Re: Are PC surge protectors needed in the UK?
Andrew Gabriel
Lem <nom...@mail.com> writes:
> Are surge protectors on the main power supply actually needed in
> the UK?
No, at least not to protect PCs.
> here in the UK we have few overhead mains power lines and have a
> relatively steady mains power supply when compared to many other
> countries (including the US).
>
> However there seem to be very many surge protector products
> advertised for sale in the UK (Argos, Maplins, etc).
Lots of people run a particular OS which is known for its
instability, and you can sell them just about anything if
you suggest it might make their systems more stable. Of
course it doesn't, but that just means they'll try something
else (except changing the OS;-). Surge protectors are one of
the many items on the list that such people will try.
The other issue is that the multi-way trailing socket blocks
got down to the point where they're only a couple of pounds
each, or even less. By adding a extra few pence worth of
components, you can call them surge protected and sell them
for 3 times the price. Brings in more profit.
> I am quite sure it is not bad practice to use a surge protector but
> in fact I have never known anyone who has has a problem from a
> surge coming in through the power supply.
True. Even if you have millions of pounds worth of equipment
on your supply, it isn't worth it, the occurance is so rare,
so it certainly isn't for a few home PC's.
What I have seen several times is damage caused by a surge
induced by lightning on a phone line. Of course, the mains
surge protector will do nothing to protect against that.
> So personally I don't bother using a surge protector on my PC.
>
> Am I being too complacent?
Not in my opinion.
If I saw something at less than rip-off prices for protecting
against surge on phone line, I might consider that.
--
Andrew Gabriel
St Dom
equipment failure not an issue then?
something as simple as a desk fan switched on & off and supplied via the
same socket as a PC can create a transient over voltage of 1kV
Sure it will not knock the PC out but it will reduce the life of the kit.
"Andrew Gabriel" <and...@cucumber.demon.co.uk> wrote in message
news:ccjpam$ltd$1...@new-usenet.uk.sun.com...
someone
"Lem" <nom...@mail.com> wrote in message
news:9520A0CF6...@130.133.1.4...
> Are surge protectors on the main power supply actually needed in
> the UK?
>
> relatively steady mains power supply when compared to many other
> countries (including the US).
>
> However there seem to be very many surge protector products
> advertised for sale in the UK (Argos, Maplins, etc).
>
> in fact I have never known anyone who has has a problem from a
> surge coming in through the power supply.
>
>
> Am I being too complacent?
I have no UK electrical power system knowledge.
I do however have extensive electrical utility knowledge.
I would never install a computer without a UPS.
What is the age of the installed cables? When a cable faults the surge is
substantial.
Transformers, capacitors, breakers, generators ...all fault at times and
they result in heavy electrical surges.
HDD are frequently corrupted due to power events.
~$40 USD (350VA) with $15000 of equipment insurance is worthwhile.
greywolf42
news:9520A0CF6...@130.133.1.4...
> Are surge protectors on the main power supply actually needed in
> the UK?
>
> here in the UK we have few overhead mains power lines and have a
> relatively steady mains power supply when compared to many other
> countries (including the US).
>
> However there seem to be very many surge protector products
> advertised for sale in the UK (Argos, Maplins, etc).
>
> I am quite sure it is not bad practice to use a surge protector but
> in fact I have never known anyone who has has a problem from a
> surge coming in through the power supply.
>
> So personally I don't bother using a surge protector on my PC.
>
> Am I being too complacent?
It all depends on your tolerance of risk.
It doesn't have to hit a power pole. A friend of mine lost 2 PCs when
lightning hit a tree next to her house, then jumped to the house (knocking
some siding off in the process).
Surge protectors are cheap. Or just back up your data regularly.
--
greywolf42
ubi dubium ibi libertas
{remove planet for return e-mail}
Soren Kuula
> It doesn't have to hit a power pole. A friend of mine lost 2 PCs when
> lightning hit a tree next to her house, then jumped to the house (knocking
> some siding off in the process).
Soren
Anthony
>
> Does anyone have a reference to HDDs getting corrupted by power
> events on the mains power supply?
References, no, practical, hands on experience, yes. I've seen this
several times, not only the HDD, but I've had transient power problems
take out motherboards too.
Examples:
Case 1: Home computer (this one) tree fell across a main line (11kva I
think), caused a surge prior to the stepdown transformer kicking out,
corrupted an almost new 40 gig hdd. Fortunately, an LLF fixed it.
Case 2: Engraving lasers at work, fed from the bus, kept killing HDD's
and motherboards. Ultimately traced to transient voltage spikes,
installed an AVR UPS. Failures were occuring once to twice a week, after
the AVR UPS was installed on each machine, we have had Zero failures, in
over a year.
--
Anthony
You can't 'idiot proof' anything....every time you try, they just make
better idiots.
Remove sp to reply via email
Pyriform
> What I have seen several times is damage caused by a surge
> induced by lightning on a phone line. Of course, the mains
> surge protector will do nothing to protect against that.
>
>> So personally I don't bother using a surge protector on my PC.
>>
>> Am I being too complacent?
>
> Not in my opinion.
> If I saw something at less than rip-off prices for protecting
> against surge on phone line, I might consider that.
I have a surge protector which includes protection for the phone line,
so that's where the ADSL modem gets plugged. Whether it would actually
work or not is another matter of course. My complacency lies in not
bothering to research the matter thoroughly, on the grounds that if I
was buying snake oil, it was at least *cheap* snake oil...
For the mains, online UPS is the proper solution, but I can't say I'm
unduly concerned about not having one. Power provision in the UK is
fairly reliable at the moment, although political and business
imperatives will probably conspire to make it worse in the future.
w_tom
more money per protected appliance. Furthermore it does not
even claim to protect from the typically destructive
transient. Protectors do not stop, block, filter, or absorb
destructive transients. Ineffective protector manufacturers
get one to wish that is how they work. In reality, the
protector is not protection. Protector and protection are two
separate components of a surge protection system. Effective
systems must include the protection. And the connection to
protection is either a hardwire (less than 3 meters) or a
protector (also part of a less than 3 meter connection).
In short, the protection is called single point earth
ground. Destructive surges may enter the building seeking
earth ground. If not earthed (either by hardwire connection
or by surge protector), then the destructive surge may find a
path to earth ground via computer. One classic example is due
to a direct strike to lines highest on utility poles - AC
electric. Incoming on AC electric, through computer and its
modem, then outgoing to earth ground via phone line. Many
then *assume* the surge entered on phone line, damaged modem,
then stopped - a violation of even primary school science.
Effective protection means all incoming utilities are
earthed before entering the building. All must be earthed to
the same single point earth ground. That means even the CATV
wire drops down to earth ground, connects ground block 'less
than 3 meters' to that earth ground, and only then rises back
up to enter building. Again, no surge protector required
because earthing is accomplished by a direct and short
hardwire connection.
These concepts are explained further including some examples
of 'whole house' protectors for AC mains at:
"RJ-11 line protection?" on 30 Dec 2003 through 12 Jan 2004 in
pdx.computing at
http://tinyurl.com/2hl53 and
"strange problem after power surge/thunderstorm" in
comp.dcom.modems on 31 Mar 2003 at
http://tinyurl.com/2gumt .
Additional information on how surge protectors work, how
they are rated, installed, etc was posted in:
"Opinions on Surge Protectors?" on 7 Jul 2003 in the
newsgroup alt.certification.a-plus at
http://tinyurl.com/l3m9 and
"Power Surge" on 29 Sept 2003 in the newsgroup
alt.comp.hardware at
http://tinyurl.com/p1rk
One industry professional demonstrates how two structures
are protected. Notice every wire entering each structure
(building and tower) must first connect to single point
ground. Even the buried phone wire carries a potentially
destructive transient which is why even buried wires must
enter building at the service entrance with the 'less than 3
meter' connection to earth ground:
http://www.erico.com/public/library/fep/technotes/tncr002.pdf
How do we identify ineffective protectors? 1) No dedicated
connection to earth ground AND 2) manufacturer avoids all
discussion about earthing. A surge protector is only as
effective as its earth ground - the protection.
Those ineffective protector manufacturers fear you might
learn about the essential earth ground AND discover that
plug-in protectors cost tens of times more money per protected
appliance.
w_tom
upon what the filesystem is. For example, if using FATxx
filesystems, then a loss of electrical power at the right time
can even erase files from that drive. Just another reason why
the technically informed want NTFS filesystems on drives; not
FAT.
Transients should never be a problem to disk drives or
memory. Based upon how these devices are connected, then a
differential type transient would be required to cause
damage. But all minimally acceptable power supplies must have
the essential function called overvoltage protection - that
makes a differential transient not possible.
That is the theory as well proven by power supplies even 30
years ago. Reality is the gross profits obtained by dumping
inferior supplies in North America where so many computer
assemblers don't even have basic electrical knowledge. Many
clones are not built and sold missing the essential
overvoltage protection because the assembler only understands
one specification - dollars. It's called a bean counter
mentality. If the power supply is sold on the cheap, (ie full
retail price is less than $60), then this and other critical
functions are simply *forgotten*. Does not matter. Consumer
is only to be fleeced.
If the destructive differential transient does occur, there
is no overvoltage protection circuit to protect that hardware
- do to power supply purchased by a bean counter. No
problem. Myth purveyors then quickly blame speculated surges,
and recommend overpriced, typically undersized, and
ineffective plug-in protectors.
Up front - does the power supply specifically state that
overvoltage protection is provided? If not, then it probably
is a man-made disaster just waiting to destroy disk drive,
data, and other computer components.
This overvoltage protection is something completely
different from another disk drive threat to FAT filesystems -
blackouts and brownouts.
w_tom
destructive transient. Kilometers of air could not stop the
transient. Do you think that silly little RCD switch contact
will do what kilometers of sky could not? Again, protection
is about shunting (diverting, connecting, shorting) a
transient to earth ground. There is no way around that
fundamental fact. Nothing is effective at stopping such
transients.
Again, read those cited discussions. Effective protection
was even demonstrated by Ben Franklin in 1752. It too is
discussed there. Did Franklin stop or block transients? Of
course not. Only products selling on myths attempt to get
others to "speculate" that protectors work by sitting between
the transient and its objective - earth ground destructively
via a computer.
