually had to be redone.
That's also the source of the RJ45 arcing problem.

Incidentally, the original package was an amazing example of a box
that could only be built by left-handed assemblers. I didn't believe
it until I tried it myself.

Sounds like the Toyota here, except you also need skinny arms.

[...]

 

[SoothSayer]

Yeah, sorta. I don't run an IT department. I do run a consulting biz
that in the past has wired various buildings, offices and homes for
ethernet. I've also helped setup server farms for an ISP. Lots of
connectors involved. I've seen failures, but they were never due to
contact failure. It was almost always failure of the insulation
displacement connection on the jack, crimping failure on the plug,
wrong plug type, and shredded cables. Most customers insist on having
the wiring tested with a cable certifier to detect split pairs and
other wiring screwups that don't show up on continuity tests. All my
crimpers are ratcheting type, which is required for a decent
connection. Lots of other precautions, such as pull tests. After
installation, the wiring is again tested using diagnostics built into
the managed switch.

When I inherit a new system, I run the certifier though the wiring and
usually find problems. Usually, it's bad patch cables with miserable
crimps to the plugs. Lately, I'm finding really awful RJ45 plugs that
really can't be properly crimped. I sometimes find stranded wire RJ45
plugs used with stranded CAT5. That won't work (although the reverse
will work).

I rarely see any "failures" that could be directly attributable to the
connector unless the connector was mechanically damaged. If I had to
assign the blame, it would be with whomever built the patch cable or
installed the wall jacks.

Our gateways have optical links, and 100% redundancy.

IF there were EVER a connection problem on ANY port(for the Enet
discussion), it would switch it out, and alert the operators.

Hp series 8 blades for the C7000 chassis series are cool. The old
blades had vertical locking bar and they moved these down onto the end.
It is amazing how much they pack inside these blades. Dual Zeon, SAS,
Optical interfaces, and Ethernet.

After the third party vendor comes in and wires up our racks (way
cheaper than paying our own people to do it, oddly), we are pretty
confident as they tag each line and run diagnostics on each Ethernet
line. Then our in house QAs verify every element of the build on each
rack.

Our RF guys test out the link side cables and get that set up. Then we
take the whole thing apart and crate the racks and ship them to the
gateway site. There, another person re-assembles them and the same wire
crew (we fly them all over the world) re-runs all the wiring to the rack
destinations. Every cable and every link of any kind gets tested here.
At the site everything works again upon re-assembly. That wiring company
makes more money from us than any other customer they ever had, and their
workers get to see the world like an active duty sailor.

We calibrate our crimpers, and so do the cable wire crews. Hell, their
gear is better than some of our gear as it relates to wires. Bad crimps
are the bane of the industry.

However, a properly constructed system will behave the way I stated.

Bad crimps make certain folks draw improper conclusions, but they are,
in fact, the reason most of the failures they observed occurred. It is
hardly ever a connector contact face, or connector mating contact wire
tangency face.

Back in my early days, I worked on upright video games. We had to
charge the customer (video game operator) a minimum of a 1/2 hour for the
service call. Most of them only took a few minutes. So to keep a
customer from becoming irate that you only spent 3 minutes fixing
something, but charged a half hour, you had to sit behind the machine and
act as if you were fixing something.

Back then card edge fingers and connectors were far more prone to
surface oxidation issues (remember re-seating hard drive connectors on
MFM?). All we had to do to fix the game usually is power down (usually
is already) and remove and re-seat all the main board's edge connectors.
Job done, 95% of the time.

I used to have problem charging a customer for a TV repair for this
reason, back when I did that. That was when I learned that they are
paying for your expertise, not your time. Customers, however, still have
problems accepting that reason, to this day.

Like I said, there are trillions of connections each day that operate
just fine, even in extreme environs.

The failure rate is literally more than an order of magnitude less than
it was 20 years ago.

Connector technology has taken big steps forward too. Smaller and
smaller, yet quite integral. Gold is the reason.

Backplanes and such didn't get added to my numbers. There are 1000s of
trillions of connections of high integrity working every day.

