Re: System Saver IIgs -- ID of burned part?-- line protection

[Rubywand <rubywand@swbell.net>]

w_tom writes ...
> 
>   You are confusing length of cord with connection to earth ground.  That is 6 feet
> of cord, maybe 50 feet of in-the-wall 12 AWG to the fuse box, then some more distance
> to earth ground.  All too far.  That distance between power strip and computer -
> irrelevant by comparison.
>

     Agree. The distance to earth ground reduces the value of using a cord's
Ground lead as a shunting point when dealing with pulses which have a V
mainly with respect to earth ground. On the other hand, so what? The
computer, etc. is subject to damage from such a pulse only in so far as it is
earth grounded. The better the earth ground connection at the computer, the
greater the relative V with no protection in place. But, then, the better the
earth ground, the more effective nearby MOV protection vis-a-vis earth ground
will be. That is, 'it's all relative'.

     Meanwhile, you have the resistance (DC and reactive) of the power cord
and house wiring acting to suppress the pulse and increase the value of
shunting at the end of the line. And you have capacitive shunting along the
length of the cord (and house wiring) plus the breakdown voltage of
insulation for the leads in the cord.

 
>   Effective surge protection uses wire as part of the surge protection system.  If a
> surge is shunted to earth ground with 50' of wire between surge protector and
> transistors, then that wire adds protection.  However, if the surge protector shunts
> a surge at the appliance, then that long wire to earth ground encourages the surge to
> find other, destructive, paths to earth ground.  One common path is through a
> computer's modem to ground via the phone line - destructively.

     Granted, if you shunt a glitch to some kind of grounding system, then
you place a V on the grounding system. At any given point, this would raise
local earth ground V shared by local equipment and cabling. The standard
glitch protections for modem and phone lines built into many power strips and
UPS should be more than adequate.


> 
>   50' of 12 AWG wire may leave the surge protector at 15,000 volts during a 100 amp
> surge.  That 15,000 volts will definitely finds alternative paths to earth ground -
> because the surge protector was too far from earth ground.  Since the surge protector
> was too close to the appliance, then one alternative path is often through the
> appliance.
>

     Sounds like a lightning strike situation. If you live in a place where
pulses like this are common, some sort of special grounding system plus
arrestors and large MOVs on the house mains might be worth the bother and
expense.

 
>   If your MOVs explode or vaporize, then they were not just undersized.  They were
> grossly undersized.  I believe that definition was expressly made in the previous
> post.  MOVs should degrade slowly with use.  And when properly sized, that
> degradation slows exponentially meaning that it is easy and not expensive to install
> MOVs that will last longer than you.
>

     Makes sense. Our System Save IIgs has been in use for over a decade. The
MOV's in it have never exploded, vaporized, or, even, fried. So, for this
location, they are properly sized. Someone who has had problems with zapped
MOVs should probably get higher capacity models.

 
> 170 V is not a surge.

     Well, not compared to a 15kV 100 Amp zonk! By "surge" I mean a
relatively long duration, usually differential mode, over-voltage condition.
The MOV that blew in Mitch's System Saver IIgs is the one which protects
against differential mode garbage.


> It is well below what any surge protector will observe.

     That is true with regard to the V130LA10. The ones in our collection all
clamp right around 200V. On the other hand, we have MOVs that reliably clamp
at 50V. You can obtain surge/glitch protection for all kinds of voltage
levels.


> 
>   UPSes, especially the sub $200 version connect your computer direct to AC mains
> when not in battery backup mode.  That UPS relay is 1000 times too slow to afford any
> surge protection.

     What you call a "surge" is often referred to as a "spike".

     UPS's work fine as protection against relatively long duration
over/under voltage situations. This, in turn, works to protect against
spiking. Spikes are more likely during sustained off-voltage (high or low)
situations. Just being taken off-line and placed on inverter power for a few
seconds prevents exposure to especially dangerous spikes.


> Furthermore, destructive surges are common mode transients.

     Which, really, is very fortunate.



> They pass right through UPSes sometimes
> without ever being seen by the UPS.  Understand the difference
> between common vs differential mode surges.  It is why UPSes are so
> stingy with detailed specs.  Look at these for a better line UPS.  Notice on page 5 -
> no common mode surge protection!:
>      http://www.apcmedia.com/litpdfs/pdfs_new/996%2D0383e.pdf
 ....

     Got a 404 error; but, thanks. I'll check around for some information on
better UPS's.




Rubywand