Second - what fast acting switch? That RCD maybe takes 10
milliseconds to respond. In the meantime, 300 consecutive and
destructive transient would complete before the RCD even
thought about tripping. That fast acting switch has the speed
of molasses. Effective protection is defined in those
previously cited posts. You have much to learn there.
"J.J." wrote:
> Could a very fast-acting switch (like in an Residual Current
> Detector) be used to cut the incoming power supply quickly enough
> to halt the transient mains electricity spike before it got to be
> too large?
Andrew Gabriel
"Pyriform" <nob...@nowhere.com> writes:
> For the mains, online UPS is the proper solution, but I can't say I'm
> unduly concerned about not having one. Power provision in the UK is
> fairly reliable at the moment, although political and business
> imperatives will probably conspire to make it worse in the future.
Yes, we've had an exceptionally stable/reliable supply in the UK
for perhaps 30 years now (or well over 40 years if you ignore the
1972 miners strike). There are a number of reasons why that's
unlikely to be maintained in the future though.
--
Andrew Gabriel
Gregory L. Hansen
> Again an assumption that something will stop or block a
>destructive transient. Kilometers of air could not stop the
>transient. Do you think that silly little RCD switch contact
>will do what kilometers of sky could not? Again, protection
>is about shunting (diverting, connecting, shorting) a
>transient to earth ground. There is no way around that
>fundamental fact. Nothing is effective at stopping such
>transients.
My surge protector came with a guarantee that the company will pay to
repair or replace any equipment damaged through the powerlines when it was
plugged in to their product.
--
"Outside the camp you shall have a place set aside to be used as a
latrine. You shall keep a trowel in your equipment and with it, when you
go outside to ease nature, you shall first dig a hole and afterward cover
up your excrement." -- Deuteronomy 23:13-14
Keith
> In article <40EDCBCC...@hotmail.com>, w_tom <w_t...@hotmail.com> wrote:
>> Again an assumption that something will stop or block a
>>destructive transient. Kilometers of air could not stop the
>>transient. Do you think that silly little RCD switch contact
>>will do what kilometers of sky could not? Again, protection
>>is about shunting (diverting, connecting, shorting) a
>>transient to earth ground. There is no way around that
>>fundamental fact. Nothing is effective at stopping such
>>transients.
>
> My surge protector came with a guarantee that the company will pay to
> repair or replace any equipment damaged through the powerlines when it was
> plugged in to their product.
Have you collected on that promise? Somehow I think you'd buy an extended
warranty on that strip too. ;-)
--
Keith
Louis Bybee
news:cckv5n$enc$1...@hood.uits.indiana.edu...
> In article <40EDCBCC...@hotmail.com>, w_tom <w_t...@hotmail.com>
wrote:
> > Again an assumption that something will stop or block a
> >destructive transient. Kilometers of air could not stop the
> >transient. Do you think that silly little RCD switch contact
> >will do what kilometers of sky could not? Again, protection
> >is about shunting (diverting, connecting, shorting) a
> >transient to earth ground. There is no way around that
> >fundamental fact. Nothing is effective at stopping such
> >transients.
>
> My surge protector came with a guarantee that the company will pay to
> repair or replace any equipment damaged through the powerlines when it was
> plugged in to their product.
> --
I've been around for some time now, and I've never known or heard of anyone
actually collecting on such an offer. I know a few that tried, but gave up
after about the fifteenth hoop (getting through the first few was a cake
walk, but they kept getting higher, and higher!). :-] I've heard that
some outfits will pay claims depending on the cost of the item involved, but
know damage will occur with their devices, and consider the payout a cost of
doing business.
Read all the fine print on the offer you have, and report back if you think
you still have a prayer collecting after damaging surge event.
Louis--
*********************************************
Remove the two fish in address to respond
Andrew Gabriel
"J.J." <nom...@mail.com> writes:
> and...@cucumber.demon.co.uk (Andrew Gabriel) wrote:
>
>> Yes, we've had an exceptionally stable/reliable supply in the
>> UK for perhaps 30 years now (or well over 40 years if you
>> ignore the 1972 miners strike). There are a number of reasons
>> why that's unlikely to be maintained in the future though.
>
> stable/reliable UK mains power supply?
In a nutshell...
o We are no longer self-sufficient in energy (became a natural gas
importer last year) and are increasingly going to have to rely on
sources from the less stable parts of the planet and sources
which require traversing the less stable parts of the planet.
o All the non-natural gas sources of generation are winding down
at the end of their service lifetimes, and no more being built
(ignoring renewables, which are currently insignificant).
o We no longer have a store of energy -- we used to have many months
supply of coal stockpiled at power stations and weeks supply of
gas stored in gasometers -- all now gone.
o We now have very little in the way of spare generating capacity.
The nationalised electricity generating board used to maintain
spare capacity to enable peaks and unexpected outages to be
handled without concern, but the privatised companies mothballed
this plant as they are only paid for what they produce. It
would take between 3 months and a year to get it back in service,
depending how long it's been mothballed, so it's no use as an
energency reserve.
The industry had a wake-up call on 10 December 2002 when the country
got within a couple of minutes of having to load shed (switch off
parts of the country in an emergency due to not enough power being
able to be generated to meet demand). In spite of this, nothing was
done. Again last winter, there was a particularly cold spell forecast
and a number of experts warned we were in an even worse state than
the year before. Fortunately, the cold spell wasn't anything like as
bad as forecast. Given these wake-up calls have been ignored by the
government, it looks like it's going to have to get worse before
any notice is taken, and we probably are going to have to suffer a
significant load-shedding incident blacking out significant parts of
the country.
--
Andrew Gabriel
Andrew Gabriel
glha...@steel.ucs.indiana.edu (Gregory L. Hansen) writes:
>
> My surge protector came with a guarantee that the company will pay to
> repair or replace any equipment damaged through the powerlines when it was
> plugged in to their product.
If you do the sums, it cheaper for them to leave out the surge
suppressor components altogether and just pay up on any such incident.
Power line surges (in the UK at least) is a vanishingly insignificant
source of damage.
--
Andrew Gabriel
Richard Herring
<and...@cucumber.demon.co.uk> writes
>In article <9520A0CF6...@130.133.1.4>,
> Lem <nom...@mail.com> writes:
>
>What I have seen several times is damage caused by a surge
>induced by lightning on a phone line.
Been there, done that :-(
>Of course, the mains
>surge protector will do nothing to protect against that.
>
>> So personally I don't bother using a surge protector on my PC.
>>
>> Am I being too complacent?
>
>Not in my opinion.
>If I saw something at less than rip-off prices for protecting
>against surge on phone line, I might consider that.
>
Many of the mains surge protectors also include connectors for looping
a phone line through them.
--
Richard Herring
Richard Herring
top-posted
> Again an assumption that something will stop or block a
>destructive transient. Kilometers of air could not stop the
>transient.
I think you need to be more quantitative here. Kilometres of air could
not stop what, exactly? They certainly do a good job of protecting me
from lightning strikes in the next village.
--
Richard Herring
w_tom
charges. Electricity travels through kilometers of
non-conductive air. Why would some millimeters of air between
two RCD contacts stop same electricity? 1,000,000 mm of air
did not stop lightning. Why would 4 mm of RCD air stop what a
million mm could not?
The bottom line is this. Nothing is going to stop
lightning. Lightning protection is about shunting - also
called diverting, redirecting, or electrically connecting - to
earth ground. Any protector that claims to stop or block
lightning (such as in the RCD question) is simply promoting a
myth. And yet that is exactly what many people do - promote
the myth - when they recommend plug-in protectors.
In response to the OP's question. 4 mm of air inside the
RCD (open switch contacts) is not going to stop a potentially
destructive surge.
Richard Herring wrote:
> In message <40EDCBCC...@hotmail.com>, w_tom <w_t...@hotmail.com>
Gregory L. Hansen
Haven't had anything break that was plugged into it, although the power
has flickered or quit many times during storms. But then, the TV and VCR
are still working, and they're not on a surge protector.
--
"Let us learn to dream, gentlemen, then perhaps we shall find the
truth... But let us beware of publishing our dreams before they have been
put to the proof by the waking understanding." -- Friedrich August Kekulé
Pyriform
> The bottom line is this. Nothing is going to stop
> lightning. Lightning protection is about shunting - also
> called diverting, redirecting, or electrically connecting - to
> earth ground. Any protector that claims to stop or block
> lightning (such as in the RCD question) is simply promoting a
> myth. And yet that is exactly what many people do - promote
> the myth - when they recommend plug-in protectors.
I am not naive enough to think that my little "surge protector" would
save anything from a full-on lightning strike. The issue is whether one
might reasonably expect more modest voltage transients to occur on the
mains supply (and phone line) which can be safely absorbed by such a
device, and which might otherwise disrupt or destroy delicate
electronics (equipped, perhaps, with rather cheap PSUs).
I freely admit to not knowing the answer. Your contributions make good
points about the requirements to be met in order to achieve 'proper'
surge protection, but do nothing to address this question.
Richard Herring
writes
[top-posting corrected]
>Richard Herring wrote:
>> In message <40EDCBCC...@hotmail.com>, w_tom <w_t...@hotmail.com>
>>> Again an assumption that something will stop or block a
>>> destructive transient. Kilometers of air could not stop the
>>> transient.
>>
>> I think you need to be more quantitative here. Kilometres of
>> air could not stop what, exactly? They certainly do a good
>> job of protecting me from lightning strikes in the next
>> village.
> Lightning is a connection from cloud to earth borne
>charges. Electricity travels through kilometers of
>non-conductive air.
No, current travels through kilometres of conductive ionised air. First
you need a strong enough potential difference to produce a strong enough
electric field to cause that ionisation. Then there's a complicated
process by which the ionisation spreads to produce a complete channel.
> Why would some millimeters of air between
>two RCD contacts stop same electricity? 1,000,000 mm of air
>did not stop lightning. Why would 4 mm of RCD air stop what a
>million mm could not?
Because the potential difference across _that_ gap, and hence the field
strength, may now be insufficient to produce ionisation. Never mind how
many megavolts there were between cloud and earth, the question is how
many there are across that contact gap.
That's why I asked you to be quantitative. Qualitative language like
"same electricity", "surge" and "transient" is not helpful here. Do you
mean a current, a potential difference, an electric field, or what?