How many do you think were utilized in getting your post to the first
server it hit? Tens of thousands likely.

 

[Nico Coesel]

Yeah, sorta. I don't run an IT department. I do run a consulting biz
that in the past has wired various buildings, offices and homes for
ethernet. I've also helped setup server farms for an ISP. Lots of

I rarely see any "failures" that could be directly attributable to the
connector unless the connector was mechanically damaged. If I had to
assign the blame, it would be with whomever built the patch cable or
installed the wall jacks.

I agree a lot can go wrong when the wiring is installed. I've seen my
share of badly crimped connectors, poorly installed wall-outlet and
bad punch-down work. But after installation dirt and moisture are the
biggest enemy.

I recently designed a board for which the customer and I decided to
use a spring loaded terminal block instead of RJ45 because it is much
more resistant against poor installation practices and moisture.

 

[Joerg]

Take the engine out & replace it with a big rubber band and it'll
still be faster than the 2CV.

Hey, my 2CV could blaze along the autobahns at 55mph. 60-65mph only with
a stiff tailwind. But ... 50mpg, and that's with an engine designed in
the late 30's.

 

[josephkk]

Folks,

Have to send 12V/500mA across a connection. So far there is only an
RJ45. Looking at datasheets I saw current ratings of 1.2A and 1.5A which
sounds like a lot.

Is this kosher? I know PoE sends up to 600mA across but I've always felt
uneasy about it when they went up from the (older?) 350mA.

Just in case you are really interested, the connector is IEC 60603-7-x
where x is a speed grade. Straight from my copy of ANSI/TIA-568 C
revision part 2.

The IEC spec is 220 US$ if i read the website correctly:

http://webstore.iec.ch/Webstore/webstore.nsf/ArtNum_PK/41668!opendocument&preview=1

 

[legg]

Interestingly the color we see doesn't matter. Paint can have totally
different properties at different wavelengths. There is black paint
which is 'white' if you look at it with an infrared camera.

Different optically clear components (as in commercial RJ45) also have
different absorption characteristics. But this besides the point, as
is the automobile comparison.

"The unit will sit in full sun occasionally and can become as hot as
the inside of a car under that condition."

For a start, very few people have any idea what the varying
temperatures are in the different areas of their car. They are aware
of how hot their seats are, maybe the passenger compartment air, and
the dashboard, where their favourite CD just got toasted. This usually
involves the IR transmissivity of auto glass and the absorptive
capacity of the upholstery or temperature indexes of toasted articles.

Electronic packages that are designed to sit in full insolation need
to be packaged appropriately, just as connectors (original topic) need
to be selected appropriately for their application.

RL

 

[Chieftain of the Carpet Crawlers]

One of the most retarded comments about IR and IR imagery and "black
paint" I have ever seen.

 

[Chieftain of the Carpet Crawlers]

"The unit will sit in full sun occasionally and can become as hot as
the inside of a car under that condition."

Put a braided sheath over it. Sheesh.

 

[Warren]

Sure, but there is also re-radiation and reflection. Except maybe for a
Hell Angels Harley or a low-rider rap mobile, vehicles are normally not
painted a dull "primer black". It's more metallic-anthracite plus a
shiny clear coat on top.

My father painted his stock car's radiator silver one time. From that
point on, the engine overheated badly and he could not understand why (I
was not old enough at the time to be of any help).

After several weeks of trying to work out the problem he eventually gave
up and discarded that radiator and put a different "black" one in. The
problem then completely vanished. It wasn't until years later, when I
took thermodynamics in high school, that I realized why a black radiator
works better for dissipating heat.

But what I don't understand is why air conditioning heat exchangers are
all aluminum in color. Why do you not see black A/C heat exchanger
coils? Even on your car, your radiator will be black, but the A/C heat
exchanger in front of it (which also must dissipate heat) is left an
aluminum (no paint) color.

Would the A/C heat exchanger work better if it were also painted a thin
layer of black paint? I've never had the guts to try it on my home's
central air coils.

Warren.