dslreports logo
 
    All Forums Hot Topics Gallery
spc
Search similar:


uniqs
1100
wirelessdog
join:2008-07-15
Queen Anne, MD

wirelessdog

Member

Proper Grounding

I figured since we haven't had an all out drag out fight on this for a while... LOL

actually my question is quite simple. I understand the importance of using shielded wire and properly installed ends to alleviate much of the Ubiquiti ESD issues. Here is my question:

Do I also need to use shielded patch cables from the PoE to the switch or just from the PoE to the radio for surge protection purposes?

Thanks!

Semaphore
Premium Member
join:2003-11-18
101010

Semaphore

Premium Member

I install the Tycon suppressor »tyconpower.com/products/ ··· heet.pdf in the AP enclosure, and another between the POE and the STP befor it leaves the enclosure to head up the tower.

Have not yet (>3 years) blown a switch port with this config. Have blown LOTS of AP Ethernet ports, POE's and switch ports before using the Tycon surge suppressors.

Have never used STP from the POE to the switch.
LLigetfa
join:2006-05-15
Fort Frances, ON

LLigetfa to wirelessdog

Member

to wirelessdog
The PoE brick is grounded via the third prong so as long as your AC protection ground is adequately bonded, then using FTP cable between the brick and the switch would see little to no benefit. If not properly bonded, then the FTP ground could in fact introduce ground loops.
wirelessdog
join:2008-07-15
Queen Anne, MD

wirelessdog

Member

said by LLigetfa:

The PoE brick is grounded via the third prong so as long as your AC protection ground is adequately bonded, then using FTP cable between the brick and the switch would see little to no benefit. If not properly bonded, then the FTP ground could in fact introduce ground loops.

Perfect. Thanks!

On another note, is it beneficial to bond the tower ground to the A/C ground?
LLigetfa
join:2006-05-15
Fort Frances, ON

LLigetfa

Member

Yes, AC and tower should be bonded.

TomS_
Git-r-done
MVM
join:2002-07-19
London, UK

TomS_ to LLigetfa

MVM

to LLigetfa
said by LLigetfa:

The PoE brick is grounded via the third prong

Is that for dissipating surges, or just an electrical safety ground?

WHT
join:2010-03-26
Rosston, TX

WHT

Member

The third wire ground is for electrical safety only for a specific problem - specifically if the live conductor inside a device touches the outside metal case. It will create a fault condition where the current flowing from the live conductor to the common (earth) ground exceeds the current rating of the circuit breaker.

It does not provide any safety from a ground fault - that is what ground fault interrupters are for.

The metallic path of the shielded cable and third wire ground will safely dissipate high voltage (and very low current) electrostatic events, nothing more.
LLigetfa
join:2006-05-15
Fort Frances, ON

LLigetfa to TomS_

Member

to TomS_
No, it is not for dissipating significant surges and does not negate the need for surge protection. It is bonded to the shield on the ethernet port and is intended to bleed off a static charge.

WHT
join:2010-03-26
Rosston, TX

WHT

Member

HA! Beat ya by five seconds.
lutful
... of ideas
Premium Member
join:2005-06-16
Ottawa, ON

lutful to wirelessdog

Premium Member

to wirelessdog
said by wirelessdog:

is it beneficial to bond the tower ground to the A/C ground?

Ideally, this bonding conductor should run below ground level between the tower's grounding system and the utility grounding rod. That way, you don't compromise lightning surge protection for electrical safety.

The absolute worst case scenario: bonding conductor starts near the top of the tower, goes down a bit, but then dangles diagonally across the yard on its way to the utility ground. That allows AC ground to jump to several thousand volts during a nearby lightning storm.
lutful

lutful to TomS_

Premium Member

to TomS_
said by TomS_:

said by LLigetfa:

The PoE brick is grounded via the third prong

Is that for dissipating surges, or just an electrical safety ground?

Two previous comments on this question imply that it can only handle "electrostatic" events.

Any indoor surge protector which complies with ANSI/IEEE C62.41 standard is designed to dissipate a defined test waveform with ~ 90 joules of energy between the ground prong and the utility ground through ~60ft of copper. They assume typical residential installation by electricians.

WHT
join:2010-03-26
Rosston, TX

WHT

Member

said by lutful:

Two previous comments on this question imply that it can only handle "electrostatic" events.

Mostly correct. To define the scope of what we're talking about...

* Electrostatic event - static buildup from blowing particulate matter. The 26 ga wire can certainly drain off almost the infinitesimal current of a high voltage static charge.
* Low current induced foreign power events - The 26 ga wire could safely drain off around 4 amps.
* High current induced foreign power events - The 26 ga wire will likely fuse open.
* Near strike lightning - The 26 ga wire will most likely fuse open.

