strange ethernet electric problem

Hello, we have to connect two computers via ethernet in an industrial enviroment. we are using a crossover cable to connect both computers.

If i connect my laptop running with bateries to one of this computers the ethernet connection works fine but when i plug my laptop to the power line the ethernet connection brokes. This power line is common to all devices and we have tested with several cables.

This problem does not happen always. Some days all seems to work fine but other days the ethernet connection does not work a all.

Any ideas?

Reply to
peancor
Loading thread data ...

Are they homemade crossover cables? It is possible that a mispaired cable could cause that problem. It must be such that pins (1,2) and pins (3,6) are each the two wires of a twisted pair.

-- glen

Reply to
glen herrmannsfeldt

glen herrmannsfeldt wrote in news:-ICdnboFdI0RP-bbnZ2dnUVZ snipped-for-privacy@comcast.com:

The cables work fine in my development enviroment. It is only in the industrial enviroment where the failure is detected and i suppose that it is related to the power line since my laptop connects fine when it is running on batteries.

Reply to
ghost in the machine

You haven't answered the question.

Reply to
Al Dykes

ghost in the machine wrote in part:

Two issues: industrial environments are electomagnetically noisier and any cable imperfections (like split pair) more likely to fail. In some environments STP is required.

Second, have an electrician check out equipment grounding. There may be a ground potential difference or ground noise.

-- Robert

Reply to
Robert Redelmeier

Sorry, the cables are cat5e but the ethernet connectors are attached to the cable manually (i dont know the name of the machine used to do that).

snipped-for-privacy@panix.com (Al Dykes) wrote in news:f5gigd$lht$ snipped-for-privacy@panix5.panix.com:

Reply to
peancor

peancor wrote in part:

Stranded or solid core conductor?

What connectors? Plugs [male] or jacks [female]? There can definitely be incompatibilities if connectors are used on the wrong cable type.

-- Robert

Reply to
Robert Redelmeier

Robert Redelmeier wrote in news:YjRei.21339$ snipped-for-privacy@newssvr23.news.prodigy.net:

Thank you all!

I thought about STP but i did a test with a very short cable with similar results. All equipment is connected to a 'good ground' and the same power lines are used. Previously the equipment was connected to diferent power lines and we never had such ethernet problems using the same components. We had to change the power installation due to grounding problems were affecting to other parts of the system.

What really disconcerts me is the power issue. All the devices are powered from the same point. When i plug my laptop to the same point (using a power source with no ground connection) the ethernet fails.

Also the problems does not happen always. Some days all is working perfect but another days the ethernet does not work.

Do you know of any industrial machine what could cause this effects through the power lines?

Reply to
peancor

Robert Redelmeier wrote in news:Z4Tei.21378$ snipped-for-privacy@newssvr23.news.prodigy.net:

it is stranded and we need a distance of about 25m.

the connectors are plugs on both sides

Reply to
peancor

glen herrmannsfeldt wrote in news:xr2dnbsv7KLGYebbnZ2dnUVZ snipped-for-privacy@comcast.com:

Could you please explain me briefly the concept of mispaired cable? I did a quick googling with no relevant results.

Thank you in advance.

We may have RF (400A Magnetic Agitation) relatively near the system although the cable is ground canalized and isolated from other conductors.

At first, we thought about the magnetic agitation but it was not enabled when the system failed. We did a test with the magnetic agitation enabled and the system worked fine in that occassion.

Reply to
peancor

A mispaired cable can stop or start working for almost any reason.

Ethernet is transformer coupled, as are laptop power supplies, so there is no obvious direct cause for failure. Unless you are in an extreme RF environment, such as near a radio transmitter, there aren't many reasons for it not to work.

-- glen

Reply to
glen herrmannsfeldt

peancor wrote in part:

OK. A little long for stranded, but well within total length spec (100m).

OK. Those need to be the correct plugs for stranded, usually single tooth. Look carefully through the plastic (magnifying glass if needed) and tell us the order of the conductors (colors).

-- Robert

Reply to
Robert Redelmeier

Robert Redelmeier wrote in news:tQTei.17490$y snipped-for-privacy@newssvr27.news.prodigy.net:

i do not have access right now to the cables but i assume that they are correctly wired (if you are referring to that) since they work perfectly in my development enviroment.

Reply to
peancor

There is some explanation in:

formatting link
It could be better, though. The two wires for transmit and the two for receive much each be twisted together. If they aren't, then the induced currents and voltages don't cancel out in the end.

I don't know what Magnetic Agitation is. (and google didn't help.)

-- glen

Reply to
glen herrmannsfeldt

glen herrmannsfeldt wrote in news:Pv6dnU-WisuMruHbnZ2dnUVZ snipped-for-privacy@comcast.com:

Thank you, yes i had considered that but we bought the cables as cat5e, so i guess that they should be well paired, and we have tested with several cables.

Basically we had a metallic fluid being shaked by means of a big magnetic field but as i told you the system has been working correctly when the field was active as well as failing when the field was inactive.

Reply to
peancor

Mispaired cables can almost work, depending on the exact pairing and length. For 10baseT they tend to work up to about 10m or so. It might depend on how the cable was coiled, what other wires or metal objects were nearby, etc.

The magnetic agitator sounds like it should be low frequency magnetic fields, which shouldn't effect ethernet so much.

My next suggestion is to try powering the laptop from an outlet close to the other end of the ethernet cable. That might need a long cord.

-- glen

Reply to
glen herrmannsfeldt

glen herrmannsfeldt wrote in news:o_-dnSvN-L8x1-HbnZ2dnUVZ snipped-for-privacy@comcast.com:

Yes, it is low frequency. The factory also uses variable-frequency drives (VFD) to control electric engines.

We tried powering the laptop (and the system) to an outlet unrelated to the first one and all worked but i do not know what conclusion extract from that. We have been running the system during several weeks with other electric setup (but the same ethernet setup) and the ethernet never give us a problem.

But the purpose of the new electric installation was preciselly to provide common power lines to the system and a good ground to avoid another problems that we had in the system so we can not return to the previous electric setup easily.

I can not see a clear relation between the power lines and the ethernet connection. The ethernet cards are correctly detected and we tried with several ethernet cards with similar results so i suppose that we have a structural problem not tied to a particular device.

Reply to
peancor

Buy a factory-made 100ft patch cord and try it.

Reply to
Al Dykes

Ethernet NICs isolate the computers at both ends from anything induced onto the cable by the RFI/EMI environment to something like 1000v. This would also isolate any voltage differences caused by grounding problems.

RFI/EMI and ground-induced interference with properly-wired Ethernet equipment is effectively non-existent in the real world.

Reply to
Al Dykes

snipped-for-privacy@panix.com (Al Dykes) wrote in news:f5j5jr$g5i$ snipped-for-privacy@panix5.panix.com:

Yes, i am going to buy a factory made cable but the fact that all cables work fine in my development enviroment and the electrical issue disconcerts me.

Reply to
peancor

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.