Power supply EXPLOSION

On Sun, 25 Jul 2004 18:25:44 +0100, "Peter Hucker"
wrote:

snip

And now it's worth nothing. You should have experimented on one of
your worse supplies because a 300W Fortron-Source will outperform many
350W and even some 400W units.

Worse? That WAS the worse one. I didn't want to break my gold plated dual fan 600 watt thing. The 300W one I broke was absolutely terrible - 10% voltage accuracy?! **** that. This 600 watt one is 2%. And what's Fortron-Source? A make or a design?

Fortron/Source, also known as SPI or Sparkle, is a major
manufactuer of PC PSU. They make more than one quality grade at
each wattage level. Your PSU could've been great for 300W,
average, or not so good. It is not certain that your PSU is even
a Fortron/Source though (but 2nd bit of evidence, that "choke" yo
applied heatsink to does suggest it), anyone could use a bolt to
secure the heatsink.. it's not very common but not an advanced
design that couldn't be duplicated by any sweatshop using such a
cheap/thin pcb that they needed some extra support of components.
I"m not suggesting that a poor pcb is why Fortron does it, but
rather that it is another reason it could be done.

Is your gold plated dual fan 600W a major name-brand? If not,
it's extremely unlikely that it's much better than a Fortron.
The thing about off-brands/generics is that (especially in their
high wattage models) the specs on the label are usually off by so
much that it may be illegal to market them in this way, yet due
to apathy no one is stopping them. Then again, this is a good
reason for a generic label, like gypsys they can just dump them
under a different label and/or channel if the need arises.

A good 600W PSU will cost over $100, potentially a lot more. If
yours cost less you might be better off assuming it's only worth
(roughly) 320W, if that. Running such a PSU on higher demand
system could damage PSU and connected parts. When a PSU is
undersized for it's rated wattage it may deviate far more from
spec to compensate, regardless if it's claimed spec. If you
tested it and found it not within specs at rated wattage (which
is likely), what is your recourse? With enough effort you might
get it replaced with same thing, but still you dont' end up with
a PSU capable of sustaining rated wattage at rated tolerance.

If you had a picture of label and/or entire open top of the
failed PSU we might be able to make positive ID.


That picture indicates it's a power factor correction choke, as
evidenced by the "PFC" printed over it, and if air flow wouldn't cool
it significantly, I doubt the manufacturer would have included those 3
air slots on each side. Also the metal box over it probably doesn't
make good contact with its core inside, so a heatsink external to it
would probably be useless. This choke is a high voltage device and
probably located next to high voltage components that aren't
completely insulated, so be very careful in installing a heatsink on
it.

It seemed pretty insulated to me, and I never even got a tickle off it.

It is fairly well insulated, but like many PSU parts it's
designed to withstand manufacturing installation and use of PSU,
but after PSU is opened and user manipulates parts or alters PSU
design, the parameters for temp or durability may be exceeded.



The thing was very hot without a heatsink, and cool with one, so I assumed that the heat was getting to the oiutre casing just fine.

When a part is encased, it's inevitable that outer casing is hot
if interior is. The heatsink you added, while potentially risky,
may've been better than nothing since the PSU ran without fan,
but issue is then if part ran as cool with this heatsink as it
would've with forced-air cooling, or at least within tolerance
for part... with specs for part you may be guessing without any
idea what part's interior temp is.

On Sun, 25 Jul 2004 19:25:40 GMT, kony wrote:

On Sun, 25 Jul 2004 18:25:44 +0100, "Peter Hucker"
wrote:

snip

And now it's worth nothing. You should have experimented on one of
your worse supplies because a 300W Fortron-Source will outperform many
350W and even some 400W units.

Worse? That WAS the worse one. I didn't want to break my gold plated dual fan 600 watt thing. The 300W one I broke was absolutely terrible - 10% voltage accuracy?! **** that. This 600 watt one is 2%. And what's Fortron-Source? A make or a design?

Fortron/Source, also known as SPI or Sparkle, is a major
manufactuer of PC PSU.

Ah - yes it said sparkle. [photographs label later in thread - no, my memory is apalling (and so is my spelling), it's a suntek]

They make more than one quality grade at
each wattage level. Your PSU could've been great for 300W,
average, or not so good. It is not certain that your PSU is even
a Fortron/Source though (but 2nd bit of evidence, that "choke" yo
applied heatsink to does suggest it), anyone could use a bolt to
secure the heatsink.. it's not very common but not an advanced
design that couldn't be duplicated by any sweatshop using such a
cheap/thin pcb that they needed some extra support of components.
I"m not suggesting that a poor pcb is why Fortron does it, but
rather that it is another reason it could be done.

