Grounded RJ-45 connections 2

[johningram]

Where can i get the rj-45 plugs (the male piece) that have the ground connection for a good price. Im looking at only buying a few like 10 or less. Although for a good price i would buy more.

Also i have read on the net it may be better to only ground one side of the wire and not to both switches. Only one switch. Is this true and why? I would think most people would have just grounded both sides.

 

[daronwilson]

If you are not going to bond BOTH ends, don't bother bonding either end.

I strongly disagree. Grounding one end allows for stray RF or noise on the wire to find a way to ground and be 'drained' from the shield. Grounding at both ends does the same thing, however, it does add the potential for ground loop problems. Also if there some other failure of the neutral/ground system in the building or between buildings, your ground can attempt to handle a major amount of current, which it will do for a very short time until meltdown. We do literally miles of 24volt 4-20ma instrumentation wiring on single pair shielded wire grounded at the cabinet and floating at the instrument.

Not grounding at either end makes a great antenna. Just for fun, lets say your non-grounded shield is reasonably long and is resonant on some AM broadcast band, pretty easy to do really. That 10,000 watt AM station across down still puts a pretty good amount of RF in your building and actually transfers to your shield. Now instead of that stray RF going to ground at one end (or either end) it floats ont he jacked and couples to the wires inside causing problems.

So, install your wire, get out your multimeter, check for differences in potential, if none bond at each switch. IF you experience problems with the link or equipment, you might consider lifting the ground on one end to see if you have some ground loop issues.

Good Luck

It is only my opinion, based on my experience and education...I am always willing to learn, educate me!
Daron J. Wilson, RCDD
daron.wilson@lhmorris.com

 

[Servamatic]

As usual you missed my point...apparently I don't explain things very well....
As for instrumentation, I have done my share and my experience completely disagrees with yours, and yes it was your typical 4-20. Since it was an oil refinery which is not a place you want stray voltages, I tend to think they did some serious consideration and determined the best way to reduce any possibility of catastrophe.
Oh well, I have said much more than I need too on this subject.


Richard S. Anderson, RCDD

http://www.servamatic.net

 

[daronwilson]

Here is some straight forward info for you Richard:

Cable terminations

Terminations are an important part of instrument wiring. Proper terminations result in reliable connections and help to eliminate problems like ground loops and electromagnetic interference. Instrument terminations are usually made to screw terminals or compression terminals, and either one can be made reliably. Two-conductor shielded twisted pair cable is terminated in two different ways, with the drain wire connected to a terminal or with the drain wire cut off and insulated.

Fig. 1 shows a typical two-conductor shielded cable prepared for termination to screw terminals with the drain wire cut off. This is usually done at the field end of an instrument cable where no shield grounding is desired. Note that insulating tape or heat-shrink tubing is used to protect the cable from contamination and to prevent accidental grounding of the shield or drain wire. An accidental ground at this point would almost certainly create an undesirable ground loop.
(Figure 1 ILLUSTRATION OMITTED)
Fig. 2 shows a typical two-conductor shielded cable prepared for termination to screw terminals where the drain wire is to be connected. Note that the drain wire, which is an uninsulated conductor, is sleeved with insulating tubing to prevent accidental grounding. The crimp-on lug is valuable in this instance to retain the tubing. Insulating tape or heat-shrink tubing is again used to protect the cable from contamination and to prevent accidental grounding, since any accidental connection between the drain wire and a chassis, frame, or enclosure would almost certainly create a ground loop.
(Figure 2 ILLUSTRATION OMITTED)


Pulled from Electrical Construction and Maintenance magazine at this link:

http://www.ecmweb.com/ar/electric_avoid_instrumentation_problems/

It is only my opinion, based on my experience and education...I am always willing to learn, educate me!
Daron J. Wilson, RCDD
daron.wilson@lhmorris.com

 

[AdmanOK]

You know... sometimes this forum reminds me the Monty Python sketch called "Argument Clinic" (

http://www.infidels.org/news/atheism/sn-python.html)

where the guy goes in and pays to have an arguement with John Cleese's character... the only difference being that you get to do it for free here, and you have an audience!

 

[fiberman40]

Forget the copper cable and run a piece of duplex multimode fiber with 2 fiber to copper trancievers and then the grounding issue goes away.

I WILL FISH FOR FOOD

 

[josepablomir]

Servamatic: I don't want to be rude, but you don't know what you are talking about!!!!!!

In a STP LAN you are NOT going to gain anything by grounding both ends of the shielding.