> The bottom line is this. Nothing is going to stop
>lightning. Lightning protection is about shunting - also
>called diverting, redirecting, or electrically connecting - to
>earth ground. Any protector that claims to stop or block
>lightning (such as in the RCD question) is simply promoting a
>myth. And yet that is exactly what many people do - promote
>the myth - when they recommend plug-in protectors.
Now you're confusing two different things. Most plug-in protectors that
I've encountered do not rely on open contacts (as in the RCD) but use
some form of voltage-dependent resistor and/or spark gap to shunt the
excess potential difference to local ground. To be sure, there may be so
much inductance in the system that they are ineffective, but that's a
different issue.
>
> In response to the OP's question. 4 mm of air inside the
>RCD (open switch contacts) is not going to stop a potentially
>destructive surge.
It _will_ stop a "surge" or even a "transient" up to some threshold
potential difference. What's that value, and how is it related to what's
going on in the sky outside?
--
Richard Herring
someone
"w_tom" <w_t...@hotmail.com> wrote in message
news:40EDC2A0...@hotmail.com...
A brownout is generally recognized as planned voltage cuts - or
undervoltage.
My utility in the US routinely implements 3% and 5% voltage cuts to shave or
reduce peaks (and thereby save $' when buying power).
There is also the issue of spikes and harmonics.
Given a current laptop/desktop will be ~ $800USD to ~$4000 for a high
performance system and a UPS sells for $40 - why subject a computer
to unexpected power events? What is the value of the data on the hdd(s)?
Andrew Gabriel
"someone" <som...@spamfree.com> writes:
>
> A brownout is generally recognized as planned voltage cuts - or
> undervoltage.
> My utility in the US routinely implements 3% and 5% voltage cuts to shave or
> reduce peaks (and thereby save $' when buying power).
>
> There is also the issue of spikes and harmonics.
>
> Given a current laptop/desktop will be ~ $800USD to ~$4000 for a high
> performance system and a UPS sells for $40 - why subject a computer
> to unexpected power events? What is the value of the data on the hdd(s)?
This just isn't an issue in the UK, about which the original question
was asked. Maybe it's more of a problem in the US for some reason?
--
Andrew Gabriel
Ron Reaugh
w_tom
the utility, is totally irrelevant to electronics and
especially irrelevant to computers. A computer works just
fine even when incandescent bulbs dim to less than 40%
intensity. Even demanded in Intel specifications. IOW what
the utility calls a voltage reduction is full power to the
computer. Utility would have to decrease voltage more than
20% for a computer to see a brownout. But if utility voltage
drops that low, then electric motors may be damaged. IOW
voltage too low to damage electric motors is even full power
to a computer - which demonstrates how resilient a computer
really is.
BTW, utility does not institute a voltage reduction to save
money. Voltage reductions are a last ditch effort to avoid
rolling blackouts.
Spikes and harmonics are (or should be) irrelevant to a
computer. Again, because the computer is so resilient.
However that internal computer protection assumes the building
has a 'whole house' protector so that spikes cannot overwhelm
computer internal protection.
All of which is irrelevant to HD protection. Either the
power supply will output correct power or it will shutdown.
This, of course, assumes the computer assembler had basic
electrical knowledge and did not install those 'defective by
design' $25 or $40 power supplies. But again, this was
explained earlier.
There is nothing cost effective adjacent to the computer. No
UPS nor power strip protector that will protect computer
hardware. Computer internal protection assumes the building
has implemented a 'whole house' protector on AC mains
connected less than 10 feet to central earth ground.
Protection as it was even done and well proven before WWII.
Pyriform
> BTW, utility does not institute a voltage reduction to save
> money. Voltage reductions are a last ditch effort to avoid
> rolling blackouts.
Or to avoid the expense of bringing additional generating capacity
online, thereby saving money...
> All of which is irrelevant to HD protection. Either the
> power supply will output correct power or it will shutdown.
A PSU shutting down is not irrelevant to HD protection. Shutdown at the
wrong moment (especially with the wrong operating system) and you end up
with a badly trashed filesystem.
someone
> A 3 or 5% reduction in voltage, also known as a brownout to
> the utility, is totally irrelevant to electronics and
> especially irrelevant to computers. A computer works just
> fine even when incandescent bulbs dim to less than 40%
> intensity. Even demanded in Intel specifications. IOW what
> the utility calls a voltage reduction is full power to the
> computer. Utility would have to decrease voltage more than
> 20% for a computer to see a brownout. But if utility voltage
> drops that low, then electric motors may be damaged. IOW
> voltage too low to damage electric motors is even full power
> to a computer - which demonstrates how resilient a computer
> really is.
>
> BTW, utility does not institute a voltage reduction to save
> money. Voltage reductions are a last ditch effort to avoid
> rolling blackouts.
True - my opinion as well - perhaps you could convince my local utility re
this procedure being used on a daily basis.
>
> Spikes and harmonics are (or should be) irrelevant to a
> computer. Again, because the computer is so resilient.
> However that internal computer protection assumes the building
> has a 'whole house' protector so that spikes cannot overwhelm
> computer internal protection.
>
> All of which is irrelevant to HD protection. Either the
> power supply will output correct power or it will shutdown.
> This, of course, assumes the computer assembler had basic
> electrical knowledge and did not install those 'defective by
> design' $25 or $40 power supplies. But again, this was
> explained earlier.
>
> There is nothing cost effective adjacent to the computer. No
> UPS nor power strip protector that will protect computer
> hardware. Computer internal protection assumes the building
> has implemented a 'whole house' protector on AC mains
> connected less than 10 feet to central earth ground.
> Protection as it was even done and well proven before WWII.
Perhaps you could explain the half dozen UPS I have seen that operated
correctly and interrupted close in electrical faults.
Naturally the UPS were scrap after the electrical event - but the protected
electronics were ok.
$40 UPS vs $800 desktop or in one situation $200 - $300 UPS vs. $5000 of
servers.
Eric Gisin
Once at the end of 100ft extension cord along with 1500W heater, lights
dimming.
The other time at 130V, light bulbs popping every month.
Even on a generator nearly out of gas I had a minute to shutdown my system, no
corruption occured.
"someone" <som...@spamfree.com> wrote in message
news:ZsAHc.14750$oh.13125@lakeread05...
w_tom
effective protection meaning that they are not even relevant
to the originally posted question. Furthermore, the little
surge protector does not absorb even modest transients.
Absorbing is not what they do.
Second, a full-on lightning strike is why we install
protectors. Modems already have significant internal
protection as part of design and to meet industry
requirements. Protection that can be compromised by a full-on
lightning strike.
Typical frequency of potentially destructive surges is once
every eight years. That number varies significantly even
between adjacent towns. So how frequent is your
neighborhood? Without information such as underlying geology
and manmade buried objects, weather trends, etc; then your
only valid information comes from history provided by long
term neighbors. Yes, even installation of new buried
utilities can change those trends.
'Whole house' protectors and earthing is so inexpensive that
US telco companies install same, for free, at every customer
interface. Question is whether the £1 per protected appliance
is necessary for a destructive transient that might occur once
every ten or so years. Points one and two define why you
would install that protector.
Ron Reaugh
> There is nothing cost effective adjacent to the computer. No
> UPS nor power strip protector that will protect computer
> hardware.
That's simply wrong/false. A good UPS or good surge protector WILL protect
a PC. Do you understand the concept of common mode?
> Computer internal protection assumes the building
> has implemented a 'whole house' protector on AC mains
> connected less than 10 feet to central earth ground.
That's simply wrong/false. There is nothing magic about 10 feet
except that an electrical pulse travels about that far in 10 nanoseconds.
What do you suppose the risetime of a lightening bolt is and how
does that relate to that 10 feet?
Define "central earth ground" and what it's important characteristics are.
Where is the "central earth ground" on an airplane, car, ISS or on the
Antartic pole US base(over a mile thick cold pure ice)? Are all PCs and
electronics gadgets there doomed?
Ever heard on the concept of a Farady cage?
> Protection as it was even done and well proven before WWII.
Yes, for those who actually understand it. What does the "central earth
ground"
look like at a major power station, substation or hydroelectric dam?
Where's it located?
w_tom
computer both connect to AC mains just like light bulbs. In
fact it would be same protection if both computer and UPS
shared same wall receptacle. Any transient from the
receptacle confronts UPS and computer equally. However
protection inside a UPS is often so grossly undersized that a
surge too small to damage a computer might still damage the
UPS. Furthermore, some computers can even act as surge
protectors - shunt a destructive surge so that it does not
seek earth ground via other computers.
Until you define specific circuits - including how every
wall receptacle is wired, then I cannot provide more
information.
I cannot say exactly why that particular event happened.
But above is one reason why a UPS may be damaged and computer
is not. Computer power supplies have internal protection.
Protection so sufficient that there is little adjacent to a
power supply that can enhance protection. But computer
internal protection can be overwhelmed if destructive
transients are not earthed before entering the building.
Bottom line is this. You had UPS failure. Therefore you
have no effective surge protection. Even surge protectors
must not be damaged due to a surge.
To provide a better answer, do as I do - autopsy the dead
body. Replace the defective part to learn what has actually
been damaged. Autopsy only complete when the failed unit is
fully functional.
If a server farm has no 'whole house' protection and a single
point earth ground, then no UPS or plug-in protector is going
to do anything better. In fact, it is just not a reliable
operation if 1) every incoming utility line does not enter at
the common service entrance all connected to the single point
earth ground and 2) building does not have necessary 'whole
house' protector on incoming AC mains. From Sun Microsystems
planning guide:
http://www.sun.com/servers/white-papers/dc-planning-guide.pdf
> Lightning surges cannot be stopped, but they can be diverted.
> The plans for the data center should be thoroughly reviewed
> to identify any paths for surge entry into the data center.
> Surge arrestors can be designed into the system to help
> mitigate the potential for lightning damage within the data
> center. These should divert the power of the surge by
> providing a path to ground for the surge energy. Protection
> should be placed on both the primary and secondary side of
> the service transformer. It is also necessary to protect
> against surges through the communications lines. The
> specific design of the lightning protection system for the
> data center will be dependent on the design of the building
> and utilities and existing protection measures.