Is your gold plated dual fan 600W a major name-brand? If not,
it's extremely unlikely that it's much better than a Fortron.

No name on it at all. Extrmely cheap, but nice - 4 coloured LEDs on each of 2 transparent fans! And it does give very accurate stable voltages.

The thing about off-brands/generics is that (especially in their
high wattage models) the specs on the label are usually off by so
much that it may be illegal to market them in this way, yet due
to apathy no one is stopping them. Then again, this is a good
reason for a generic label, like gypsys they can just dump them
under a different label and/or channel if the need arises.

True enough, I have not drawn over 200 watts from it. But it's working so well at 200, I assume it'd manage 600.

A good 600W PSU will cost over $100, potentially a lot more.

I paid $38.50! There is no way a 600W supply is worth $100. I'd expect a fanless one for that. Where on earth are you sourcing your equipment?

If
yours cost less you might be better off assuming it's only worth
(roughly) 320W, if that. Running such a PSU on higher demand
system could damage PSU and connected parts. When a PSU is
undersized for it's rated wattage it may deviate far more from
spec to compensate, regardless if it's claimed spec. If you
tested it and found it not within specs at rated wattage (which
is likely), what is your recourse?

Get a refund of course. But as it is within 2% at 200 watts, I would assme it would be fine at 600. It's doing the job 5 times better than the suntek did, so I'm happy. I won't ever be drawing 600 watts.

With enough effort you might
get it replaced with same thing, but still you dont' end up with
a PSU capable of sustaining rated wattage at rated tolerance.

No, I'd get my money back or a voucher for a different supply that worked correctly. You come from the land of sueing people for slightly squint paving stones - you should know this.

If you had a picture of label and/or entire open top of the
failed PSU we might be able to make positive ID.

For a post mortem? ;-)

http://80.229.155.158/temp/psulabel.jpg

That picture indicates it's a power factor correction choke, as
evidenced by the "PFC" printed over it, and if air flow wouldn't cool
it significantly, I doubt the manufacturer would have included those 3
air slots on each side. Also the metal box over it probably doesn't
make good contact with its core inside, so a heatsink external to it
would probably be useless. This choke is a high voltage device and
probably located next to high voltage components that aren't
completely insulated, so be very careful in installing a heatsink on
it.

It seemed pretty insulated to me, and I never even got a tickle off it.

It is fairly well insulated, but like many PSU parts it's
designed to withstand manufacturing installation and use of PSU,
but after PSU is opened and user manipulates parts or alters PSU
design, the parameters for temp or durability may be exceeded.

Hey that's a little unfair - I've only ever set fire to two electrical items (well 3 now if you call fireworks a fire).

The thing was very hot without a heatsink, and cool with one, so I assumed that the heat was getting to the oiutre casing just fine.

When a part is encased, it's inevitable that outer casing is hot
if interior is. The heatsink you added, while potentially risky,
may've been better than nothing since the PSU ran without fan,
but issue is then if part ran as cool with this heatsink as it
would've with forced-air cooling, or at least within tolerance
for part... with specs for part you may be guessing without any
idea what part's interior temp is.

Er whoops...

--
*****TWO BABY CONURES***** 15 parrots and increasing http://www.petersparrots.com
93 silly video clips http://www.insanevideoclips.com
1259 digital photos http://www.petersphotos.com
Served from a pentawatercooled dual silent Athlon 2.8 with terrabyte raid

Eagles may soar, but weasels aren't sucked into jet engines.

Peter Hucker wrote:
Ok, maybe watercooling the power supply was not a good idea.

It certainly kept the power transistors with the big heatsinks cool, but
what of the poor little diodes and a transformer, which didn't get the
airflow they expected?

Been working fine for a few weeks, then suddenly..... about 15 bangs,
flashes, sparks etc flew out of it, as though I had lit an entire box of
fireworks under the desk. Strangely the PC continued to run while this
happened (for about 10 seconds, at which point the PSU gave up and it
went off. Fuses all intact! Replaced the PSU, and the PC booted ok!
Just one drive of the mirror/stripe appeared to be blank/corrupted, but
it's autorebuilding it in the background.

http://80.229.155.158/temp/psufail


Did you consider that maybe the water was conducting between the
heatsinks in the powersupply -- could be why more than one heatsink is
used there? This could cause an undesired bias to be applied to the
reg. leading to the failure.

