Underground Data Connection Between Buildings

[Darkness]

There is a lot of talk these days when linking two buildings together underground using Cat5e/6 due to earthing differential between the two buildings from what I read and why Fiber is recommended.

Does this earthing differential is a factor if the two buildings had separate electric supplies with their own earth such as connecting between two neighbours houses?

So if you had a garage/workshop 30m or more down the yard and you wanted to run cat6, you would be fine especially if the garage's power electric supply comes from the same meter and same earthing system?

Thanks

 

[deadshort]

Hi, i would always advise installing surge suppressors at at the port on the switch which is connected to an external data cable.

Regards,

DS

 

[Aragorn84]

If you use unsheilded cable, there would be no linked earth between the two buildings anyway.

 

[Darkness]

I know some people use a Lightning Arrestor on runs especially for Poe equipment such as cameras but I've never seen anything that is Gigabit compatible and for use with cat6 between switches.

@Aragorn84 If that is the case, why do they still state it being a problem?

 

[bernardgreen]

Whether a difference in Earth potentials will create a problem for CAT 5/6 interconnection depends a lot on the construction of the data ports in the equipment that are interconnected.

If one of the interfaces has opto coupling and full galvanic separation then ground potential differences are unlikely to cause any significant problems.

The opto coupling and full galvanic separation can be in a module external to the equipment.

The exception is when a lighting strike creates a potential difference that exceeds the break down voltage in the opto coupling and galvanic separation, this break down voltage is typically 2000 volts or more

 

[plugwash]

AIUI standard for UTP ethernet is transformer isolation rated to 1.5KV. So if the devices are in-spec then they should survive fault voltages resulting from the LV mains system.

 

[JohnW2]

AIUI standard for UTP ethernet is transformer isolation rated to 1.5KV. So if the devices are in-spec then they should survive fault voltages resulting from the LV mains system.

Is not at least some of the perceived (hypothetical?) risk related to potential fault currents which might theoretical flow in non-isolated 'earth' connections between buildings? In fact, if there were such a non-isolated 'earth' connection (which I doubt is common) between buildings, would it not have to be treated as an extraneous-conductive part at at least one, probably both, ends?

Kind Regards, John

 

[bernardgreen]

It is the ground connection via a screen cable that gets involved carrying high currents. The signal conductors being isolated make it seem OK, but it is when the metal case of an external isolation module that is connected to the screen of the inter-site data cable gets in contact with the local ground that sparks can fly.

In answer to John's question.

The screen of an incoming data cable is an extraneous conductor but bonding it to the local MET creates a path along which some very high currents can flow. Screened cables between sites are the ones mostl likely to suffer from burn out.

When the data is transmitted as the difference in voltage between two wires ( as apposed to a single wire with 0v return ) then screening is not always needed and the 0 volt and / or Grounds of the two systems do not need to be connected to each other.

 

[JohnW2]

In answer to John's question. ... The screen of an incoming data cable is an extraneous conductor but bonding it to the local MET creates a path along which some very high currents can flow. Screened cables between sites are the ones mostl likely to suffer from burn out.

Indeed - that's what I was referring to.

When the data is transmitted as the difference in voltage between two wires ( as apposed to a single wire with 0v return ) then screening is not always needed and the 0 volt and / or Grounds of the two systems do not need to be connected to each other.

They certainly don't need to be connected. One hopes that the data pairs will be properly isolated but, if they're not, is there not a possibility that they would be earth-referenced?

Kind Regards, John

 

[Lectrician]

There used to be problems with token ring and twinnax type networks using coax, but not ethernet. It's a myth. Fibre between buildings is used due to lengths usually.

 

[plugwash]

They certainly don't need to be connected. One hopes that the data pairs will be properly isolated but, if they're not, is there not a possibility that they would be earth-referenced?

Well if you are talking about devices that grossly violate the standard then anything is possible.

 

[JohnW2]

They certainly don't need to be connected. One hopes that the data pairs will be properly isolated but, if they're not, is there not a possibility that they would be earth-referenced?

Well if you are talking about devices that grossly violate the standard then anything is possible.

Fair enough - in that case, what I suggested may not usually be 'possible'. I have no idea what the standards require.

Kind Regards, John

 

[plugwash]

40.6.1.1 Isolation requirement
A PHY with a MDI that is a PI (see 33.1.3) shall meet the isolationrequirements defined in 33.4.1 .
A PHY with a MDI that is not a PI shall provide electrical isolation between the port device circuits, including frame ground (if any) and all MDI leads. This electrical isolation shall withstand at least one of thefollowing electrical strength tests:
a) 1500V rms at 50Hz to 60Hz for 60s, applied as specified in subclause 5.2.2 of IEC 60950-1:2001.
b) 2250Vdc for 60s, applied as specified in subclause5.2.2 of IEC 60950-1:2001.
c) A sequence of ten 2400V impulses of alternating polarity, applied at intervals of not less than 1s. The shape of the impulses shall be 1.2/50us (1.2us virtual front time, 50us virtual time of half value), as defined in IEC 60950-1:2001 AnnexN.

There shall be no insulation breakdown, as defined in subclause 5.2.2 of IEC60950-1:2001, during the test. The resistance after the test shall be at least 2MΩ , measured at 500V dc.

Thats for 1000base-t but i'm pretty sure the standards for the other speeds are much the same. The "PI" reference reffers to poe stuff which has similar requirements but expressed slightly differently.