If you are suffering transient damage, then the human is
reason for failure. What Sun writes is so well proven and
understood that it was standard even before WWII. Protection
is only as effective as its earth ground.
As for your brownouts - if any voltage is too low for a
computer, then the utility has grossly violated national
standards. A PUC call would create a massive response - if
your AC voltage drops so low as to be problematic to a
computer.
w_tom
simplistic file systems. It is not a problem to superior
(journalizing) filesystems.
Will a disk drive write to the platter as voltage drops? Of
course not. The disk drive controller is a complete computer
that also monitors voltage. It does not matter to disk
hardware when power is turned off. But it does matter to some
'simplistic' disk filesystems that power is not removed during
a write operation.
Just another reason why FAT was obsoleted by HPFS which in
turn was obsoleted by NTFS.
w_tom
protection. Also cite the manufactuer's spec that claims that
common mode protection (and good luck).
Also please explain how those 130 ohms impedance in 50' of
12 AWG wire is not significant when earthing even a trivial
100 amp surge?
In the meantime, please explain how earth ground at a hydro
electric plant is at all related to single point earth ground
for a building, for a PA or stereo system, for the PC board
layout of A/D converters, or any other simple electronic
system where ground loops can be a problem. You do understand
the concept of ground loop? Good. Please then show us how
the ground at a power station has any relevance?
Now for the completely irrelevant topic of ground in an
airplane:
Pyriform
> Furthermore, the little
> surge protector does not absorb even modest transients.
> Absorbing is not what they do.
I say absorb; you say shunt. We mean the same thing. Energy that would
have entered the 'protected' load instead goes somewhere else. So unless
you are arguing purely on the basis of semantics, your claim that even
"modest transients" are not absorbed by plug-in surge suppressors is
clearly false. What I want to know is whether such transients are
actually found, and whether they pose a threat to 'unprotected'
equipment.
Your "unless it protects against everything, it protects against
nothing" argument is not entirely convincing.
w_tom
Now for that airplane. Notice earth ground is the tail
section:
http://bm6aak.myweb.hinet.net/file/456.gif
Of course this airplane ground is completely beyond the
scope of the current discussion. Airplanes are more difficult
to ground. A Pan Am 707 was destroyed by lightning over
Elkland MD because internal grounding was not sufficient. An
airplane must be grounded so that any part can become an earth
ground; making airplane design more challenging. We, on the
other hand, are having enough trouble discussing simple
structural earthing - a well proven 1930 technology. Why then
complicate it with airplanes and other irerelevant questions?
w_tom wrote:
> ...
> Now for the completely irrelevant topic of ground in an
> airplane:
>
> Ron Reaugh wrote:
>> "w_tom" <w_t...@hotmail.com> wrote in message
>> news:40EF0C7C...@hotmail.com...
>>> ...
>>> Computer internal protection assumes the building
>>> has implemented a 'whole house' protector on AC mains
>>> connected less than 10 feet to central earth ground.
>>
>> That's simply wrong/false. There is nothing magic about 10
>> feet except that an electrical pulse travels about that far
>> in 10 nanoseconds. What do you suppose the risetime of a
>> lightening bolt is and how does that relate to that 10
>> feet? Define "central earth ground" and what it's
>> important characteristics are. Where is the "central earth
>> ground" on an airplane, car, ISS or on the Antartic pole
>> US base(over a mile thick cold pure ice)? Are all PCs and
>> electronics gadgets there doomed? Ever heard on the concept
>> of a Farady cage?
>> ...
Ron Reaugh
> Ok sir. Explain to us how a plug-in UPS provides common mode
> protection. Also cite the manufactuer's spec that claims that
> common mode protection (and good luck).
No problem. Here:
http://sturgeon.apcc.com/techref.nsf/partnum/990-7015/$FILE/7015-1.pdf
Section 9.4
Do you understand what common mode means with respect to the above spec and
do you understand how that differs/same as the general concept of common
mode and how that relates to these issues?
In order to protect a device from an undesirable voltage arriving over the
power cable one simply shunts that unwanted voltage such that it appears
equally on all the wires on that power cable i.e. AC-hot, AC-neutral and
the ground wire which connects to the chassis of the device. That shunt is
done with capacitors and surge diodes or MOV devices etc. or in the old days
on your phone line with a spark gap. That's basically what a surge
suppressor does. The device's input components therefore see no intolerable
VOLTAGES and it survives. It makes no difference to the device if the whole
device(chassis and all) jumps to a million volts during the episode. Ever
heard of a Faraday Cage?
> Also please explain how those 130 ohms impedance in 50' of
> 12 AWG wire is not significant when earthing even a trivial
> 100 amp surge?
Not relevant. What happens on an airplane? The protection issue isn't
grounding a surge current; the issue is maintaining important components at
a stable/safe voltage with respect to one another such that nothing gets
damaged. Ever heard of a Faraday Cage?
> In the meantime, please explain how earth ground at a hydro
> electric plant is at all related to single point earth ground
> for a building,
In large buildings as in hydroelectric plants there is NO SINGLE POINT
ground but multiple connections to a common ground cage/plane. On an
airplane there isn't even one point to ground except an ionized air column t
hat chooses its own path for a few microseconds at a time on occasion.
Ever heard of the concept of a ground plane? Every heard on the concept of
capacitive coupling and AC impedance? How does a Faraday Cage relate to a
ground plane?
> for a PA or stereo system, for the PC board
> layout of A/D converters, or any other simple electronic
> system where ground loops can be a problem. You do understand
> the concept of ground loop?
You're walkin into my backyard now. Now what's the difference in design of
that DA for 8 bit 800MHz conversion and 24 bit 200KHz conversion? DAs are
more fun that ADs.
> Good. Please then show us how
> the ground at a power station has any relevance?
>
> Now for the completely irrelevant topic of ground in an
> airplane:
It's entirely relevant to demonstrating that you have no understanding of
the issues in question.
The answer is that the principles of how to do system input protection on an
airplane are IDENTICAL to how to do them at home or in a high rise.
The NUMBER ONE FAILURE of an incompetent designer in this arena is to
mistake that a ground connection has much to do with the issue; it does
NOT. The ground connection has much more to do with other issues like the
safety of the guy who is using the box and the UL and the national
electrical code. Why do you suppose that US home wiring didn't even include
a 3rd wire(ground) until the 50's? Do you think some new physical law was
suddenly discovered?
One can protect a gadget from surges WITHOUT a ground wire or any ground at
all; that's the WHOLE point. Shunt the surge voltage such that it's common
moded and to the chassis at the input and outputs and the device is
protected.
Ron Reaugh
> Transmitted an incomplete post. Sorry for the mistake.
> Now for that airplane. Notice earth ground is the tail
> section:
> http://bm6aak.myweb.hinet.net/file/456.gif
OH, cool but of course it simply proves my points.
> Of course this airplane ground is completely beyond the
> scope of the current discussion.
You mean beyond you.
> Airplanes are more difficult
> to ground.
No, airplanes are impossible to proactively ground while in flight(save a
high energy beam) and much more importantly the is no need to ground an
airplane in flight. There is just a need to have a good continuous Faraday
Cage. Damn, how did anyone ever survive when planes were made of wood or
paper(or are they non-conductors)?
> A Pan Am 707 was destroyed by lightning over
> Elkland MD because internal grounding was not sufficient.
Oh, you mean the Faraday Cage was discontinuous or flawed or maybe it was a
super bolt of the kind that has punched holes in heavy gauge steel petroleum
tanks.
In any case I'll bet that the cockpit radio was undamaged at least until
impact.
> An
> airplane must be grounded so that any part can become an earth
> ground; making airplane design more challenging.
HUH?
> We, on the
> other hand, are having enough trouble discussing simple
> structural earthing - a well proven 1930 technology. Why then
> complicate it with airplanes and other irerelevant questions?
Because device/PC protection design has little to do with earth grounding.
Ron Reaugh
> Was UPS between AC mains and computer? No.
Wrong. Topologically the UPS is "between".
> UPS and
> computer both connect to AC mains just like light bulbs. In
> fact it would be same protection if both computer and UPS
> shared same wall receptacle.
WRONG! About a critical 10 nanoseconds WRONG nevermind the impedances and
common mode condiderations!
> Any transient from the
> receptacle confronts UPS and computer equally.
Not in the critical time domain.
> However
> protection inside a UPS is often so grossly undersized that a
> surge too small to damage a computer might still damage the
> UPS. Furthermore, some computers can even act as surge
> protectors - shunt a destructive surge so that it does not
> seek earth ground via other computers.
>
> Until you define specific circuits - including how every
> wall receptacle is wired, then I cannot provide more
> information.
>
> I cannot say exactly why that particular event happened.
> But above is one reason why a UPS may be damaged and computer
> is not.
NO, the first component with surge suppression topologically is usually the
one that takes the HIT. Do you suppose that's by design?
> Computer power supplies have internal protection.
> Protection so sufficient that there is little adjacent to a
> power supply that can enhance protection. But computer
> internal protection can be overwhelmed if destructive
> transients are not earthed before entering the building.
OH, you mean unless the building is a heavily constructed Faraday cage and
all wiring has feedthru bypass and surge suppression, then a destructive
transient could get through and that has NOTHING to do with you high
transient impedance ground wire.. Why is it that we all knew that?
> Bottom line is this. You had UPS failure. Therefore you
> have no effective surge protection. Even surge protectors
> must not be damaged due to a surge.
HUH, frequently good surge protectors are destroyed by big surges just as
they are designed to do. The good one FAIL closed circuit where protection
is even better!
> To provide a better answer, do as I do - autopsy the dead
> body. Replace the defective part to learn what has actually
> been damaged. Autopsy only complete when the failed unit is
> fully functional.
>
> If a server farm has no 'whole house' protection and a single
> point earth ground,
NO, a large server farm has grid ground and power grid firewalls.
> then no UPS or plug-in protector is going
> to do anything better. In fact, it is just not a reliable
> operation if 1) every incoming utility line does not enter at
> the common service entrance all connected to the single point
> earth ground
No, you are getting closer to reality.
> and 2) building does not have necessary 'whole
> house' protector on incoming AC mains. From Sun Microsystems
> planning guide:
> http://www.sun.com/servers/white-papers/dc-planning-guide.pdf
> > Lightning surges cannot be stopped, but they can be diverted.