I'd tend to buy a much larger powersupply than req'd and mod it's fan to
turn real slow. If you are using 200W and only have a 300W supply
that's not much margin IMHO...

Mike

On Sun, 25 Jul 2004 20:56:10 +0100, "Peter Hucker"
wrote:



Ah - yes it said sparkle. [photographs label later in thread - no, my memory is apalling (and so is my spelling), it's a suntek]

It said sparkle but it's a suntek?
We must be on a different page.



Is your gold plated dual fan 600W a major name-brand? If not,
it's extremely unlikely that it's much better than a Fortron.

No name on it at all. Extrmely cheap, but nice - 4 coloured LEDs on each of 2 transparent fans! And it does give very accurate stable voltages.


The thing about off-brands/generics is that (especially in their
high wattage models) the specs on the label are usually off by so
much that it may be illegal to market them in this way, yet due
to apathy no one is stopping them. Then again, this is a good
reason for a generic label, like gypsys they can just dump them
under a different label and/or channel if the need arises.

True enough, I have not drawn over 200 watts from it. But it's working so well at 200, I assume it'd manage 600.

It is definitely not a safe assumption to make. Quite the
opposite, this is the very issue I pointed out, that operation at
low wattage is all the design can manage.



A good 600W PSU will cost over $100, potentially a lot more.

I paid $38.50! There is no way a 600W supply is worth $100. I'd expect a fanless one for that. Where on earth are you sourcing your equipment?

On the contrary, it is NECESSARY to spend that much to get a 600W
PSU. Ink on a label, text on a website is essentially free, of
course they can sell a 600W for $38 if the only thing "600W"
about it is the labeling.

$38 will get you a median grade ~ 450W at most, unless you're
buying something as a (salvage, OEM closeout not offered though
regular channels, etc). Even sale prices would be necessary to
get the 450W for $38. Do not be confused that a fancy casing,
neon glow LED fans or a cable sheath make a decent PSU... those
parts cost a couple dollars, ten at most and added to a PSU that
in it's plain-clothes version would sell for $10 less, they've
priced it appropriately higher as well.

As for a fanless 600W PSU, capable of lasting more than a couple
months outputting 600W (if that long, before failing), you would
have to spend far more than $200... probably several thousand
because there are no passively cooled PC PSU that can manage
600W. It is pushing the limits of PS/2 casing to even do that
with a fan(s). Servers do not use much larger PSU just for the
heck of it... some are twice the volume and not from multiple
fans.



If
yours cost less you might be better off assuming it's only worth
(roughly) 320W, if that. Running such a PSU on higher demand
system could damage PSU and connected parts. When a PSU is
undersized for it's rated wattage it may deviate far more from
spec to compensate, regardless if it's claimed spec. If you
tested it and found it not within specs at rated wattage (which
is likely), what is your recourse?

Get a refund of course. But as it is within 2% at 200 watts, I would assme it would be fine at 600. It's doing the job 5 times better than the suntek did, so I'm happy. I won't ever be drawing 600 watts.


Who said you'd be entitled to a refund? Check your warranty
statements, if you have one (otherwise you may be SOL) and see if
there is a money-back guarantee.

Again, there is no reason to assume a PSU working fine at 200W
can do so at 600W. You have not mentioned how you are sure it's
"5 times better" than the suntek, but as I mentioned previously
Sparkle makes multiple grades of PSU, the "suntek" is not their
high-end AFAIK. One thing to remember about wattage is that
beyond a unit's capable wattage, suddenly ripple may be much
worse. If the suntek was just beyond that output level, it's not
surprising if it's output wasn't so clean. Likewise, once your
generic is past it's max point, it too will have to drop voltage
every time it's amperage capacity is exceeded, else it'll shut
off, (or both).


With enough effort you might
get it replaced with same thing, but still you dont' end up with
a PSU capable of sustaining rated wattage at rated tolerance.

No, I'd get my money back or a voucher for a different supply that worked correctly. You come from the land of sueing people for slightly squint paving stones - you should know this.

You'd get money back because you "say so"?
From who?
I don't sell junk PSU, but if I did, after the (clearly
specified, short reseller period is over, typically around 30
days or less) you wouldn't get a refund from me or any other
major reseller. You would be referred to manufacturer, who is
also not obligated to money back if not specified in the
warranty.