If in some special environment you have found that doing it solve a communication problem then that environment should require another solution instead of the STP you have used.

If you think different maybe you can explain me why.
Maybe Parseval, Fourier, Kirchoff or Maxwell have some secret theorem about this subject and you may point me in the proper direction.

I'm sorry, I'm starting to sound like Mr. Daron Wilson. It won't happen again, I promise!

Jose.-



_________________
Pablo Mir

http://www.pablomir.com

pm@pablomir.com
NJ, USA

 

[daronwilson]

I can't figure out if i was just insulted or not, if i was, same to ya

Oh well, just to clear it up, I don't have a good reference on what you should do on ScTp (I don't think we were ever considering STP). I think you will likely have ground loops if you ground both ends. But...my point of contention was with Richard, and illustrating that not all shielded (screened) wire should be grounded at both ends, such as instrumentation wiring.

Pablo...tell Kirchoff hello for me

It is only my opinion, based on my experience and education...I am always willing to learn, educate me!
Daron J. Wilson, RCDD
daron.wilson@lhmorris.com

 

[Servamatic]

Pablo
Well you are damn rude.... and frankly it is you that doesn’t know what you are talking about, to make such a blanket statement proves your ignorance.

I have worked on the manufacturing side of this and with the standards committees, so I have talked with a lot of people that know way more than you or I do.
If grounding both ends was bad, you would not find it in the manufacturers instructions...do you think they want to have product installed that causes problems. Get a clue.

If you don’t properly ground and bond then yes you are very likely to have ground loop issues, so if that is your practice, which I assume it is since you have so many problems...don’t bond both ends. But don’t act so high and mighty and tell everyone that it should never be done.

Just out of curiosity, on STP (not ScTP) which end would YOU bond? or would you bond at all?

Richard S. Anderson, RCDD

http://www.servamatic.net

 

[daronwilson]

What might really help out about now is a quote from the 'manufacturer' that specifies how to hook up the equipment. If the 'manufacturer' has some specific requirements for this product, why not just copy and paste that in here and we can all benefit?

It is only my opinion, based on my experience and education...I am always willing to learn, educate me!
Daron J. Wilson, RCDD
daron.wilson@lhmorris.com

 

[Servamatic]

Have you ever read the instruction sheets that come with the jacks and panels?
In a BICSI training class all methods are taken from manufacturer instruction sheets, and I gaurantee you if you don't bond both ends in a BICSI test, you will not pass.

But if you insist:

http://www.siemon.com/int/download/installation_instructions/us/Screened_MAX_6_Modules.pdf

http://www.siemon.com/int/download/installation_instructions/us/HD5_Screened_Patch_Panels.pdf

As you will see, they clearly tell you how to terminate the foil/braid and drain wire on both panels and connectors, and how to bond the patch panel to ground.

Richard S. Anderson, RCDD

http://www.servamatic.net

 

[josepablomir]

Daronwilson: It was not an insult, just a joke. Kirchoff didn't answer my last email, but Thevenin says "Hello".
Servamatic: Sorry for being rude. My fault!

I think we have to make some ponting.
For example, ground loops occur when one conductor (shielding, screen, or whatever) connects two diferent ground points. For example your wiring closet ground point with the accounting PC ground point. This is the ONLY possible way to create a ground loop in the data wire. If you don't ground one extreme of the shield then you CAN'T have a ground loop (created by the data wire). Exception to this is when you design a star-wired isolated ground system, or planning to comply with GR-1089-CORE. But in those cases the ground loops are avoided and/or "managed" by special configurations and/or equipment.
In low cost simple Shielded and/or Screened Twisted Pair wiring schemas I bond to ground only in the wiring closet, so the system won't have any ground loop because of the data wiring.

What about EMI/RFI in one-end-only ground-bonding?
No problem! RF systems smaller than 500 mts. are considered local systems. In local systems the external electromagnetic interference is, ussually, common-mode noise, so one-end ground-bonding of the shield should be enough! Shielded (or Screened) Twisted Pair wiring is intended to solve this common-mode electrical noise (not the diferential-mode noise wich is solved (!?) by the twist density and electronic filtering inside the active devices).





_________________
Pablo Mir

http://www.pablomir.com

pm@pablomir.com
NJ, USA

 

[daronwilson]

Richard thanks for the links to the manufacturers instructions, those were great, and would be applicable if I were running ScTP wiring system to the workstations. However, that was never the question. We were talking about connection two switches together, and I see no manufacturers recommendations regarding that in the documentation.

