[snip]Jay - who claims transformers provide no galvanic isolation
- also denies the longitudinal mode transient. He must avoid
discussing the various types of transients to promote myths.
Longitudinal mode transients damage hardware. A most famous
longitudinal mode transient is lightning.
[snip]Nuh, live in a farady cage. No more problems then.
A transformer has *nothing* to do with "galvanic isolation", ...
Jay T. Blocksom said:[snip]Jay - who claims transformers provide no galvanic isolation
- also denies the longitudinal mode transient. He must avoid
discussing the various types of transients to promote myths.
Longitudinal mode transients damage hardware. A most famous
longitudinal mode transient is lightning.
You think so, eh?
Then please explain why:
<http://search.yahoo.com/search?p="longitudinal+mode+transient"+lightning>
Produces ZERO hits, and why the ONLY hit produced by:
<http://www.google.com/search?q="longitudinal+mode+transient"+lightning&btnG=Google+Search>
is <http://www.talkaboutvideo.com/group/alt.tv.tech.hdtv/messages/99839.html>
which is a WWW-based Usenet mirror/archive of your own post to
<alt.tv.tech.hdtv> on [Thu, 22 Jul 2004 19:56:14 -0400] (Message-ID:
<[email protected]>]
Ditto for:
<http://web.ask.com/web?q="longitudinal+mode+transient"+lightning>
which produces ONLY <http://www2.usenetarchive.org/Dir88/File532.html>, which
is yet another Usenet mirror/archive, this time of someone else quoting your
earlier rant on <alt.tv.tech.hdtv>.
So... If this term is so accepted and so "famous", then why is it not
documented ANYWHERE, except in your own fertile imagination? Hmmm?
I repeat: Get back under your bridge and resume your meds.
Going off topic ...
Has anyone ever seen a UPS that *replaces* a computer's PSU
instead of merely sitting between the PSU and the wall outlet ?
I've often wondered if a PSU could be more efficient if it took
the AC input from the wall inlet, converted it to DC, and then
never had to deal with AC again. In other words, if the UPS took
the AC power from the wall outlet, converted it to DC, and then
used the DC to do two things: maintain the charge on the battery
and provide the 12V, 5V and 3.3V for the computer.
As things currently stand, every UPS I've seen takes in AC power
and outputs only AC power. When the battery is not needed, this
is not a big deal. However, when the battery *is* needed it seems
silly to take DC from the battery and convert it to AC - only to
send that AC to a PSU which has to convert it back to DC again.
Alternatively, how about a new kind of PSU designed to work in
conjunction with a new kind of UPS: the UPS outputs DC at
voltage "X" instead of AC and the PSU is designed to take only DC
at "X" volts ?
When running a computer from automotive power, one is
cautioned about another transient: load dump. Load dump is
defined in specs such as SAE J1455 and ISO 7637-1. Defined
are automotive transients up to 270 volts and energy up to 50
joules. These are worst case numbers. But a power supply for
computers (in automotive and marine environments) should
withstand load dump spikes of 100+ volts on 12 volt power.
Those whose only number is price will buy a $25 power
supply. This is the classic bean counter decision. Those who
make decisions using numbers may spend $50 for the same supply
that includes other computer safety features such as load dump
protection.
68% power efficiencies is on the low side. Intel requires
68% as the absolute minimum when a power supply is 100%
loaded. Most power supplies are more like 75% efficient.
This number may also be impacted by line harmonics. European
industry standards require power factor correction which means
AC line power is used more efficiently.
Yes these are off topic subjects because they don't do
anything for the OP's question. These alternative supplies
neither add nor impact motherboard hardware protection.
Using dedicated batteries for 3.3, 5, and 12 volts is not a
valid solution. Battery voltages vary too much. Computers
require much better voltage regulation. Not as critical as
required by 8080s back when I was soldering together computers
(Jay is reminded who here has dirt under the fingernails).
But still battery voltages vary so much as to require voltage
regulators - which then creates efficiencies found in current
supplies. Dedicated batteries provide nothing useful and can
even create other destructive problems not relevant to the OPs
question.
I've often wondered if a PSU could be more efficient if it
took the AC input from the wall inlet, converted it to DC,
and then never had to deal with AC again.
[snip]Going off topic ...
Has anyone ever seen a UPS that *replaces* a computer's PSU
instead of merely sitting between the PSU and the wall outlet ?
[snip]I've often wondered if a PSU could be more efficient if it took
the AC input from the wall inlet, converted it to DC, and then
never had to deal with AC again.
... near-impossibility, really -- of stuffing all the necessary
components (possibly including several batteries) into the
confines of a "standard" PC PSU form-factor.
