Read both Electrical Engineering Times articles entitled
"Protecting Electrical Devices from Lightning Transients" to
appreciate which type of transient does damage. Destructive surges
seek earth ground. Never forget a damning question: where is surge
energy dissipated? An MOS does not absorb that energy. Review its
spec numbers. It cannot. Where is that energy dissipated? That EE
Times article describes what an effective MOS device does.

MOVs, gas discharge tubes, 3 mm 'carbons', and avalanche diodes (ie
MOS devices) are all shunt mode devices. All protect by doing same
concept. Same concept also makes lightning rods effective. All
connect a typically destructive surge to earth. Again, the damning
question. Where is that surge energy dissipated? No earth ground
means no effective protection. Applies even to MOS devices.

We routinely fixed things by learning how damage happened. For
example, a network of powered off computers were connected to plug-in
protectors. Protectors damaged adjacent computers. A surge entered on
black (hot) wire. Adjacent plug-in protector performed as an MOS
device does. Surge was shunted from black wire to white (neutral) and
green (safety ground) wire - just like an MOS device. Surge
distributed (shunted) to other wires and still seeking earth ground.
We learned by doing and by asking damning questions such as "where is
surge energy dissipated?" An adjacent protector (MOS, MOV, GDT,
etc) simply distributed that surge to more paths; to find earth ground
destructively. Surge also used network cable to find a third
computer and earth ground via that computer's modem (this sentence
specifically relevant to the OP's question). All three 'powered off'
computers were damaged. We replaced every semiconductor to make all
three computers working - by following the surge path. A subtle hint
as to how much knowledge and experience is behind this post. Surge
found earth ground destructively to earth when a plug-in protector was
adjacent to a computer AND too far from earth ground.

Where was surge energy dissipated? Protector (MOS, GDT, MOV, etc)
does not stop, absorb, or "block" the typically destructive surge.
Does not even claim to. Do you really believe anything will stop
what three miles of sky could not? That is what your GDT example
claims; what the MOS device must do according to DTC's post.

Show me those MOS energy numbers if it absorbs a surge. Absorbing or
blocking surges is not what an MOS device does. If MOS protectors
'block' surges, then neither phone nor ethernet communication would be
possible - just another reason why that MOS device does not 'block' as
assumed.

Telco also uses MOS devices. MOS devices are located where?
Adjacent to earth ground and up to 50 meters distant from
electronics. What makes a telco's MOS devices so effective? Superior
earthing AND that 50 meters separation means even better protection.
Telcos have over 100 years experience making surge damage irrelevant.
Same solutions apply to the OP's question.

The OP asked about an ethernet cable between two buildings. OP
needs a solution routinely installed in all telco COs where maybe 100
surges during every thunderstorm create no damage. Their effective
MOS device (and yes, that is what many COs use to eliminate lightning
damage) is located where each wire enters the building, adjacent to
the earthing electrode used by all protectors, and up to 50 meters
distant from electronics.

Surges are not voltage events, as was implied. Destructive surges
are current events. Notice that protectors are rated in current -
amps. Voltage rises as necessary so that same current will flow.
There is no 'stopping' a typically destructive surge. Voltage will
rise as necessary to 'blow through' that protector. Any attempt to
block a surge current means voltage increases. More energy is
dissipated in the wrong place. Same principle even makes Ben
Franklin's lightning rods effective. Effective protector shunts a
surge current to earth meaning less voltage and meaning less energy
dissipates in a building. No effective protector stops or blocks the
typically destructive surge. That Electrical Engineering Times
article describes what is necessary for surge protection. Where is
surge energy dissipated? Bud routinely ignores that question to
promote ineffective and profitable 'magic box' protectors. Where is
that energy dissipated - the damning question? Two front page EE
Times articles define surge protection. The NIST also states bluntly
where surge energy must be dissipated:
> You cannot really suppress a surge altogether, nor
> "arrest" it. What these protective devices do is
> neither suppress nor arrest a surge, but simply
> divert it to ground, where it can do no harm.

"What these protective devices do ..." - and that includes the MOS
device.

How does every ethernet wire make a connection to earth? Some use
MOS devices. But again, why does the MOS device provide protection?
Because massive energy is shunted to and absorbed by earth. MOS
protector to absorb that energy would, instead, just fail. No
protection. A protector that 'protects' by failing means no effective
protection. A protector must earth a surge AND remain functional.
What makes the MOS device effective? A short connection to earth
ground - a solution to the OP's question.

Ethernet cable must be earthed where it enters a building. That
means a low capacitance device. MOVs typically have excessive
capacitance which is why MOS type devices are sometimes used.
Protector (ie MOS device) is only as effective as its earth ground.
Again, the EE Times articles.

Provided are more answers to the OP's original question including
what makes that earthing connection - ie MOS devices. Stated is what
provides the protection and where energy must be dissipated. A long
list of reasons and examples why a single solution means protection
inside electronics (ie 1500 volt ethernet interface) is not
overwhelmed. This solution that has been standard and understood for
over 100 years.

On Jan 28, 12:26 pm, DTC <m...@nothingtoseehere.zzx> wrote:
> I think ya'll aren't on the same page here, compounded by taking
> quotes and references out of context.
>
> There are two kinds of protection. The exact description may vary,
> but its still a function of clamping the over voltage and shunting
> the over voltage to ground, may it be done in a single device or with
> two separate devices.
>
> A MOS "surge protector" found in an outlet strip merely clamps
> (or blocks) excessive voltage up to the point where the MOS device
> fails. It offers virtually no lightning protection, per se.
>
> A gas tube (or old style carbon block) arrestor shunts the over
> voltage to ground. Its very robust, but if the voltage is not high
> enough to arc over the arrestor (as in the leading edge of the spike),
> then the "surge protector" will block it.
> ...