What PSU rating for Athlon 2400 with 6 HDDs?

[Watson A.Name - \Watt Sun, the Dark Remover\]

It's true that an insufficient power supply can cause both
instability and eventual damage, but the typical Athlon
system does not need 400W, and there is almost no "PC"
system that needs 500W, even if the vast majority of the
current were concentrated on only the 5V or 12V rail.

With a typical PC, that being current-gen CPU, a couple hard
drives, budget/low-end video card, etc, 300W PSU in a good
name brand is sufficient. SFF systems demonstrate every day
that even a 180-250W PSU will run a modern built with enough
margin for another hard drive or two... but the PSU may need
replaced sooner.

Generics on the other hand, are a lottery. Their wattage
rating means almost nothing, they can only be assumed to be
somewhere inbetween 200W and 400W without further evidence.

I'm surprised at all the posts that have so many assertions, yet so
little thought and consideration. I don't think I've yet seen an
authoritative URL or two to back up those assertions.

Very little solid advice. Like check the hard disk and see what the
current ratings are, and add them up to get a total. Especially the
+12VDC because the motor takes quite a bit of current from that.

And I really detest those claims that PCs don't use very much power.
The most I can get on a 20A circuit is about a dozen, and the breaker
blows. I figure that's about 150W per PC. And those are old timers,
like P233s.

If you want to play it safe, buy a couple external HD enclosures with
power brick, and use them with a USB 2.0 card. That will relieve the
case and PS of a lot of power and heat. I have a couple of these, and
they get fairly warm, after an hour or so, so they do suck a fair amount
of power.

 

[Chip]

[snip]

Generics on the other hand, are a lottery. Their wattage
rating means almost nothing, they can only be assumed to be
somewhere inbetween 200W and 400W without further evidence.

Absolutely correct. "Qtec" - for example - are a very popular low-end
brand. You might refer to them as a no-name, but arguably they are one step
up from that. They produce a very popular 550w PSU.

The first thing to note is that by 550w, they mean *Peak*. Which basically
means nothing at all. They don't specify for how long this peak is
available, nor under what conditions. Its a bit like saying these pills
will increase your penis length by "up to" 4 inches.

Furthermore, they give no indication at all about what the sustained power
is. But don't worry, 550w is lots isn't it, so surely it can't be a
problem. LOL

So what is the max. current available on the 12v rail on a Qtec 550w? 30A
maybe? 28A? 25A?

No. Its 14A. Read it and weep. 14A.

It wouldn't power my laptop.

Chip

 

[Matt]

I'm surprised at all the posts that have so many assertions, yet so
little thought and consideration. I don't think I've yet seen an
authoritative URL or two to back up those assertions.

Very little solid advice. Like check the hard disk and see what the
current ratings are, and add them up to get a total. Especially the
+12VDC because the motor takes quite a bit of current from that.

 

[Gareth Tuckwell]

[snip]

Absolutely correct. "Qtec" - for example - are a very popular low-end
brand. You might refer to them as a no-name, but arguably they are one
step up from that. They produce a very popular 550w PSU.

The first thing to note is that by 550w, they mean *Peak*. Which
basically means nothing at all. They don't specify for how long this peak
is available, nor under what conditions. Its a bit like saying these
pills will increase your penis length by "up to" 4 inches.

You can get a power supply that will increase your penis length by 4
inches??
I gotta get me one of those!!

Lol!

 

[Michael Brown]

Chip wrote:
[...]

No. Its 14A. Read it and weep. 14A.

It wouldn't power my laptop.

Damn, you must have short battery life on that laptop

 

[Chip]

[snip]

Absolutely correct. "Qtec" - for example - are a very popular low-end
brand. You might refer to them as a no-name, but arguably they are one
step up from that. They produce a very popular 550w PSU.

The first thing to note is that by 550w, they mean *Peak*. Which
basically means nothing at all. They don't specify for how long this
peak is available, nor under what conditions. Its a bit like saying
these pills will increase your penis length by "up to" 4 inches.

You can get a power supply that will increase your penis length by 4
inches??
I gotta get me one of those!!

"Up to" four inches!

Chip

 

[Chip]

Chip wrote:
[...]

No. Its 14A. Read it and weep. 14A.

It wouldn't power my laptop.

Damn, you must have short battery life on that laptop

Bloody right. It has got the power to hibernate.