Keep in mind that your laws might offer better protection in the
UK, this is how it is in the US. Even so, I question whether
your connected equipment would be covered, it could be a long and
costly exercise "proving" the PSU caused equipment damage.



If you had a picture of label and/or entire open top of the
failed PSU we might be able to make positive ID.

For a post mortem? ;-)

http://80.229.155.158/temp/psulabel.jpg

I have not seen inside that PSU, cannot speculate about it's
exact design but do note that it's not same labeled specs as most
300W Fortron/Sparkles.

On Sun, 25 Jul 2004 17:41:41 -0400, M.Hockings wrote:

Peter Hucker wrote:
Ok, maybe watercooling the power supply was not a good idea.

It certainly kept the power transistors with the big heatsinks cool, but
what of the poor little diodes and a transformer, which didn't get the
airflow they expected?

Been working fine for a few weeks, then suddenly..... about 15 bangs,
flashes, sparks etc flew out of it, as though I had lit an entire box of
fireworks under the desk. Strangely the PC continued to run while this
happened (for about 10 seconds, at which point the PSU gave up and it
went off. Fuses all intact! Replaced the PSU, and the PC booted ok!
Just one drive of the mirror/stripe appeared to be blank/corrupted, but
it's autorebuilding it in the background.

http://80.229.155.158/temp/psufail

Did you consider that maybe the water was conducting between the
heatsinks in the powersupply -- could be why more than one heatsink is
used there? This could cause an undesired bias to be applied to the
reg. leading to the failure.

Could be, but as I didn't get an electric shock when I touched either of the sinks to test how hot they were without sinks, I assumed both were grounded / not connected.

I'd tend to buy a much larger powersupply than req'd and mod it's fan to
turn real slow. If you are using 200W and only have a 300W supply
that's not much margin IMHO...

Possibility - but I'm trying to get hold of a good fanless one first. I've already got two slow fans blowing across the motherboard for the 3 graphics cards (with big zalman fanless sinks), the 2GB of worryingly hot memory even with big heatsinks strapped over them, and the motherboard CPU power supply.

For christs sake I wish they'd start making chips that don't give off so much bloody heat!


--
*****TWO BABY CONURES***** 15 parrots and increasing http://www.petersparrots.com
93 silly video clips http://www.insanevideoclips.com
1259 digital photos http://www.petersphotos.com
Served from a pentawatercooled dual silent Athlon 2.8 with terrabyte raid

I go fishing; I catch nothing. I go to orgies; I catch everything.

On Sun, 25 Jul 2004 22:20:36 GMT, kony wrote:

On Sun, 25 Jul 2004 20:56:10 +0100, "Peter Hucker"
wrote:



Ah - yes it said sparkle. [photographs label later in thread - no, my memory is apalling (and so is my spelling), it's a suntek]

It said sparkle but it's a suntek?
We must be on a different page.

I THOUGHT it was a sparkle - when you said sparkle it rang a bell - must have been thinking of another one.

Is your gold plated dual fan 600W a major name-brand? If not,
it's extremely unlikely that it's much better than a Fortron.

No name on it at all. Extrmely cheap, but nice - 4 coloured LEDs on each of 2 transparent fans! And it does give very accurate stable voltages.


The thing about off-brands/generics is that (especially in their
high wattage models) the specs on the label are usually off by so
much that it may be illegal to market them in this way, yet due
to apathy no one is stopping them. Then again, this is a good
reason for a generic label, like gypsys they can just dump them
under a different label and/or channel if the need arises.

True enough, I have not drawn over 200 watts from it. But it's working so well at 200, I assume it'd manage 600.

It is definitely not a safe assumption to make. Quite the
opposite, this is the very issue I pointed out, that operation at
low wattage is all the design can manage.

Doesn't matter, as I won't be drawing more than that - anyway this is a temporary one till I get a fanless one - this one will go in my server which I am currently setting up in the loft where I can't hear it!

A good 600W PSU will cost over $100, potentially a lot more.

I paid $38.50! There is no way a 600W supply is worth $100. I'd expect a fanless one for that. Where on earth are you sourcing your equipment?

On the contrary, it is NECESSARY to spend that much to get a 600W
PSU. Ink on a label, text on a website is essentially free, of
course they can sell a 600W for $38 if the only thing "600W"
about it is the labeling.

And I thought we had imported lawsuits from you lot.