To a good Faraday cage or ground plane. A circuit of LOW AC IMPEDANCE.
> > The plans for the data center should be thoroughly reviewed
> > to identify any paths for surge entry into the data center.
> > Surge arrestors can be designed into the system to help
> > mitigate the potential for lightning damage within the data
> > center. These should divert the power of the surge by
> > providing a path to ground for the surge energy. Protection
> > should be placed on both the primary and secondary side of
> > the service transformer. It is also necessary to protect
> > against surges through the communications lines. The
> > specific design of the lightning protection system for the
> > data center will be dependent on the design of the building
> > and utilities and existing protection measures.
What that all boils down to is design the Faraday Cage or ground plane well
and has little to do with actual earth grounding save the UL and electrical
code.
Chaos Master
> Does anyone have a reference to HDDs getting corrupted by power
> events on the mains power supply?
Happened here. Power surge, while I was working on a MS-Word document.
Returned, Windows booted, but crashed at the desktop, so I entered using a Win98
boot disk and ran Scandisk.
2 or 3 bad blocks, and some files/Windows' registry were corrupted.
[]s
--
© Chaos Master. |"These wounds won't seem to heal
Posting from Brazil. | This pain is just too real
http://marreka.no-ip.com | There's just too much that time can't erase"
(most often offline... ) | -- Evanescence, "My Immortal"
w_tom
> ... good surge protectors are destroyed by big surges just
> as they are designed to do. The good one FAIL closed
> circuit where protection is even better!
That claim is classic urban myth. MOV data sheets define
normal operation. MOV is at end of life typically when it
degrades by about 5%. How can it degrade 5% and yet
vaporize? It cannot. Bottom line remains - a properly sized
protector shunts the transient and remains fully operational.
Eventually MOV degrades; does not vaporize. Vaporizing is
when the MOV grossly exceeds manufacturer specification - is
grossly undersized for the task. But purveyors of undersized
and ineffective protectors want consumers to believe their
overpriced protector should vaporize on every surge. Scam is
the better word.
In the meantime, there is no topology in electronic
circuits. Electrically, a shunt mode protector is not
"between" the appliance and a surge no matter how much junk
science topology is rationalized. But then this thread is
full of myth purveyors promoting such junk science reasoning -
such as MOVs are designed to protect by vaporizing. Which
plug-in manufacturer do you work for, Ron Reaugh?
>> ...
>> From Sun Microsystems planning guide:
>> http://www.sun.com/servers/white-papers/dc-planning-guide.pdf
>>> Lightning surges cannot be stopped, but they can be diverted.
w_tom
of aluminum. And inside that aluminum 'faraday cage' was a
completely enclosed fuel tank - another faraday cage. How did
lightning get through two faraday cages to explode the fuel
tank on that Boeing 707? And why was the correction to
install more grounds inside that 'faraday cage'? Could it be
that no sufficient 'faraday cage' exists? Yep.
Ron, when you get some real world experience with 'faraday
cages', then come back and share your experiences. In the
meantime, grounding inside that 'faraday cage' is essential
for safe airline operations - so that lightning will pass
through the inside of that 'faraday cage' without doing
damage. Lightning caused damage inside that 'faraday cage'
over Elkton MD.
Ron Reaugh wrote:
> "w_tom" <w_t...@hotmail.com> wrote in message
w_tom
is the reference to common mode protection? Where are any
numbers that apply to common mode protection? Once APC used
to provide far more numbers when they claimed to protect from
Normal mode transients. Now they don't even make those
claims. Numbers provided by Ron's citation:
> Surge energy rating (one time, 10/1000 盜
> waveform): 160 J.
> Surge current capability (one time, 8/20 盜
> waveform): 6500 Amp peak maximum.
Hell. They don't even list dBs for the noise filter. What
kind of spec is that? Noise filter for what? Incoming AC
line? Output power line? Clearly these are specs for the
technically naive.
After one surge, the entire UPS is toast? Look at those
pathetic numbers. Only 160 joules? Only 6500 amps?
Effective protection starts at about 1000 joules and 50,000
amps. Thank you Ron for demonstrating pathetic protection
from that plug-in UPS.
Oh - where do they mention anything about 'faraday cage'
protection?
In the meantime, Ron describes normal mode protection:
> In order to protect a device from an undesirable voltage
> arriving over the power cable one simply shunts that unwanted
> voltage such that it appears equally on all the wires on that
> power cable ...
Where is the common mode protection? Not in that citation.
Not in what Ron describes. Just another reason why that
plug-in UPS does not provide effective protection.
Ron Reaugh wrote:
> "w_tom" <w_t...@hotmail.com> wrote in message
> news:40EF2C44...@hotmail.com...
>> Ok sir. Explain to us how a plug-in UPS provides common mode
>> protection. Also cite the manufactuer's spec that claims that
>> common mode protection (and good luck).
>
> No problem. Here:
> http://sturgeon.apcc.com/techref.nsf/partnum/990-7015/$FILE/7015-1.pdf
> Section 9.4
>
> Do you understand what common mode means with respect to the
> above spec and do you understand how that differs/same as
> the general concept of common mode and how that relates to
> these issues?
>
> In order to protect a device from an undesirable voltage
> arriving over the power cable one simply shunts that unwanted
> voltage such that it appears equally on all the wires on that
> power cable i.e. AC-hot, AC-neutral and the ground wire
> which connects to the chassis of the device. That shunt is
> done with capacitors and surge diodes or MOV devices etc. or
> in the old days on your phone line with a spark gap. That's
> basically what a surge suppressor does. The device's input
> components therefore see no intolerable VOLTAGES and it
> survives. It makes no difference to the device if the whole
> device(chassis and all) jumps to a million volts during the
> episode. Ever heard of a Faraday Cage?
> ...
Ron Reaugh
> Please feel free to show us an MOV datasheet that says
> > ... good surge protectors are destroyed by big surges just
> > as they are designed to do. The good one FAIL closed
> > circuit where protection is even better!
>
> That claim is classic urban myth. MOV data sheets define
> normal operation. MOV is at end of life typically when it
> degrades by about 5%. How can it degrade 5% and yet
> vaporize? It cannot. Bottom line remains - a properly sized
> protector shunts the transient and remains fully operational.
> Eventually MOV degrades; does not vaporize. Vaporizing is
> when the MOV grossly exceeds manufacturer specification - is
> grossly undersized for the task. But purveyors of undersized
> and ineffective protectors want consumers to believe their
> overpriced protector should vaporize on every surge. Scam is
> the better word.
Gibber ignored.
> In the meantime, there is no topology in electronic
> circuits.
And this wacko nonsense is from the guy who brought up DA design! At high
frequency it is most ALL about topology!
> Electrically, a shunt mode protector is not
> "between" the appliance and a surge no matter how much junk
> science topology is rationalized.
Did you ever hear about the speed of light or about 1 foot per nanosecond?
> But then this thread is
> full of myth purveyors promoting such junk science reasoning -
> such as MOVs are designed to protect by vaporizing. Which
> plug-in manufacturer do you work for, Ron Reaugh?
Vaporizing...are you gonna bring in Klingons now as we seem to be having a
bit to drink?
Ron Reaugh
> Ron. Did you read your citation before posting it? Where
> is the reference to common mode protection?
Right where I said it was as anyone who reads it can see for themselves.
Gain some technical background before you tackle such technical issues.
Ron Reaugh
> Well lets see. The Boeing 707 was an entire 'faraday cage'
> of aluminum. And inside that aluminum 'faraday cage' was a
> completely enclosed fuel tank - another faraday cage. How did
> lightning get through two faraday cages to explode the fuel
> tank on that Boeing 707? And why was the correction to
> install more grounds inside that 'faraday cage'? Could it be
> that no sufficient 'faraday cage' exists? Yep.
If you'd read my previous post I already said exactly that.
>
> Ron, when you get some real world experience with 'faraday
> cages', then come back and share your experiences. In the
> meantime, grounding inside that 'faraday cage' is essential
> for safe airline operations - so that lightning will pass
> through the inside of that 'faraday cage' without doing
> damage. Lightning caused damage inside that 'faraday cage'
> over Elkton MD.
No, inside a discontinuous Faraday cage....get a clue.
w_tom
from a type of surge that does not typically exist. That way,
the myth purveyors will assume it protects from all types of
surges. And so we have a common quote - "a surge got past my
surge protector". In reality, the surge took a left turn to
get the surge protector and a right turn to attack the
computer - simultaneously.
Get a surge protector that protects from a type of surge
that actually does damage. So which one do you install? The
one that costs tens of times more money per protected
appliance, OR the one that costs so much less and even
protects from the destructive type of surge? The latter is
the single, properly sized, and properly earthed 'whole house'
protector. Now that the best protector is also the most cost
effective, we can compare that price to what we might lose
without installing it. IOW the 'whole house' protector
provides a basis to decide protection for everything inside
the house.
Is a PC surge protector needed in the UK. No: if protectors
is the ineffective plug-in type that may even contribute to
damage of an adjacent computer. Maybe: if it is the less
expensive and more effective 'whole house' type.
w_tom
protector absorbs surges, then it sits between surge and the
protected device - in series. It must stop or block surges.
But then it must be a series mode protector. It is not. It
is shunt mode. Shunt mode protectors don't function if they
absorb surge energy. Difference between 'shunt' and 'absorb'
is important to keep the consumer confused.
A shunt mode protector connects off to the side; is not
located electrically between protected device and incoming
transient. A shunt mode protector connects as if it was
another appliance - albeit much closer to the single point
ground.
Absorbing is what the plug-in protector manufacturer hopes
you assume. That way they need not discuss earthing and hope
you assume it is a series mode protector. But they are,
instead, shunt mode protectors. Effective shunt mode
protectors must be connected short to earth ground. If you
*assume* it absorbs surges then they can avoid an earthing
discussion; let myths purveyors promote their ineffective
product.
Series mode protectors absorb. Shunt mode protectors are
similar to electric switches or electric wires - they shunt.
If others believe that it absorbs, then critical earth ground
may be overlooked. Essential to selling that ineffective
protector is to avoid all mention of earthing. And so they
hope other will *assume* it absorbs. If it shunts, then one
may ask what it shunts to? Those would be embarrassing
questions.
w_tom
impulse voltage first be seen? On the near end where the
impulse starts? Obviously not. Impulse first appears at the
far end of wire. How can this be when it takes the impulse 1
foot per second to get to that far end? Topology as defined
by Ron would erroneously conclude impulse first appears where
impulse was first applied to wire.