Chip

 

[Chip]

Chip wrote:
[...]

No. Its 14A. Read it and weep. 14A.

It wouldn't power my laptop.

Damn, you must have short battery life on that laptop

Bloody right. It has got the power to hibernate.

Damned typos.

It *hasn't* got the power to hibernate.

 

[Don Klipstein]

I'm surprised at all the posts that have so many assertions, yet so
little thought and consideration. I don't think I've yet seen an
authoritative URL or two to back up those assertions.

Very little solid advice. Like check the hard disk and see what the
current ratings are, and add them up to get a total. Especially the
+12VDC because the motor takes quite a bit of current from that.

And I really detest those claims that PCs don't use very much power.
The most I can get on a 20A circuit is about a dozen, and the breaker
blows. I figure that's about 150W per PC. And those are old timers,
like P233s.

How many watts, how many volt-amps, and what's the efficiency as well
as the power factor of the power supply?

- Don Klipstein ([email protected])

 

[kony]

I'm surprised at all the posts that have so many assertions, yet so
little thought and consideration. I don't think I've yet seen an
authoritative URL or two to back up those assertions.

Hmmmmm. Claiming little thought and consideration and
mentioning lack of URLs, yet you DON'T PROVIDE ANY YOURSELF.
Seems a bit odd, no?

We all know that usenet is not a place where every statement
is backed up with essays and links... that's what google is
for.

Very little solid advice. Like check the hard disk and see what the
current ratings are, and add them up to get a total. Especially the
+12VDC because the motor takes quite a bit of current from that.

That will tell you little to nothing about the needed
spin-up current.

And I really detest those claims that PCs don't use very much power.
The most I can get on a 20A circuit is about a dozen, and the breaker
blows. I figure that's about 150W per PC. And those are old timers,
like P233s.

1) Do not run a dozen systems on a single 20A circuit then.

2) 150W sounds about right. Typical PSU efficiency is 75%,
so system is using 113W from psu, in line with what might be
expected from a P233 box.

If you want to play it safe, buy a couple external HD enclosures with
power brick, and use them with a USB 2.0 card. That will relieve the
case and PS of a lot of power and heat. I have a couple of these, and
they get fairly warm, after an hour or so, so they do suck a fair amount
of power.

Not too safe for the data is it, running an enclosure
that'll "get fairly warm"?

External enclosures don't necessarily have very good power
supplies either, the drive may be safer in the system
regardless of cooling issues.

It is a good idea to add up current consumption of all
devices, providing the HDD spin-up current is taken into
account, and that it is recognized that not all devices can
possibly consume max current simultaneously. That's where
a lot of online calculators make mistakes, suggesting 500W+
PSU for systems with lower power consumption than many
running fine from 300W PSU. Then someone comes along and
claims their gee-whiz-bang generic that's rated for 400W
couldn't cut it... not because it didn't have high enough
wattage but because the labeled rating was only wishful
thinking.

 

[kony]

[snip]

Generics on the other hand, are a lottery. Their wattage
rating means almost nothing, they can only be assumed to be
somewhere inbetween 200W and 400W without further evidence.

Absolutely correct. "Qtec" - for example - are a very popular low-end
brand. You might refer to them as a no-name, but arguably they are one step
up from that. They produce a very popular 550w PSU.

The first thing to note is that by 550w, they mean *Peak*. Which basically
means nothing at all. They don't specify for how long this peak is
available, nor under what conditions. Its a bit like saying these pills
will increase your penis length by "up to" 4 inches.

Furthermore, they give no indication at all about what the sustained power
is. But don't worry, 550w is lots isn't it, so surely it can't be a
problem. LOL

So what is the max. current available on the 12v rail on a Qtec 550w? 30A
maybe? 28A? 25A?

No. Its 14A. Read it and weep. 14A.

It wouldn't power my laptop.

14A of stable, clean power is enough for most basic systems.
It might not be sufficient for 6 drives PLUS a CPU using 12V
rail for power, nor adding a high-end modern video card, but
it would run any typical mid-grade system fine (if that 14A
were from a decent PSU).

The problem though is how many corners are cut, 14A from a
PSU with undersized inductors and poor filter caps may have
horrible ripple... it also says nothing about how the PSU
behaves when it goes out of spec, what tolerances it's
designed to meet and how quickly it'd shut down if it
failed. Sometimes you really have to dig to get specs for
individual components in a PSU, but most often you don't, if
a corner was cut in one area to save on construction costs
then other corners were obviously cut too, with the end
result being that a PSU worth only 250W, looks like a 250W
PSU inside.