$38 will get you a median grade ~ 450W at most, unless you're
buying something as a (salvage, OEM closeout not offered though
regular channels, etc). Even sale prices would be necessary to
get the 450W for $38. Do not be confused that a fancy casing,
neon glow LED fans or a cable sheath make a decent PSU... those
parts cost a couple dollars, ten at most and added to a PSU that
in it's plain-clothes version would sell for $10 less, they've
priced it appropriately higher as well.

It's on www.ebuyer.co.uk - I've not seen them that cheap anywhere else - that company has lots of low price stuff - a RAID card for a tenner ffs!

As for a fanless 600W PSU, capable of lasting more than a couple
months outputting 600W (if that long, before failing), you would
have to spend far more than $200... probably several thousand
because there are no passively cooled PC PSU that can manage
600W. It is pushing the limits of PS/2 casing to even do that
with a fan(s). Servers do not use much larger PSU just for the
heck of it... some are twice the volume and not from multiple
fans.

I'm not intending to get a 600W fanless - I'm looking at a 300W and a 420W - I've never seen a 600W fanless!

If
yours cost less you might be better off assuming it's only worth
(roughly) 320W, if that. Running such a PSU on higher demand
system could damage PSU and connected parts. When a PSU is
undersized for it's rated wattage it may deviate far more from
spec to compensate, regardless if it's claimed spec. If you
tested it and found it not within specs at rated wattage (which
is likely), what is your recourse?

Get a refund of course. But as it is within 2% at 200 watts, I would assme it would be fine at 600. It's doing the job 5 times better than the suntek did, so I'm happy. I won't ever be drawing 600 watts.


Who said you'd be entitled to a refund? Check your warranty
statements, if you have one (otherwise you may be SOL) and see if
there is a money-back guarantee.

That's irrelevant, I'm going by LAW.

Again, there is no reason to assume a PSU working fine at 200W
can do so at 600W. You have not mentioned how you are sure it's
"5 times better" than the suntek,

I did. 2% voltage inaccuracy instead of 10%.

but as I mentioned previously
Sparkle makes multiple grades of PSU, the "suntek" is not their
high-end AFAIK.

Suntek is a sparkle?

One thing to remember about wattage is that
beyond a unit's capable wattage, suddenly ripple may be much
worse. If the suntek was just beyond that output level, it's not
surprising if it's output wasn't so clean.

It was at two thirds!

Likewise, once your
generic is past it's max point, it too will have to drop voltage
every time it's amperage capacity is exceeded, else it'll shut
off, (or both).

What would happen if I connected all the rails of TWO PSUs in parallel? I'll ask you first rather than get a double firework display :-) Just wondering if one would try to keep at 5.05 volts and one at 4.95 volts and both fight like crazy or something.

With enough effort you might
get it replaced with same thing, but still you dont' end up with
a PSU capable of sustaining rated wattage at rated tolerance.

No, I'd get my money back or a voucher for a different supply that worked correctly. You come from the land of sueing people for slightly squint paving stones - you should know this.

You'd get money back because you "say so"?
From who?
I don't sell junk PSU, but if I did, after the (clearly
specified, short reseller period is over, typically around 30
days or less) you wouldn't get a refund from me or any other
major reseller. You would be referred to manufacturer, who is
also not obligated to money back if not specified in the
warranty.

Keep in mind that your laws might offer better protection in the
UK, this is how it is in the US. Even so, I question whether
your connected equipment would be covered, it could be a long and
costly exercise "proving" the PSU caused equipment damage.

1 year standard - up to SIX yearss if the item should last that long. I just don;'t understand how you put up with that 30 day rubbish. It doesn't even cover early deaths. You need a year to be sure you've got well into the bathtub curve or whatever it is.

If you had a picture of label and/or entire open top of the
failed PSU we might be able to make positive ID.

For a post mortem? ;-)

http://80.229.155.158/temp/psulabel.jpg

I have not seen inside that PSU, cannot speculate about it's
exact design but do note that it's not same labeled specs as most
300W Fortron/Sparkles.



--
*****TWO BABY CONURES***** 15 parrots and increasing http://www.petersparrots.com
93 silly video clips http://www.insanevideoclips.com
1259 digital photos http://www.petersphotos.com
Served from a pentawatercooled dual silent Athlon 2.8 with terrabyte raid

Birthdays are good for you. Statistics show that the people who have the most live the longest. (Rev. Larry Lorenzoni)

On Mon, 26 Jul 2004 01:13:54 +0100, "Peter Hucker"
wrote:


Who said you'd be entitled to a refund? Check your warranty
statements, if you have one (otherwise you may be SOL) and see if
there is a money-back guarantee.