Why do we know that topology not relevant? Its a simple
second year course called E-M fields. One first learned basic
concepts before promoting rubbish such as topology and
'faraday cage'.
Notice that Ron Reugh also cannot provide MOV datasheets to
demonstrate protection by vaporization. He is typical of
those who would recommend plug-in protectors. Facts remain
unchallenged: a surge protector that vaporizes during a surge
is ineffective and even violates the MOV manufacturer's own
specifications. Ron's best technical response:
> Gibber ignored.
Ron Reaugh wrote:
> "w_tom" <w_t...@hotmail.com> wrote in message
> ...
w_tom
mode transient protection. Even worse, they no longer make the
specific claim for normal mode protection. Ron again
demonstrates propaganda used to promote plug-in UPSes for
ineffective surge protection.
John Gilmer
>
> Just another reason why FAT was obsoleted by HPFS which in
> turn was obsoleted by NTFS.
I'l bite:
I think I understant FAT systems.
But the only thing I "think" I know about the NTFS is that, effectively, the
system first makes a record of what it is about to do, then it does it, and
then it either erases the original record or somehow marks it.
SO: can someone "explain" the NTFS to me. (Please don't tell me to "look
it up.")
The Ghost In The Machine
<w_t...@hotmail.com>
wrote
on Sat, 10 Jul 2004 02:06:11 -0400
<40EF8752...@hotmail.com>:
> Put an impulse down a wire. Where on that wire will the
> impulse voltage first be seen? On the near end where the
> impulse starts? Obviously not. Impulse first appears at the
> far end of wire. How can this be when it takes the impulse 1
> foot per second to get to that far end?
Erm, 1 foot per *nano*second. Light speed.
Perhaps it's somewhat less than that; AIUI electric current
is along the lines of 2/3 c, but it's still pretty darned
fast.
[rest snipped]
--
#191, ewi...@earthlink.net
It's still legal to go .sigless.
Ron Reaugh
> Put an impulse down a wire.
You can't even define "impulse".
> Where on that wire will the
> impulse voltage first be seen? On the near end where the
> impulse starts? Obviously not. Impulse first appears at the
> far end of wire.
Now you've self contradicted and imploded. Ask Albert.
Ron Reaugh
> IOW there is no claim in that APC sales brochure for common
> mode transient protection. Even worse, they no longer make the
> specific claim for normal mode protection. Ron again
> demonstrates propaganda used to promote plug-in UPSes for
> ineffective surge protection.
Your wacko claims have already been refuted by my citation earlier in this
thread.
APC does include "common mode" as I cited.
Mike Tomlinson
Ron Reaugh <ron-r...@worldnet.att.net> writes
>Vaporizing...are you gonna bring in Klingons now as we seem to be having a
>bit to drink?
I suggest you consult the thread with the same title crossposted to the
following groups before wasting any more time on w_tom:
uk.comp.vendors,uk.comp.homebuilt,alt.comp.hardware,alt.comp.hardware.pc
-homebuilt
--
A. Top posters.
Q. What's the most annoying thing on Usenet?
No Spam
> Hard drives can be corrupted for various reasons based also
> upon what the filesystem is. For example, if using FATxx
> filesystems, then a loss of electrical power at the right time
> can even erase files from that drive. Just another reason why
> the technically informed want NTFS filesystems on drives; not
> FAT.
>
Or RieserFS. But that is used by OS's other than Gate$
Lemingware! :-)
no....@hotmail.com
The Ghost In The Machine
<gil...@crosslink.net>
wrote
on Sat, 10 Jul 2004 10:53:27 -0400
<40f007dc$0$12...@dingus.crosslink.net>:
NTFS stands for NT File System, presumably. (I've no idea what NT
stands for. Certain jokesters have their own opinions, mine among
them.)
A file system is a method by which the unorganized data in
a disk partition -- basically, a very long chain of blocks,
or perhaps a mapping from an integer (the logical block
address) to a fixed-size chunk of data (the block) -- can
be organized into something more appetizing to humans:
files, directories, symbolic links, or in Microsoft
parlance (perhaps), documents, folders, and shortcuts.
DOS 1.0's FAT filesystem didn't even have directories
(that was added in 1.1 or 2.0; I forget which). NTFS is
fairly sophisticated; it has, among other things:
- per-file locking (to the intense annoyance of UNIX and Linux
programmers, this appears to be on by default)
- resource streams a la Macintosh (which aren't apparently used yet?)
- Access Control Lists
- Unicode support
- Case-preserving filenames
- a master file table, which is where the small files live
- sparse files (files with "holes" in their blocklists -- a
useful capability in some contexts related to databases, AIUI)
- short file name capability for DOS backwards compatibility
- hidden files
- support for running a defragmenter while the volume is mounted.
(Don't ask.)
There are a few other capabilities but I'd have to look.
If you really do want to look it up, you can try
the Linux source code -- an NTFS implementation
is in the kernel under /usr/src/linux/fs/ntfs or
/usr/src/linux/Documentation/filesystems/ntfs.txt. It is
definitely not for the faint of heart. There should
be some documentation somewhere on Microsoft's website,
of course; again, I'd have to look.
HTH
JM
>w_tom wrote:
>> Hard drives can be corrupted for various reasons based also
>> upon what the filesystem is. For example, if using FATxx
>> filesystems, then a loss of electrical power at the right time
>> can even erase files from that drive. Just another reason why
>> the technically informed want NTFS filesystems on drives; not
>> FAT.
The files aren't actually gone forever. If the power fails during the
writing of the file, but the FAT have not been updated yet, the data will be
found as "lost clusters" by Scandisk. It'll probably be in a bunch of
pieces though, due to fragmentation.
There are little backup programs that back up the disk's FAT's, boot sector,
etc. in the event that that any of the disk's reserved sectors get
obliterated by some other means.
John Gilmer
>
> NTFS stands for NT File System, presumably. (I've no idea what NT
> stands for. Certain jokesters have their own opinions, mine among
> them.)
Yeah, yeah.
And NT stands for New Technology. It was written as a "Windows Like"
operating system to run on hardware stuff by Sun Micro and the old DEC which
used UNIX. At some point Micro$oft is to make the NT and Windows
essentially the same operating system.
>
> A file system is a method by which the unorganized data in
> a disk partition -- basically, a very long chain of blocks,
> or perhaps a mapping from an integer (the logical block
> address) to a fixed-size chunk of data (the block) -- can
> be organized into something more appetizing to humans:
> files, directories, symbolic links, or in Microsoft
> parlance (perhaps), documents, folders, and shortcuts.
> DOS 1.0's FAT filesystem didn't even have directories
> (that was added in 1.1 or 2.0; I forget which).
Well, you lost me again. The directory is supposed to point to the entry
in the FAT corresponding to the first "allocation unit" of the file. From
my old memoery, the Intel development system just had a fixed directory.
Directory entry #1 was the first allocation unit, etc. Longer files were
accomodated by "chaining" directories.
> NTFS is
> fairly sophisticated; it has, among other things:
>
> - per-file locking (to the intense annoyance of UNIX and Linux
> programmers, this appears to be on by default)
Well, WTF does it "lock?'
> - resource streams a la Macintosh (which aren't apparently used yet?)
That doesn't help.
> - Access Control Lists
OK.
> - Unicode support
Which means ...
> - Case-preserving filenames
OK
> - a master file table, which is where the small files live
Huh?
> - sparse files (files with "holes" in their blocklists -- a
> useful capability in some contexts related to databases, AIUI)
Neat!
> - short file name capability for DOS backwards compatibility
OK
> - hidden files
Old stuff.
> - support for running a defragmenter while the volume is mounted.
> (Don't ask.)
Well, I understand was the defragmenter does in a FAT system but since I
still don't understand how files are stored I can't understand how that are
either fragmented or defragmented.
>
> There are a few other capabilities but I'd have to look.
OK
>
> If you really do want to look it up, you can try
> the Linux source code -- an NTFS implementation
> is in the kernel under /usr/src/linux/fs/ntfs or
> /usr/src/linux/Documentation/filesystems/ntfs.txt. It is
> definitely not for the faint of heart. There should
> be some documentation somewhere on Microsoft's website,
> of course; again, I'd have to look.
Sorry, you have just asked me to think and work harder than I care to.
eh...@bellatlantic.net
> Ron. Did you read your citation before posting it? Where
>is the reference to common mode protection? Where are any
>numbers that apply to common mode protection?
>
>
<snip>
> Where is the common mode protection? Not in that citation.
>Not in what Ron describes.
>
<snip>
Yes, it IS in that citation, exactly where he said it is.
It is the line below the two you quoted from the page
he cited.
It says: "Surge response time: 0 ns (instantaneous)
normal mode, < 5ns common mode."
jmfb...@aol.com
"John Gilmer" <gil...@crosslink.net> wrote:
<snip-valiantly refraining from comment>
>> - support for running a defragmenter while the volume is mounted.
>> (Don't ask.)
>
>Well, I understand was the defragmenter does in a FAT system but since I
>still don't understand how files are stored I can't
>understand how that are
>either fragmented or defragmented.
Consider a file system that writes empty blocks in numerical sequential
order. Now think of a file that's deleted. This leaves an empty
"hole" in the filled blocks. Now make a file whose size is less
than the "hole". Now you have a smaller hole that will be filled
with the next file that is written. That file is larger than the
hole so the hole gets filled, then the next block that isn't filled
is found and written into. Over time, all files, when viewed from
the geometry of the physical disk look like swiss cheese.
A defragmenter takes the whole file system and rewrites each file
such that all its block numbers are monotonically increasing.
Now, where this gets really, really fucked up is when the defragger
program "forgets" which should be the next block (real easy to do
with off-by-one bugs) or has to call its error handling when it
can't do a fit or the block chain pointers become broken. The
last one is a feature of all Misoft OSes because of memory
management problems--but that's another nightmare in the not-an-OS
biz.
<snip>
/BAH
Subtract a hundred and four for e-mail.
Leonard Caillouet
"Mike Tomlinson" <mi...@NOSPAM.jasper.org.uk> wrote in message
news:1hZuM5I+...@jasper.org.uk...