... need to either dedicate a battery (or bank of batteries)
and the attendant charging/monitoring circuits to *each*
different voltage rail in the system ...
Jay T. Blocksom said:It's been done; but it never caught on commercially.
There are several problems inherent in such an approach, not he least of which
is the difficulty -- near-impossibility, really -- of stuffing all the
necessary components (possibly including several batteries) into the confines
of a "standard" PC PSU form-factor.
Then there's the fact that, at least in most applications, keeping the PC
"alive" during an outage doesn't do you all that much good, if you cannot also
keep at least some of the peripherals (monitor, network hub/switch, etc.)
alive. So you'd still need a "general purpose" UPS for these devices, which
sort'a defeats the point.
All in all, it's just not a cost-effective or practical approach, unless
you've got a *really* weird set of "special needs".
[snip]I've often wondered if a PSU could be more efficient if it took
the AC input from the wall inlet, converted it to DC, and then
never had to deal with AC again.
In theory, yes. But in "real life", the first major problem is you'd need to
either dedicate a battery (or bank of batteries) and the attendant
charging/monitoring circuits to *each* different voltage rail in the system
(for a standard desktop PC, that's currently: +3.3V, +5.0V, +12V, -12V, and
-5.0V -- for a total of five batteries, minimum), which gets both expensive
and bulky in a large hurry (not to mention a maintenance headache), or suffer
a major efficiency "hit" via one or more voltage-divider circuits. Then
there's the fact that standard batteries do not offer the "correct" voltages
needed by current PC designs (no, the output of a "12V" battery is *not*
really 12 volts); so we still need to throw away (primarily in the form of
heat, BTW) some more voltage, even if we do go the mutiple-battery route.
In addition to these problems, there are (as touched on above) packaging
issues: One battery or several, the overall system inevitably becomes
significantly larger and heavier -- precisely the opposite of what "The
Market" generally wants. And even if we manage to cram all the necessary
"stuff" into a sufficiently small package, who wants a potentially leaky
lead-acid battery inside their PC case? And if we go with more exotic battery
types (such as Lithium-Ion, which would be the best choice among currently
viable technologies) to bypass that issue, we both complicate the required
circuitry still further *and* blow the budget sky high. (Have you priced
high-capacity Li-Ion batteries? Sit down before you do. <~>)
This is one of the big reasons (tho' of course, not the only one) why laptop
PCs use fundamentally different processors, chipsets, etc., than their desktop
counterparts do. By making the various system components all run at
approximately the same voltages (and reducing the total power demand as much
as possible), the overall system design is greatly simplified.
Jay said:[snip]Going off topic ...
Has anyone ever seen a UPS that *replaces* a computer's PSU
instead of merely sitting between the PSU and the wall outlet ?
It's been done; but it never caught on commercially.
There are several problems inherent in such an approach, not he least of which
is the difficulty -- near-impossibility, really -- of stuffing all the
necessary components (possibly including several batteries) into the confines
of a "standard" PC PSU form-factor.
Then there's the fact that, at least in most applications, keeping the PC
"alive" during an outage doesn't do you all that much good, if you cannot also
keep at least some of the peripherals (monitor, network hub/switch, etc.)
alive. So you'd still need a "general purpose" UPS for these devices, which
sort'a defeats the point.
All in all, it's just not a cost-effective or practical approach, unless
you've got a *really* weird set of "special needs".
[snip]I've often wondered if a PSU could be more efficient if it took
the AC input from the wall inlet, converted it to DC, and then
never had to deal with AC again.
In theory, yes. But in "real life", the first major problem is you'd need to
either dedicate a battery (or bank of batteries) and the attendant
charging/monitoring circuits to *each* different voltage rail in the system
(for a standard desktop PC, that's currently: +3.3V, +5.0V, +12V, -12V, and
-5.0V -- for a total of five batteries, minimum), which gets both expensive
and bulky in a large hurry (not to mention a maintenance headache), or suffer
a major efficiency "hit" via one or more voltage-divider circuits. Then
there's the fact that standard batteries do not offer the "correct" voltages
needed by current PC designs (no, the output of a "12V" battery is *not*
really 12 volts); so we still need to throw away (primarily in the form of
heat, BTW) some more voltage, even if we do go the mutiple-battery route.
In addition to these problems, there are (as touched on above) packaging
issues: One battery or several, the overall system inevitably becomes
significantly larger and heavier -- precisely the opposite of what "The
Market" generally wants.
