🚀 go-pugleaf

As soon as a read about 'fusible' components that protect other
components, then I know the poster does not have electrical training. 
Yets put up a few simple numbers.  Destructive transients occur in
microseconds.  Even the fastest blowing fuse takes tens of
milliseconds.  This is a classic myth no based upon technical
knowledge:
> either have a fusible transformer that 'sacrifices' itself in the
> event of a surge in order to protect the PC, or an optical coupling 
> that sort of does the same thing.

  There is no sacrifical component to save electronics - except where
urban myth replaces electrical knowledge.  Modems are rarely damaged
by phone line surges.  In the UK, this is different because in North
America, all incoming phone lines have a 'whole house' protector
installed free by the telco.  For some reason, before the transistor
became a household standard, BT stopped installing premise interface
protection.  UK modems suffer surges both from AC electric and from
phone line because neither utility has 'whole house' protectors.


  Furthermore the concept of breakdown voltage is totally unknow to
the poster.  Even optocouplers have breakdown voltages typically
starting at 2000 volts.  But computer power supplies also have
internal protection that will withstand common mode transients of
about 2000 volts - as has been industry standard for 30 years.   IOW
appliances already have effective protection that assumes the 'whole
house' protector will earth the direct strike before entering the
building.  Internal protection sufficient to make those indirect
lightning strikes nothing but urban myth. Why then would a transient
cause hardware damage?  Because of something called breakdown voltage
and because a 'whole house' protector did not shunt before that surge
could enter a building.

  Effective protection shunts a surge to earth ground before the surge
can enter a building.  It is why phone line surges in North America
are usually not the reason for modem damage.  Every incoming utility
must have some sort of 'less than 3 meter' connection to central earth
ground - either by hardwire or by 'whole house' protector.  Once
inside the building - especially on AC electric - then the surge has a
destructive path to earth ground via appliances - ie that modem.  Most
modems in North America suffer damage from surges on AC mains.  Surge
damage is easier in UK where even the phone line has no 'whole house'
protector.

  Surges do not enter a building, damage some component, then stop. 
They first seek a circuit to earth ground.  Only then does something
in that circuit fail. The only effective protection is earthing a
surge before it can enter the building - as Franklin demonstrated in
1752 - as research papers from Westinghouse and GE demonstrated in the
1930s - as is standard everywhere that surge damage is not an option -
as is standard when the home owner wants effective protetion at costs
tens of times less per protected appliance.

  That claim of EMP induced surges is only another urban myth that
falls flat on its face when the myth purveryor cannot provide numbers
to support that claim.  Appliances have such good internal protection
that only the direct strike causes damage.  Damage traceable to human
failure because human did not install effective 'whole house'
protectors for a 'less than 3 meter' connection to central earth
ground.

"Nigel Kendrick" <nkendrick@stiknowledge.dotcodot.uk> wrote in message news:<bh929v$okn$1@news8.svr.pol.co.uk>...
> Sorry, but...
> 
> >Modems are rarely damaged by phone line surges.
> 
> Not so - modems are **often** damaged...etc
> 
> >Modems even have galvanic isolation that separates modem for phone line.
> 
> They'll either have a fusible transformer that 'sacrifices' itself in the
> event of a surge in order to protect the PC, or an optical coupling that
> sort of does the same thing.
> 
> But modems have a
> > direct connection to one AC mains electric wire.
> >
> 
> Nope - full isolation via the power adaptor (so-called 'double insulated' if
> in a plastic case). If internal to a PC, the only power connections with be
> to the DC power bus. There should be no circuit path between the AC input
> and the low-voltage DC side unless an auto-transformer has been used -
> highly unlikely.
> 
> >   Surges must first form a complete electrical circuit.  >Current passes
> completely through everything in that circuit.  Only then does one part of
> that circuit fail - become surge damaged.
> 
> In the event of a lightning strike or phone line surge caused by a distant
> strike, 'circuits' will include air paths due to the high potential
> difference ('voltage' if you must) of the surge. Simply having the kit on a
> surface that has a low-to-high resistance path to earth (ie a table with
> metals legs or a nearby metal desk lamp or grounded PC case) may form a
> circuit. Also remember that nearby/direct lighting strikes can contain
> high-frequency components that will drive voltage through insulation.
> 
> >   Telephone lines already have effective surge protector.
> 
> Not for a major surge.
> 
> > Phones that do not connect to AC mains are typically
> > not damaged by such surges.
> 
> Sorry, not so. HF components and EMP-induced currents can fry kit in the
> vicinity even if there's no direct connection to anything. - a mobile phone
> left in the middle of a desk can be fried by a nearby lightning strike.
> 
> NK