> Q. What's the most annoying thing on Usenet?
A. Cross posting. Second only to people who bitch about top posters .
Leonard
The Ghost In The Machine
<jmfb...@aol.com>
wrote
on Sun, 11 Jul 04 09:29:22 GMT
<40f11921$0$1167$61fe...@news.rcn.com>:
Indeed. In Linux, there's no defragger [*], because the file
code in Linux is a little smarter. I'd admittedly have
to look for the details though, and ext2's organization
is quite different from FAT's or NTFS. FAT in particular
is terrible, basically every file is a single chain --
but you probably knew that already. NTFS is more or less
as I've described it in my prior post, at a high level,
and it feels like an engineered solution, whereas Linux's
ext2 is more elegant, even if it's still engineered.
But there's no perfect solution anyway; as you've described
the problem, there's always going to be a hole or two,
and a determined program can probably fragment any file
system if it does something like the following:
open big file
write block to big file
open little file
write block to little file
close it
write block to big file
open little file
write block to little file
close it
write block to big file
...
(It's a good thing the trend is towards centralized dedicated-machine
syslog-type logging. :-) )
I'll admit to wondering whether NT had the rather interesting
capability or not of "let's just write it here". I base
this hypothesis on observations using DiskKeeper Lite, which
copy I had at the time on a machine at my then-employer.
Basically, the notion is to simply write the new block at
an open sector in the cylinder over which the head is
flying.
Of course this would fragment things terribly, and I have no proof.
But things did fragment pretty badly when I used such tools
as Visual C++.
>
>
> <snip>
>
> /BAH
>
> Subtract a hundred and four for e-mail.
[*] actually, there is, but it's very rarely used.
w_tom
it conduct 1 microamp in 5 ns to the safety ground? What kind
of protection is that? Based upon facts and numbers provided,
then my digital multimeter is even a better surge protector -
a claim I can make because specs are better called an
'executive summary'.
At least that manufacturer once provided insufficient specs
for Normal mode protection that the manufacturer does claim to
provide:
> Normal mode surge voltage let through <5% of test peak
> voltage when subjected to IEEE 587 Cat. A 6kVA test
Now manufacturer cannot bother to provide even that
insufficient information. After all, they are not trying to
sell a 'common mode' claim to the informed. They dumb down
the numbers into rubbish so that one who wishes MOVs absorb
the energy of a surge will see what he wants.
My car tires have a common mode response time AND that
proves those tires are effective protection? Common mode
what? Does not matter. That tiny phrse would be enough even
for a poet to believe what he wants to know.
How much common mode current in less than 5 ns? From what
or which one wire is that common mode response? Is that
common mode response really just a response inside the UPS
controller circuit? Or is that a common mode response on the
serial port. RS-232 is a common mode communication ports. So
does the serial port haves a less than 5 ns response? Wow.
That means the UPS must provide massive lightning protection -
if living in the world of Harry Potter.
IOW they mention 'common mode response' but give not one
indication that the UPS provides common mode surge
protection. It only does something - and they don't even say
what or how much. That woefully insufficient and deceptive
information is enough for some to loudly declare that a UPS
provides lightning protection. IOW another urban myth has
been promoted.
There are no claims of common mode transient protection on
the incoming AC input. Provided are words without relevance so
that a poet might hope for common mode response to something -
which therefore must be a direct lightning strike? It's
called wild speculation on your part - the same person who
foolishly believes shunt mode devices (such as wire) are
designed to absorb energy. But an engineer says, "What is
this crap. There is no numerical information to work with."
That UPS does not claim common mode protection. It simply
claims some undefined of response to common mode noise from or
to an undefined location. It does not even say those 160
joules are involved in such protection. Furthermore it admits
to being grossly undersized - only 160 joules. A poet then
can assume the response time means the UPS will conduct 50,000
amps? A poet can. So can Harry Potter. Those who must deal
in reality cannot.
There is nothing in those specs beyond gobbledygook. Using
ehsjr and Ron Reaugh reasoning, should we assume the UPS is
sufficient even for aeronautical environments? After all,
they do
claim 'something' that myth purveyors can distort into a real
world miracle.
ehsjr - when will you claim that a faraday cage also makes
that UPS so effective?
I have this 741 op amp (a semiconductor amplifier). It also
has a common mode rating. So that operational amplifier (that
little IC) is also a lightning protector? Yes according to
how ehsjr reasons. Give me a break. That UPS does not even
claim to provide common mode protection - which is why they
must all but encrypt their specifications. Its called name
dropping. They dropped the phrase "common mode". That
without any numbers is enough for ehsjr to loudly claim the
UPS provides common mode protection. It is called Junk
Science reasoning.
Ron Reaugh
> Common mode to what?
Nice try liar.
w_tom
purchases power supplies without consulting specifications.
Intel specs for ATX power supplies demand that PSU not damage
motherboard and other components:
> Section 3.5.1 Overvoltage Protection
> The overvoltage sense circuitry and reference shall reside
> in packages that are separate and distinct from the
> regulator control circuitry and reference. No single
> point fault shall be able to cause a sustained overvoltage
> condition on any or all outputs. The supply shall provide
> latch-mode overvoltage protection as defined below.
> Table 11: Overvoltage Protection
> Output
> +12 VDC Max is 15.6 volts
> +5 VDC Nominal is 6.3 volts
> +3.3 VDC Nominal is 4.2 volts
Too many computer assemblers don't have necessary technical
knowledge and therefore don't even know that overvoltage
protection (OVP) has been a defacto standard for 30+ years.
That motherboard damage probably may be traceable to the
ill-informed computer assembler (who does not demand specs) or
a power supply manufacturer who outrightly lies on his
specifications.
There is nothing in a UPS that will accomplish the necessary
OVP.
Other essential functions that should be found in the power
supply specification, but that many 'bean-counter' selected
supplies may be missing:
Specification compliance: ATX 2.03 & ATX12V v1.1
Short circuit protection on all outputs
Over voltage protection
Over power protection
EMI/RFI compliance: CE, CISPR22 & FCC part 15 class B
Safety compliance: VDE, TUV, D, N, S, Fi, UL, C-UL & CB
Hold up time, full load: 16ms. typical
Efficiency; 100-120VAC and full range: >65%
Dielectric withstand, input to frame/ground: 1800VAC, 1sec.
Dielectric withstand, input to output: 1800VAC, 1sec.
Ripple/noise: 1%
MTBF, full load @ 25°C amb.: >100k hrs
Power supplies missing these and other functions are sold at
good profit in the North American computer clone market. OVP
must be in all computer supplies but is often missing in clone
computers.
"J.J." wrote:
> If a computer PSU fails then I have heard that it may (or may
> not) blow the mainboard and perhaps various other components
> with a power surge or soemthing like that.
>
> It seems that better PSUs are designed so that when they fail
> they have some circuitry which protects the other components
> in the PC.
>
> Is this failsafe feature of the PSU I am referring to pretty
> much the same feature you are referring to? Or are they
> separate features?
>
> Does anyone know how common it is to get this failsafe
> feature in a PSU?
Charles Perry
<snip>
> There is nothing in a UPS that will accomplish the necessary
> OVP.
You need to specify which kind of UPS. Some UPS will provide excellant over
voltage protection.
Charles Perry P.E.
Mike
>> HDD are frequently corrupted due to power events.
>Does anyone have a reference to HDDs getting corrupted by power
>events on the mains power supply?
Reference Microsoft :)
"Because you didn't shut down Windows properly, Scandisk is now trashing
your disk to complete the job.
In future, always shut down Windows properly"
Scandisk is not *always* going to pull your nuts out of the fire after a
power interruption. In fact, it might make things worse :)
Wouldn't you *expect* data corruption, if the system was writing/about to
write cached data, and the mains power went off?
You don't need references to figure out that it's a bad idea to just
lose power in an uncontrolled way.
--
--------------------------------------+------------------------------------
Mike Brown: mjb[at]pootle.demon.co.uk | http://www.pootle.demon.co.uk/
w_tom
12 volt outputs. Table 11 from Intel specs even defines where
and what that OVP must do. There is nothing in a plug-in UPS
- outputting 120 VAC or 230 VAC to power supply - that is
going to over voltage protect those DC outputs. Nothing.
Only OVP that a UPS can provide - limit 120 VAC or 230 VAC.
That will not solve an overvoltage problem on DC output of a
'defective by design' power supply.
J.J. asked:
> If a computer PSU fails then I have heard that it may
> (or may not) blow the mainboard and perhaps various
> other components ...
Yes, if power supply does not have OVP. No if supply does
have the required OVP. No UPS will solve this missing OVP
problem on DC outputs of power supply.
w_tom
one technical fact to post in response. At least a junk
scientist would try to invent a fact. Ron Reaugh is even
worse than a junk scientist. He insults. Some claim that a
plug-in UPS provided hardware protection. They can insult.
That alone proves they must be right. Hey Ron. Is god on
your side? No wonder you just 'know' these things.
Keith
> Again, the trashed filesystem is a problems of FAT and other
> simplistic file systems. It is not a problem to superior
> (journalizing) filesystems.
Which would be?
>
> Will a disk drive write to the platter as voltage drops?
Certainly, at least to some point.
> Of course not.
You're *ONCE AGAIN* yalking through your ass.
> The disk drive controller is a complete computer
> that also monitors voltage.
Really? I'm not from Missouri, but close enough. An IDE port monitors
its supply voltage? You're simply talking out your ass, since it's been
shot off repeatedly.
> It does not matter to disk
> hardware when power is turned off. But it does matter to some
> 'simplistic' disk filesystems that power is not removed during a write
> operation.
>
> Just another reason why FAT was obsoleted by HPFS which in
> turn was obsoleted by NTFS.
You haven't a clue (as usual). NTFS is a slight modification to HPFS
(written by the same SB, AFAIK) to make sure that OS/2 couldn't access NT
systems. Neither is a JFS, nor is either less corrruptable than FAT.
Indeed NT systems are far more susceptable to corruption than other
similar OSs because of the agressive write buffering. Even (non-JFS) OS/2
systems are better at self-healing than NT. Of course JFS is a standard
part of OS/2 now. Windows? YMBK!