Depending on the building and configuration, the two switches could be at very different ground potentials. Different floors, different services, etc. I was kind of hoping to see a switch manufacturer's recommendation regarding interconnecting switches with an ScTP backbone. Not that it will really make much difference to me, if I have a ground loop issue I'm dropping one end of the shield and going to have it running just fine.

FWIW, I passed my BICSI test the first time, even doing grounding my way.

Oh that Thevenin....what a character in high school!




It is only my opinion, based on my experience and education...I am always willing to learn, educate me!
Daron J. Wilson, RCDD
daron.wilson@lhmorris.com

 

[Servamatic]

I guess we will have to agree to disagree, you do it your way, I will do it the way it should be done.
One more time for the cheap seats, proper bonding and grounding followed by testing before connecting, will eliminate ground loop issues.

I wont get into your BICSI test issue Daron, lets just say I have serious doubts.

Richard S. Anderson, RCDD

http://www.servamatic.net

 

[daronwilson]

SO....I guess you CAN'T provide manufacturers instruction sheets showing us all how to hook up switches with ScTP cabling. I'll try to look when I get a chance, hopefully the folks that make the switch have some guidelines.

Richard, you bragged about your 3 months of instrumentation cabling where you grounded both ends and told us that was the correct way. I provided you documentation from a nice trade journal showing that you were doing it wrong. I suppose I can get our instrument engineer to provide me some manufacturer data sheets for specific equipment if you still don't believe it.

So, go ahead and do it 'the way it should be done' even though the professional electrical engineers see it differently. That's mighty smug.

Unfortunately it is damn hard to 'eliminate ground loop issues'. If you have done anything with broadcast facilities or musical stage setups, you would know they crop up all the time, even on seemingly well grounded setups.

Best of luck

It is only my opinion, based on my experience and education...I am always willing to learn, educate me!
Daron J. Wilson, RCDD
daron.wilson@lhmorris.com

 

[wires]

I was always taught that if you were going to connect two grounds together it had best be done with beefy conductors. The entire idea allowing even the possibility of "drain wires" to carry current from one area to another gives me the willies.

You can get all the advantages of shielding without engaging in such a dangerous practice by using a star grounding configuration. Personally I have seen little need for shielded/screened cable except for coax, audio and instrumentation.

The only upside I can see to connecting the dinky shield wire at both ends is that in a failure situation it would likley function as a non-resetable circuit breaker (i.e. as a fuse).

 

[daronwilson]

One more time for the cheap seats, proper bonding and grounding followed by testing before connecting, will eliminate ground loop issues

*ouch* the cheap seats? Are you center stage and all us peons are the cheap seats? Oh well, here is an interesting link Richard, apparently not everyone knows your method eliminates ground loop issues.

http://www.cbmusa.com/products/patch.php3

I'll keep looking for more info for us peons...


It is only my opinion, based on my experience and education...I am always willing to learn, educate me!
Daron J. Wilson, RCDD
daron.wilson@lhmorris.com

 

[daronwilson]

I know we've beat this one to death, but...

IF proper grounding and bonding will eliminate ground loops, and you have an installation that is properly grounded and bonded, what is the point of checking for voltage on the shield between devices? It simply cannot be there can it if the installation is properly grounded and bonded can it? If there is voltage there, what does one do since the entire system is already properly grounded and bonded?

No cushion on the cheap seats, so it keeps me awake thinking about this stuff....


It is only my opinion, based on my experience and education...I am always willing to learn, educate me!
Daron J. Wilson, RCDD
daron.wilson@lhmorris.com

 

[johningram]

Thanks guys.
I think im going to try to un-ground one side.

I never did understand how to use a multi meter to check for differences in ground potential. Did he want me to stick the black wire to the sctp ground wire and the red to the ground in the wall socket? Or what?

 

[daronwilson]

Using the multimeter could be done a couple of ways, but I would basically check between drain wire of the sctp coming in and a good ground at the switch end, frame ground, ground bus bar, etc. BEFORE you plug it in. You should not be able to get much of a difference in voltage between the two points if the grounds are 'common'. You could also use the milliamp setting and see if any current flows between the two points.

Again, if you see less than one volt or so, you might be well enough bonded between points to bond both ends. Just be aware that if that does NOT allow good communications, or you burn out a port or NIC or something, don't rule out lifting the shield on one end and trying it that way.

Good Luck from the cheap seats!

It is only my opinion, based on my experience and education...I am always willing to learn, educate me!
Daron J. Wilson, RCDD
daron.wilson@lhmorris.com