That's irrelevant, I'm going by LAW.

Perhaps, I can't speculate what effort you'd have to go though,
but I find it likely that it would be about principle, not $38,
as your time spent collecting would easily be worth more than
$38.



Again, there is no reason to assume a PSU working fine at 200W
can do so at 600W. You have not mentioned how you are sure it's
"5 times better" than the suntek,

I did. 2% voltage inaccuracy instead of 10%.

I meant, how you're sure the 2% and 10% are accurate. It is no
great feat for an ATX PSU to be within 5%, I wonder if the 10%
was merely a recant of Intel's guideline, which is for max of 10%
on 12V rail. It's VERY rare for a PSU that is working "properly"
per it's design, and not running a load beyond it's capacity, to
be off by anywhere near 10%.


but as I mentioned previously
Sparkle makes multiple grades of PSU, the "suntek" is not their
high-end AFAIK.

Suntek is a sparkle?

Based on the two points of evidence presented so far it seems
likely, but as mentioned I have not seen that PSU in person. All
of the PSU on their website (http://www.suntekgroup.com/atxa.htm)
looks like old designs that correspond to same wattage Sparkles,
nor do any of their pictures look like Sparkles but pictures
might be near a decade old, it is rather odd for a manufacturer's
highest rated (300W) offering to have only 140W 3V+5V combined
rating, that was about right for the Pentium I era.


One thing to remember about wattage is that
beyond a unit's capable wattage, suddenly ripple may be much
worse. If the suntek was just beyond that output level, it's not
surprising if it's output wasn't so clean.

It was at two thirds!

I can't speculate further about it without knowing exactly what
you had, since I don't recall ever running across that specific
model, or if I did it was long enough ago that I don't recall...
could easily have same thing in a pile but wearing a different
label, but can't know this for sure. It could've been very old
or margin for it's rated wattage too. Today's top Sparke/Fortron
offerings are better than 10% tolernace, likely nearly same
capacity as the "600W" one you have.



Likewise, once your
generic is past it's max point, it too will have to drop voltage
every time it's amperage capacity is exceeded, else it'll shut
off, (or both).

What would happen if I connected all the rails of TWO PSUs in parallel? I'll ask you first rather than get a double firework display :-) Just wondering if one would try to keep at 5.05 volts and one at 4.95 volts and both fight like crazy or something.

I'd expect the one with the higher voltage to keep running but
the other one to shut down, due to it trying to reduce voltage to
what it "though" was right. I could be wrong though, I don't
parallel PSU outputs and if I were to try I'd probalby use
isolation diodes.



Keep in mind that your laws might offer better protection in the
UK, this is how it is in the US. Even so, I question whether
your connected equipment would be covered, it could be a long and
costly exercise "proving" the PSU caused equipment damage.

1 year standard - up to SIX yearss if the item should last that long. I just don;'t understand how you put up with that 30 day rubbish. It doesn't even cover early deaths. You need a year to be sure you've got well into the bathtub curve or whatever it is.

.... when that's the way it's been all along you operate
accordingly. Although, when I buy equipment I typically buy as
If I had no warranty at all except non-DOA, as I'd rather pop
open a PSU, voiding it's warranty, to be certain nothing is
wrong... rarely a capacitor might be mounted backwards, a load
resistor resting against a capacitor, or just making sure there
is some filtration or minimal surge protection on the AC side
before going to the trouble of installing it.

On Mon, 26 Jul 2004 01:33:52 GMT, kony wrote:

On Mon, 26 Jul 2004 01:13:54 +0100, "Peter Hucker"
wrote:


Who said you'd be entitled to a refund? Check your warranty
statements, if you have one (otherwise you may be SOL) and see if
there is a money-back guarantee.

That's irrelevant, I'm going by LAW.

Perhaps, I can't speculate what effort you'd have to go though,
but I find it likely that it would be about principle, not $38,
as your time spent collecting would easily be worth more than
$38.

What is this with people liking time to money? If I had an infinite supply of paid jobs, that would be true.

Again, there is no reason to assume a PSU working fine at 200W
can do so at 600W. You have not mentioned how you are sure it's
"5 times better" than the suntek,

I did. 2% voltage inaccuracy instead of 10%.