--
Keith
jmfb...@aol.com
w_tom <w_t...@hotmail.com> wrote:
People might take you a tad more seriously if you exhibited
a scintilla of trainability. The next line is a hint...
<snip topposting>
jmfb...@aol.com
Keith <k...@att.bizzzz> wrote:
>On Fri, 09 Jul 2004 19:27:13 -0400, w_tom wrote:
>
>> Again, the trashed filesystem is a problems of FAT and other
>> simplistic file systems. It is not a problem to superior
>> (journalizing) filesystems.
>
>Which would be?
>>
>> Will a disk drive write to the platter as voltage drops?
>
>Certainly, at least to some point.
>
>> Of course not.
>
>You're *ONCE AGAIN* yalking through your ass.
>
>> The disk drive controller is a complete computer
>> that also monitors voltage.
>
>Really? I'm not from Missouri, but close enough. An IDE port monitors
>its supply voltage? You're simply talking out your ass, since it's been
>shot off repeatedly.
>
>
>> It does not matter to disk
>> hardware when power is turned off. But it does matter to some
>> 'simplistic' disk filesystems that power is not removed during a write
>> operation.
>>
>> Just another reason why FAT was obsoleted by HPFS which in
>> turn was obsoleted by NTFS.
>
>You haven't a clue (as usual). NTFS is a slight modification to HPFS
>(written by the same SB, AFAIK) to make sure that OS/2 couldn't access NT
>systems. Neither is a JFS, nor is either less corrruptable than FAT.
How in the world does he think that FAT has anything to do with
physical disk specs?
>Indeed NT systems are far more susceptable to corruption than other
>similar OSs because of the agressive write buffering.
That's really, really, really too bad. It used to know how.
<snip>
jmfb...@aol.com
m...@posie.local.dom (Mike) wrote:
>In article <9520B3D00...@208.42.66.156>, J.J. <nom...@mail.com>
wrote:
>
>>> HDD are frequently corrupted due to power events.
>
>>Does anyone have a reference to HDDs getting corrupted by power
>>events on the mains power supply?
>
>Reference Microsoft :)
>
>"Because you didn't shut down Windows properly, Scandisk is now trashing
> your disk to complete the job.
>
> In future, always shut down Windows properly"
>
>Scandisk is not *always* going to pull your nuts out of the fire after a
>power interruption. In fact, it might make things worse :)
>
>Wouldn't you *expect* data corruption, if the system was writing/about to
>write cached data, and the mains power went off?
Not if you have an OS that knows what it's doing. A great part
of file system code is supposed to anticipate such things. For
instance, one approach is to not put the file entry in the user
file directory file until after the last bit is written and the
user's prog has issued a close.
>
>You don't need references to figure out that it's a bad idea to just
>lose power in an uncontrolled way.
Of course it's a bad idea :-) but most systems, that went through
a design, allocate quite a bit of their code to Murphy anticipation.
Richard Herring
writes
That objection misses the point. You may lose a lot of data because of
the buffering, but the commit/rollback transaction processing supposedly
means that what you lose is a complete transaction and what you have
left will always be consistent.
At least, that's the theory...
--
Richard Herring
w_tom
intelligent functions? A disk drive is an embedded computer
complete with voltage monitoring circuits. This controller is
not the same as an IDE port.
Long before voltage drops below what digital circuits on
disk drive require, the disk drive has detected the falling
voltage and stopped writing. Disk drive hardware protects
itself from voltage drop which is also why a power down will
not interfere even with normal disk drive housekeeping.
The idea that an IDE port monitors voltage is an erroneous
assumption and was not made in any previous post. An IDE port
is nothing more than a bidirectional repeater. An IDE port
has no intelligent functions and does not monitor voltage.
Where did you get this idea that IDE port functions were even
being discussed? Are you confusing IDE with some other
hardware interface? Or did you just fail to read that
previous post will sufficient care?
Posted previously was:
> The disk drive controller is a complete computer that
> also monitors voltage.
Where is an IDE interface even implied here?
Keith wrote:
> On Fri, 09 Jul 2004 19:27:13 -0400, w_tom wrote:
>> Will a disk drive write to the platter as voltage drops?
>
> Certainly, at least to some point.
>
>> Of course not.
>
> You're *ONCE AGAIN* yalking through your ass.
>
>> The disk drive controller is a complete computer
>> that also monitors voltage.
>
> Really? I'm not from Missouri, but close enough. An IDE port
> monitors its supply voltage? You're simply talking out your
> ass, since it's been shot off repeatedly.
> ...
jmfb...@aol.com
Sigh! Those problems were solved (one product was JMF's) in a variety
of ways in an OS that smells like NT. Consistency wasn't the major
problem with the airline reservation system.
>
>At least, that's the theory...
Yea. And then somebody tries to program the thing. Then you
find out what the real theory is ;-).
jmfb...@aol.com
w_tom <w_t...@hotmail.com> wrote:
> You knew that an IDE port is not a computer; has no
>intelligent functions? A disk drive is an embedded computer
>complete with voltage monitoring circuits.
Oh, my! You are young.
<pins>
Folkert Rienstra
"Keith" <k...@att.bizzzz> wrote in message news:pan.2004.07.12...@att.bizzzz...
> On Fri, 09 Jul 2004 19:27:13 -0400, w_tom wrote:
>
> > Again, the trashed filesystem is a problems of FAT and other
> > simplistic file systems. It is not a problem to superior
> > (journalizing) filesystems.
>
> Which would be?
> >
> > Will a disk drive write to the platter as voltage drops?
>
> Certainly, at least to some point.
>
> > Of course not.
>
> You're *ONCE AGAIN* yalking through your ass.
>
> > The disk drive controller is a complete computer
> > that also monitors voltage.
>
> Really? I'm not from Missouri, but close enough. An IDE port monitors
> its supply voltage? You're simply talking out your ass, since it's been
> shot off repeatedly.
And yours blew-up just now when you can't make the distinction between
an IDE Disk Controller and an IDE Hostbus Adapter.
>
>
> > It does not matter to disk hardware when power is turned off. But it
> > does matter to some 'simplistic' disk filesystems that power is not
> > removed during a write operation.
*And* to the drive when it may encounter a bad sector afterwards,
nomatter what filesystem is in use, though the 'damage' is temporary.
Folkert Rienstra
<jmfb...@aol.com> wrote in message news:40f278b4$0$1190$61fe...@news.rcn.com...
> In article <40F257AD...@hotmail.com>, w_tom <w_t...@hotmail.com> wrote:
> > You knew that an IDE port is not a computer; has no
> > intelligent functions? A disk drive is an embedded computer
> > complete with voltage monitoring circuits.
>
> Oh, my! You are young.
Well, in that case you are probably old as methusalem.
Not a working braincell left in your cranium.
w_tom
concepts:
> That objection misses the point. You may lose a lot of
> data because of the buffering, but the commit/rollback
> transaction processing supposedly means that what you
> lose is a complete transaction and what you have left
> will always be consistent.
jmfb...@aol.com
Unix also has a different OS philosophy w.r.t. file organizations.
> ... I'd admittedly have
>to look for the details though, and ext2's organization
>is quite different from FAT's or NTFS. FAT in particular
>is terrible, basically every file is a single chain --
>but you probably knew that already.
No, it's worse than that. NOte that I have never read the code
nor the specs of FAT. However, based on the way it "behaves"
on my machine FAT treats the whole disk as a single chain.
This does not honor directory boundaries the way sane people
would expect. This would also explain all the werid-assed bugs
DOS and its layers have.
> ... NTFS is more or less
>as I've described it in my prior post, at a high level,
>and it feels like an engineered solution,
I would hope so. It's been getting "developed" since 1971.
> ...whereas Linux's
>ext2 is more elegant, even if it's still engineered.
>But there's no perfect solution anyway;
There isn't going to be one, and only one, solution because
each choice solves different problems and has orthogonal
design goals. I don't know about today but in the olden
days, the choice was between "fast" retrieval or humungous
files, a.k.a. data bases. If your system had to maintain
one file that fit on 100 disk packs, your file system OS
code would look and behave differently from a system that
needed to maintain 200,000 small files for 10,000 different
users who accessed them on unpredictable days and times.
> ... as you've described
>the problem, there's always going to be a hole or two,
>and a determined program can probably fragment any file
>system if it does something like the following:
>
>open big file
>write block to big file
>open little file
>write block to little file
>close it
>write block to big file
>open little file
>write block to little file
>close it
>write block to big file
>....
>
>(It's a good thing the trend is towards centralized dedicated-machine
>syslog-type logging. :-) )
See my description above; I just threw a kink in your POV.
>
>I'll admit to wondering whether NT had the rather interesting
>capability or not of "let's just write it here".
That depends on the definition of "here". Are you talking
about the logical placement of the file or the physical
placement of the file? There are other flavors of "here"
but I go into them. :-)
> .. I base
>this hypothesis on observations using DiskKeeper Lite, which
>copy I had at the time on a machine at my then-employer.
>Basically, the notion is to simply write the new block at
>an open sector in the cylinder over which the head is
>flying.
Ah, you were talking about physical. Note that physical has
nothing to do with the FAT nor NTFS.
Now the problem with cylinders is that they spin. The problem
with tracks is that they're a circle.
>
>Of course this would fragment things terribly, and I have no proof.
No,no,no. Now you're confusing logical bit placement with physical
bit placement. The two really (TW's going to kill me) don't have
much to do with each other.
>But things did fragment pretty badly when I used such tools
>as Visual C++.
Now you're talking about logical placement.
You are confused :-). However, I can't help you very much
because it would be a case of the blind leading the blind.
There are other people who are bit gods in this area.
jmfb...@aol.com
"Folkert Rienstra" <see_re...@myweb.nl> wrote:
>
><jmfb...@aol.com> wrote in message
news:40f278b4$0$1190$61fe...@news.rcn.com...
>> In article <40F257AD...@hotmail.com>, w_tom
<w_t...@hotmail.com> wrote:
>> > You knew that an IDE port is not a computer; has no
>> > intelligent functions? A disk drive is an embedded computer
>> > complete with voltage monitoring circuits.
>>
>> Oh, my! You are young.
>
>Well, in that case you are probably old as methusalem.
Not quite.
>Not a working braincell left in your cranium.
At least I'm not dying of PCitis.