With that said the 26 ga can handle at least, but not much more, simple electrostatic events.
lutful
... of ideas
Premium Member
join:2005-06-16
Ottawa, ON

lutful

Premium Member

Click for full size
said by WHT:

said by lutful:

Two previous comments on this question imply that it can only handle "electrostatic" events.

Mostly correct. To define the scope of what we're talking about ...
...
With that said the 26 ga can handle at least, but not much more, simple electrostatic events.

Let me post a composite photo illustrating what I was talking about.

I sort of pioneered this "power-over-ethernet" thingy in 1999, a few years before the term PoE was coined, designed quite a few high-end PoE circuits a decade ago, and since then a few test circuits that can blast Chinese crap with various IEEE surge waveforms to see how they actually cope.
plugnothing
join:2014-09-03
US

plugnothing to wirelessdog

Member

to wirelessdog
Here was one frustrating experience I had with STP.

I mounted a Nano Loco M900 as an AP on a residential outbuilding, using STP to the POE and UTP to the switch. I had frequent intermittent failures of the radio--everytime I used electricity in the building. I thought it was related to power/voltage problems, so I put in a UPS. No help there. So I put in a ping watchdog-rebooter. No help; multiple reboots, no service.

Finally I trouble-shot the problem successfully, with the help of somebody on the other end of the link, with a cell phone: it was the lighting dimmers. Turn them on, the network goes down. The radio, on the other hand, stayed up. So it wasn't radio interference; it was network interference. It seems this happens with DSL a lot, so there was a pretty good body of help there.

I thought it might be the shielding acting as an antenna, so I replaced the STP with a UTP cable. Problem solved--but, no ground for the radio! So I put in an APC like in the picture. Problem recurs. LIfted the ground on the APC unit--problem gone.

Net: the ground is somehow picking up the dimmer interference, presumably in a ground loop, even though the electric in the barn is installed with code grounding, using brand new materials, by a well-respected licensed electrician.

Any ideas on how to solve that, so I can have a grounded radio and a functional link to the AP?
wirelessdog
join:2008-07-15
Queen Anne, MD

wirelessdog

Member

said by plugnothing:

Any ideas on how to solve that, so I can have a grounded radio and a functional link to the AP?

Hire an electrician before something catches on fire!
plugnothing
join:2014-09-03
US

plugnothing

Member

Ha, ha, ha. Did you see that the electrician did the installation? He was there during the troubleshooting process. No expense has been spared on this.

The real solution is to buy a radio with its own ground screw, but that means no UBNT.
LLigetfa
join:2006-05-15
Fort Frances, ON

LLigetfa

Member

Grounding problems at barns are common due to the voltage drop on the long neutral run and unbalanced multi-wire circuits. I contend the electrician was either constrained by your budget or was not well qualified.

As for radio grounding, one can ground at the mast with one of these.
»www.ubnt.com/accessories ··· otector/
plugnothing
join:2014-09-03
US

plugnothing

Member

Thanks for the link. It looks a lot like the APC one that Lutful posted above; that's what I am using, and it seems to do the same thing. Is it functionally different in some way I am missing?
LLigetfa
join:2006-05-15
Fort Frances, ON

LLigetfa

Member

I did not post it as an answer to your ground loop problem. If you have stray voltage on ground, then having multiple disparate grounds (not well bonded) cause problems. If the radio had a ground post that was bonded to either the ethernet shield ground or the PoE ground, you could have the same problem.
plugnothing
join:2014-09-03
US

1 edit

plugnothing

Member

Thanks for the help, wirelessdog and Lligetfa.

WHT
join:2010-03-26
Rosston, TX

WHT to wirelessdog

Member

to wirelessdog
LLigetfa points out the problems of multiple disparate grounds.

I stray a bit off topic to mention some things overlooked. Consider four different needs:
1. ESD event grounding. This isn't very difficult to achieve. Shielded cable bonded to the electrical outlet's third-wire ground is adequate.

2. Ground loops (different voltage potentials) between different buildings can be reduced by adequately sized return ground wire to master utility ground to reduce voltage drop, but...

3. An adequately sized wire for the above may not be of sufficient gauge to create a fault condition, where a fault condition is where sufficient current can flow to trip a circuit breaker.
At the far-end location you might have a $5,000 PolyPhaser ground system that will safety dissipate a direct lightning hit, but without adequately sized (for current) to the head-end you may have a fire safety problem if there is a short between the live conductor and the PolyPhaser ground that won't be cleared by a head-end circuit breaker tripping.