I meant, how you're sure the 2% and 10% are accurate. It is no
great feat for an ATX PSU to be within 5%, I wonder if the 10%
was merely a recant of Intel's guideline, which is for max of 10%
on 12V rail. It's VERY rare for a PSU that is working "properly"
per it's design, and not running a load beyond it's capacity, to
be off by anywhere near 10%.

10% was on the 3.3 and 5 volt positive rails (it was under).

but as I mentioned previously
Sparkle makes multiple grades of PSU, the "suntek" is not their
high-end AFAIK.

Suntek is a sparkle?

Based on the two points of evidence presented so far it seems
likely, but as mentioned I have not seen that PSU in person. All
of the PSU on their website (http://www.suntekgroup.com/atxa.htm)
looks like old designs that correspond to same wattage Sparkles,
nor do any of their pictures look like Sparkles but pictures
might be near a decade old, it is rather odd for a manufacturer's
highest rated (300W) offering to have only 140W 3V+5V combined
rating, that was about right for the Pentium I era.

This is a supply that came with a full tower case I bought about 2 years ago. The case is no longer manufactured.

This wattage is getting insane - in 10 years computers will consume a kilowatt at this rate.

Likewise, once your
generic is past it's max point, it too will have to drop voltage
every time it's amperage capacity is exceeded, else it'll shut
off, (or both).

What would happen if I connected all the rails of TWO PSUs in parallel? I'll ask you first rather than get a double firework display :-) Just wondering if one would try to keep at 5.05 volts and one at 4.95 volts and both fight like crazy or something.

I'd expect the one with the higher voltage to keep running but
the other one to shut down, due to it trying to reduce voltage to
what it "though" was right. I could be wrong though, I don't
parallel PSU outputs and if I were to try I'd probalby use
isolation diodes.

Wouldn't you lower the voltages too much with diodes?




--
*****TWO BABY CONURES***** 15 parrots and increasing http://www.petersparrots.com
93 silly video clips http://www.insanevideoclips.com
1259 digital photos http://www.petersphotos.com
Served from a pentawatercooled dual silent Athlon 2.8 with terrabyte raid

The evening news is where they begin with "Good evening", and then proceed to tell you why it isn't.

On Mon, 26 Jul 2004 03:17:19 +0100, "Peter Hucker"
wrote:

On Mon, 26 Jul 2004 01:33:52 GMT, kony wrote:

On Mon, 26 Jul 2004 01:13:54 +0100, "Peter Hucker"
wrote:


Who said you'd be entitled to a refund? Check your warranty
statements, if you have one (otherwise you may be SOL) and see if
there is a money-back guarantee.

That's irrelevant, I'm going by LAW.

Perhaps, I can't speculate what effort you'd have to go though,
but I find it likely that it would be about principle, not $38,
as your time spent collecting would easily be worth more than
$38.

What is this with people liking time to money? If I had an infinite supply of paid jobs, that would be true.

Even without a supply of paid jobs, is not your free time worth
something... I mean, if you had work and work and then more work
with no time to do anything else... Sure it can seem like that
happens to most of us, but that would make free time ever more
valuable, not less so, wouldn't it?


I meant, how you're sure the 2% and 10% are accurate. It is no
great feat for an ATX PSU to be within 5%, I wonder if the 10%
was merely a recant of Intel's guideline, which is for max of 10%
on 12V rail. It's VERY rare for a PSU that is working "properly"
per it's design, and not running a load beyond it's capacity, to
be off by anywhere near 10%.

10% was on the 3.3 and 5 volt positive rails (it was under).

Sounds like it was not a matter of tolerance at all, rather that
system was too great a load on that (combined 3V+5V) rail... as I
mentioned in the previous email the models on Sunteks website
look like very old designs that had atypically low 3V+5V ratings.
Today's Sparkle/Fortrom 300W is much hardier in that respect.


but as I mentioned previously
Sparkle makes multiple grades of PSU, the "suntek" is not their
high-end AFAIK.

Suntek is a sparkle?

Based on the two points of evidence presented so far it seems
likely, but as mentioned I have not seen that PSU in person. All
of the PSU on their website (http://www.suntekgroup.com/atxa.htm)
looks like old designs that correspond to same wattage Sparkles,
nor do any of their pictures look like Sparkles but pictures
might be near a decade old, it is rather odd for a manufacturer's
highest rated (300W) offering to have only 140W 3V+5V combined
rating, that was about right for the Pentium I era.

This is a supply that came with a full tower case I bought about 2 years ago. The case is no longer manufactured.