And even if we manage to cram all the necessary
"stuff" into a sufficiently small package, who wants a potentially leaky
lead-acid battery inside their PC case? And if we go with more exotic battery
types (such as Lithium-Ion, which would be the best choice among currently
viable technologies) to bypass that issue, we both complicate the required
circuitry still further *and* blow the budget sky high. (Have you priced
high-capacity Li-Ion batteries? Sit down before you do. <~>)
[snip][snip]Going off topic ...
Has anyone ever seen a UPS that *replaces* a computer's PSU
instead of merely sitting between the PSU and the wall outlet ?
It's been done; but it never caught on commercially.
There are several problems inherent in such an approach, not he least of
which is the difficulty -- near-impossibility, really -- of stuffing all
the necessary components (possibly including several batteries) into the
confines of a "standard" PC PSU form-factor.
I never assumed that it would go inside the computer in place of
the current types of PSUs - simply because of battery sizes and
the unpleasant prospect of having leaking batteries inside a
computer.
[snip]In other words, instead of having an external UPS and an internal
PSU, people would have an external (UPS + PSU).
[snip]I wasn't thinking of something that targets "The Market" - just
something for people who use UPSes anyway.
[snip]Most people I know - including myself - don't need a UPS with a
lot of run time for a home PC, so a small Li-Ion battery could be
financially feasible.
[snip]I need a UPS to get me through the momentary power dropouts that
happen once in a while - particularly during thunderstorm season.
Around here, blackouts lasting more than a few seconds happen
only once or twice a year and I can live without a computer
during those - so long as I have a minute or two to save and shut
down nicely.
[snip]Heck, even the people I know who have big UPSes to get them
through a 20 minute blackout have without exception wasted their
money. Instead of continuing to use their computers during the
blackouts, they still do what everyone else does: look for
candles/flashlights and phone people in other parts of town to
ask if they are blacked out too. ;-) Long run-times are great
for a server UPS, but human nature being what it is makes it a
waste for desktops.
[snip]This reminds me of something ...
After a blackout last year, I set a local business up with 1000VA
UPSes at every computer to see them through the next blackout -
they wanted something that would let people keep working for up
to 30 minutes.
[snip]Two days later I was there as an ordinary
customer when, lo and behold, the next blackout happened. Turns
out they had a policy of shutting everything down, kicking the
customers out, and locking the doors. Much cheaper and smaller
UPSes with a two minute run-time would have done the job.
[snip]We built a PC containing even screen and keyboard inside a
12 inch by 12 inch by 4 inch box. The battery (lead acid) and
power supply were also inside (late 1990s design). It
executed
Windows and was networked. But then Jay somehow knows:
Package was much smaller than a standard PC form-factor.
[snip]The battery does not go inside a power supply for other and
obvious human safety reasons.
[snip]Somehow Jay would have one believe another mythical
requirement; a:
One output regulator provided all voltages from only one
battery - even though Jay says this working design is not
possible.
Jay said:[snip]On Fri, 01 Apr 2005 19:27:18 GMT, in <alt.comp.periphs.mainboard.asus>,
Going off topic ...
Has anyone ever seen a UPS that *replaces* a computer's PSU
instead of merely sitting between the PSU and the wall outlet ?
[snip]
It's been done; but it never caught on commercially.
There are several problems inherent in such an approach, not he least of
which is the difficulty -- near-impossibility, really -- of stuffing all
the necessary components (possibly including several batteries) into the
confines of a "standard" PC PSU form-factor.
I never assumed that it would go inside the computer in place of
the current types of PSUs - simply because of battery sizes and
the unpleasant prospect of having leaking batteries inside a
computer.
OK, but such packaging was rather strongly implied by the "*replaces* a
computer's PSU" phrase. And FWIW, the early incarnations of this idea that I
am aware of did indeed do just that. As you might imagine, the run-times on
battery power were abysmal (a minute or two, as I now rather vaguely recall).
[snip]In other words, instead of having an external UPS and an internal
PSU, people would have an external (UPS + PSU).
Effectively, what you're talking about is a two-piece PC. While, in theory,
the so-called UPS/PSU unit could be used with any similar "PSU-less" PC, it is
easily foreseeable that in practice it would not work out that way at least
the vast majority of time, in part because the PSU-less PC portion would be
useless without its "other half". Hence, at least the vast majority of PCs
designed for such as system would need to be sold with an effectively
dedicated UPS/PSU unit; so we're back to the size/weight issues mentioned
earlier (and the fact that, since this arrangement is obviously more
complicated than "plug it into a standard wall jack and go", it's not likely
to have much appeal to John Q. Average).
[snip]I wasn't thinking of something that targets "The Market" - just
something for people who use UPSes anyway.
But that *is* the market -- at least, the market for the product you're
envisioning.