4. Protection of devices with surge protectors.
LLigetfa
join:2006-05-15
Fort Frances, ON

LLigetfa

Member

The defacto NEC standard ground of two ground rods is intended to be able to transfer enough current to trip a normal non-GFI circuit breaker. In actuality, that is not always the case.

The service neutral is bonded to the protection ground at the service entrance. I have seen where the neutral had failed and the bond to a copper water line was passing so much current that it warmed the water in the pipe. I discovered this while troubleshooting issues on a computerized fuel pump delivery system.
plugnothing
join:2014-09-03
US

plugnothing to wirelessdog

Member

to wirelessdog
Keep talking, guys. This is a fascinating and useful discussion.

One of the problems is that the main utility ground is so poor that there is a standing ground loop between the local ground and the utility ground in every house in the neighborhood. How is that even possible? Welcome to rural New England!

WHT
join:2010-03-26
Rosston, TX

WHT

Member

What do you mean by "local" ground. The "utility" ground would be the ground under the meter box.

nunya
LXI 483
MVM
join:2000-12-23
O Fallon, MO
·Charter

1 edit

nunya to LLigetfa

MVM

to LLigetfa
said by LLigetfa:

The defacto NEC standard ground of two ground rods is intended to be able to transfer enough current to trip a normal non-GFI circuit breaker. In actuality, that is not always the case.



I disagree with that. The intent of grounding electrodes is not to clear faults. Earth is never considered an effective path for clearing faults. In some cases it may coincidentally do just that, but should never be relied upon. It's a safety net.
Proper bonding (creating a fault current path back to the neutral point) is what is intended to clear faults.


The terms grounded, grounding, ground, bond, etc... get mixed up and tossed around a lot.
The purpose of supplemental grounding electrodes is to simply lower the impedance to ground (earth).

Typical 8-10 copper clad ground rods are not the best grounding electrodes. They are mediocre. Good enough for most applications with damp soils. They have a finite effective life span. This is why the ground rod / plate is often referred to as the "supplemental" ground.

Keep in mind, the NEC / CEC is the bare minimum you can get away with. There are are better standards out there.
These manufactures paid a lot smart engineers to design a proper installation method for their product. The #1 problem I see is that installers simply do not follow the directions provided with the product (be it an antenna, radio, mast, dish, etc...).

Just yesterday I was on a job where lighting came in through a DirecTV dish (focused from a nearby strike). $15,000 and counting. The DirecTV installer didn't make any attempt to properly ground the satellite system.
nunya

1 recommendation

nunya to plugnothing

MVM

to plugnothing
said by plugnothing:

One of the problems is that the main utility ground is so poor that there is a standing ground loop between the local ground and the utility ground in every house in the neighborhood. How is that even possible?

For clarification, the utility ground is usually made up at each transformer (pole or pad). The local ground is at the meter socket.
A dirty open on the neutral can cause that on a single service. The utility (transformer) ground may also be dirty.
If you are seeing it on multiple services (my theory without knowing the whole scenario): When you have a neighborhood with each house bonded through the metallic underground water pipes, and get a neighbor with an open neutral - all that current is trying to find it's way back through the adjoining neighbors houses. You can test for this by turning off the main breaker and measuring for objectionable current on the grounding electrode conductor coming from the water pipe.
I wouldn't call it a "ground loop" per se, but objectionable current instead.
LLigetfa
join:2006-05-15
Fort Frances, ON

LLigetfa to nunya

Member

to nunya
said by nunya:

said by LLigetfa:

The defacto NEC standard ground of two ground rods is intended to be able to transfer enough current to trip a normal non-GFI circuit breaker. In actuality, that is not always the case.

I disagree with that.
...
Proper bonding (creating a fault current path back to the neutral point) is what is intended to clear faults.

Of course the neutral is the primary path. That is a given and I should not even have to mention it. The protection ground is a backup (and a supplement) to when the primary fails.

The neutral can and often does have potential in less-than-ideal situations such as the barn mentioned above and also often seen around docks.
plugnothing
join:2014-09-03
US

plugnothing to nunya

Member

to nunya
That is the kind of information I am looking for. I have copied it and will run it past the electricians at the first opportunity. Thanks again for you patience.
"Explain as you would to a child...." (GalaxyQuest)
raytaylor
join:2009-07-28

raytaylor to wirelessdog

Member

to wirelessdog
I will probably be mocked, but I dont use shielded cat5. I only use unshielded.
And my radio failure rate has only been something like 3 out of 500 radios deployed over 4 years if we exclude 4 bullets filling with water.
LLigetfa
join:2006-05-15
Fort Frances, ON

LLigetfa

Member

Ja, there are two schools of thought. You're in the bird-on-a-wire school.

Grounding is something that needs to be done right or not done at all. No grounding stands a better chance than poor grounding.