Ahh, now I"m beginning to understand, Suntek simply contracted
out the cheapst PSU that could get to include with cases and
Sparkle being a major manufacturer, could do it cost-effectively.
Not suprisingly, many generic labels are made by large
manufacturers, yet I don't consider them "name-brand", as the
manufacturer's name isn't on the label. Switching power supplies
are not difficult for a company to make fairly well, rather the
cost is constrained.


This wattage is getting insane - in 10 years computers will consume a kilowatt at this rate.

We see a practical limit with air-cooling of CPU/GPU and so Intel
and AMD now develop dual-core CPUs. Your memory is more of an
anomoly, usually memory is only mildly warm. It might be more of
a motherboard "issue" than anything else.


Likewise, once your
generic is past it's max point, it too will have to drop voltage
every time it's amperage capacity is exceeded, else it'll shut
off, (or both).

What would happen if I connected all the rails of TWO PSUs in parallel? I'll ask you first rather than get a double firework display :-) Just wondering if one would try to keep at 5.05 volts and one at 4.95 volts and both fight like crazy or something.

I'd expect the one with the higher voltage to keep running but
the other one to shut down, due to it trying to reduce voltage to
what it "though" was right. I could be wrong though, I don't
parallel PSU outputs and if I were to try I'd probalby use
isolation diodes.

Wouldn't you lower the voltages too much with diodes?

Yes there would be voltage drop, but too much is relative only to
initial voltage, and heat generation (waste). Many PSU have
voltage adjustment POTs. Even so, if a redundant current sharing
PSU is desired they are of course available ready-made.

"Peter Hucker" wrote in message news:opsbpeo6tdaiowgp@blue...
On 25 Jul 2004 00:17:43 -0700, do_not_spam_me wrote:

while the heatsink attached to the low voltage components
poses no shock hazard, my concern here was over the heatsink
attached to the high voltage transistors because even if
it measures 0V, as I believe it does in most Fortron-Source
supplies, it's not safe to water cool it or allow it to
have any chassis or external contact. This is because its
electrical insulation is very thin, and good safety practices
call for using much better insulation for such high voltages.

You have made a mistake. You are assuming I am safe.

I don't understand what you mean, but I assume you want to be safe.

I got it for a fiver!

And now it's worth nothing. You should have experimented on
one of your worse supplies because a 300W Fortron-Source
will outperform many 350W and even some 400W units.

Worse? That WAS the worse one. I didn't want to break my gold
plated dual fan 600 watt thing. The 300W one I broke was
absolutely terrible - 10% voltage accuracy?! **** that. This
600 watt one is 2%. And what's Fortron-Source? A make or a design?

Fortron-Source, www.fsp-group.com, is one of the largest makers of
power supplies and sells its products under several brand names,
including Fortron, Sparkle, Powerman, and Trend, and they're also used
by many large PC makers. Most but not all FSP products have model
numbers that start with "FSP". They also provide the basic boards
used by other power supply companies, including PC Power & Cooling
(not all -- some others are from Zippy), Q-Technology (not to be
confused with Q-Tec), and Zalman. In tests for maximum output
capacity, Fortron-based supplies have tended to perform among the
best, so I don't see how your 300W Fortron's output fluctuated by 10%
with just a 200W load (probably an accurate estimate). In contrast,
that "600W" gold plated supply is probably inferior to a 400W Fortron
or 380W or 430W Antec Truepower, and very few high quality supplies
exist that are both gold plated and contain lighted fans.

I found one hot inductor/coil/transformer/whatever it is
(just two wires to it) that was warm, I put a passive
heatsinnk on that and it kept it at less than doby temperature.

http://80.229.155.158/temp/coil.jpg

This choke is a high voltage device and probably located
next to high voltage components that aren't completely
insulated, so be very careful in installing a heatsink on it.

It seemed pretty insulated to me, and I never even got a tickle off it.

Not the PFC choke itself but the components near it, such as the
bridge rectifier, transistors, line filter chokes, and fuse, all which
can have exposed high voltage on their leads. The thin insulation for
the transistors or on the large capacitors can fail and shouldn't be
trusted.

The thing was very hot without a heatsink, and cool with one,
so I assumed that the heat was getting to the oiutre casing just fine.

Unlikely. The steel laminated core doesn't conduct heat very well,
and its steel or aluminum cover probably doesn't fit tightly against
it. The choke's DC resistance is roughly 20 ohms, making it dissipate
roughly 35W with a 200W load.