Let me ask you this... Just what significant benefit to the consumer would
such an approach yield, which cannot or is not already provided by "standard"
PCs and "standard" plug-in UPS systems? Before you answer, consider that
whatever that benefit is, it would have to out-weigh the several advantages
the current approach provides that your pseudo-dedicated approach can't,
particularly in terms of flexibility (i.e., a single UPS can "protect" several
pieces of equipment simultaneously; the particular assortment of equipment
chosen at any given moment is completely up to the user; no need to replace
the UPS when replacing the PC; the UPS is not actually *needed* to operate the
PC; etc.).
[snip]Most people I know - including myself - don't need a UPS with a
lot of run time for a home PC, so a small Li-Ion battery could be
financially feasible.
Maybe. But *five* of them? I don't think so.
[snip]I need a UPS to get me through the momentary power dropouts that
happen once in a while - particularly during thunderstorm season.
Around here, blackouts lasting more than a few seconds happen
only once or twice a year and I can live without a computer
during those - so long as I have a minute or two to save and shut
down nicely.
I don't debate that. However...
A "standard" AC-powered PC in combination with a "standard" plug-in UPS can
already do everything you want, and is *cheaper* than your hypothetical
pseudo-dedicated PSU/UPS system likely could ever be. So those other "side
benefits" (such as the longer run time) that you feel you don't need can
simply be ignored -- you're *still* better off (in the cost/benefit sense)
with the conventional approach.
And besides, what if you're not sitting right at the computer to immediately
initiate that shut-down, when the power failure happens to come along?
[snip]Heck, even the people I know who have big UPSes to get them
through a 20 minute blackout have without exception wasted their
money. Instead of continuing to use their computers during the
blackouts, they still do what everyone else does: look for
candles/flashlights and phone people in other parts of town to
ask if they are blacked out too. ;-) Long run-times are great
for a server UPS, but human nature being what it is makes it a
waste for desktops.
Much depends on how you look at it, of course. But from where I sit, it's not
a waste; it's cheap insurance.
The cost difference between two otherwise identical UPSs, one with a 5-minute
run-time under a given load, the other with, say, a 20-minute run-time with
that same load, is usually relatively minimal, perhaps even negligible.
There's no point in being "penny-wise and pound-foolish".
[snip]This reminds me of something ...
After a blackout last year, I set a local business up with 1000VA
UPSes at every computer to see them through the next blackout -
they wanted something that would let people keep working for up
to 30 minutes.
Did the systems in question actually *require* a UPS with that much load
capacity, or was this simply your expedient approach to providing extended
run-rimes?
If the latter, it would very likely have been cheaper to use
smaller UPS systems with extra battery packs.
(And I presume you did not neglect such things as the servers, routers,
switches, printers, etc., right?)
[snip]Two days later I was there as an ordinary
customer when, lo and behold, the next blackout happened. Turns
out they had a policy of shutting everything down, kicking the
customers out, and locking the doors. Much cheaper and smaller
UPSes with a two minute run-time would have done the job.
I think you're exaggerating a bit, but I get your point.
In any event, this
is not an indictment of using conservatively-sized UPS systems.
have had *other* reasons for that policy -- like liability concerns, for
example.
[snip]Jay T. Blocksom wrote: [snip][snip]Let me ask you this... Just what significant benefit to the consumer
would such an approach yield, which cannot or is not already provided by
"standard" PCs and "standard" plug-in UPS systems?
I had hoped that there would be significant energy efficiency
gains by combining the UPS and the PSU into a single unit.
Apparently that is not the case.
[snip]This business was just one franchise in a chain. The UPSes to be
used were spec'd out by the chain.
[snip]They had originally had a server+terminals setup, with UPSes for
everything but the terminals. When they switched to a
server+windows clients setup they protected everything but the
clients. They operated that way for three years before their HQ
found out and told them to add the UPSes for the clients. I also
visited a different franchise in a nearby city to see if they had
those big UPSes at each client - and they did.
[snip][snip]Two days later I was there as an ordinary
customer when, lo and behold, the next blackout happened. Turns
out they had a policy of shutting everything down, kicking the
customers out, and locking the doors. Much cheaper and smaller
UPSes with a two minute run-time would have done the job.
I think you're exaggerating a bit, but I get your point.
Not exaggerating at all. This particular franchise was in a
shopping mall. I wondered if the chain has other stores that are
not in malls, and if so what is their policy about expelling
customers in the middle of a thunderstorm ? Would they just
send everyone into the little waiting/reception area ?
[snip]I didn't intend it to be. I intended it just as an example of
bad UPS usage.
They may well
This business had agents sitting at a desk and customers sitting
across the desk from the agents. I would have thought the safe
thing to do would be for everyone to just stay where they were
when the power went out.