 

[Chip]

[snip]

Generics on the other hand, are a lottery. Their wattage
rating means almost nothing, they can only be assumed to be
somewhere inbetween 200W and 400W without further evidence.

Absolutely correct. "Qtec" - for example - are a very popular low-end
brand. You might refer to them as a no-name, but arguably they are one
step
up from that. They produce a very popular 550w PSU.

The first thing to note is that by 550w, they mean *Peak*. Which
basically
means nothing at all. They don't specify for how long this peak is
available, nor under what conditions. Its a bit like saying these pills
will increase your penis length by "up to" 4 inches.

Furthermore, they give no indication at all about what the sustained power
is. But don't worry, 550w is lots isn't it, so surely it can't be a
problem. LOL

So what is the max. current available on the 12v rail on a Qtec 550w? 30A
maybe? 28A? 25A?

No. Its 14A. Read it and weep. 14A.

It wouldn't power my laptop.

14A of stable, clean power is enough for most basic systems.
It might not be sufficient for 6 drives PLUS a CPU using 12V
rail for power, nor adding a high-end modern video card, but
it would run any typical mid-grade system fine (if that 14A
were from a decent PSU).

Well whoopy do.

I suggest people who buy this crap are wanting a big powerful PSU. That's
why they buy the thing labelled 550W. Otherwise they would be the even more
crap 450W version. And then they get their nice new 550W PSU to run their
powerful graphics card, overclocked CPU and a couple of disks and case fans
and.... they find it doesn't cut it.

The problem though is how many corners are cut, 14A from a
PSU with undersized inductors and poor filter caps may have
horrible ripple...

Yep. I'll bet its crap at that too. BTW I forgot to mention it runs a
"whisper quiet" 34dbA. As compared to perhaps 22 dBA from a Quality PSU.

it also says nothing about how the PSU
behaves when it goes out of spec, what tolerances it's
designed to meet and how quickly it'd shut down if it
failed. Sometimes you really have to dig to get specs for
individual components in a PSU, but most often you don't, if
a corner was cut in one area to save on construction costs
then other corners were obviously cut too, with the end
result being that a PSU worth only 250W, looks like a 250W
PSU inside.

Frankly I think you'd be better off with a 250W Antec than one of these
"550W" Qtec things.

Chip.

 

[Tom MacIntyre]

[snip]

Generics on the other hand, are a lottery. Their wattage
rating means almost nothing, they can only be assumed to be
somewhere inbetween 200W and 400W without further evidence.

Absolutely correct. "Qtec" - for example - are a very popular low-end
brand. You might refer to them as a no-name, but arguably they are one step
up from that. They produce a very popular 550w PSU.

The first thing to note is that by 550w, they mean *Peak*. Which basically
means nothing at all. They don't specify for how long this peak is
available, nor under what conditions. Its a bit like saying these pills
will increase your penis length by "up to" 4 inches.

Furthermore, they give no indication at all about what the sustained power
is. But don't worry, 550w is lots isn't it, so surely it can't be a
problem. LOL

So what is the max. current available on the 12v rail on a Qtec 550w? 30A
maybe? 28A? 25A?

No. Its 14A. Read it and weep. 14A.

It wouldn't power my laptop.

14A of stable, clean power is enough for most basic systems.
It might not be sufficient for 6 drives PLUS a CPU using 12V
rail for power, nor adding a high-end modern video card, but
it would run any typical mid-grade system fine (if that 14A
were from a decent PSU).

The CPU doesn't need the 12 VDC, does it?

That is part of the problem with PS choice...the consideration must be
made for amperage needs for all voltages, not just total power.

Tom

 

[Franklin]

Nexus relabels Sparkle PSU. If you can find a Sparkle (or
Fortron) 350W for lower price it would be better value.

350W is enough for your system but if motherboard uses 12V
for CPU (one sign of that would be that the board uses the
"Intel" P4 4-pin 12V connector in addition to the ATX 20 pin
connector) then it would provide more margin to choose 400W
or higher (Nexus/Sparkle/Fortron will still be a good choice
in 400W or higher).

Kony, aren't the power requirements (current at certain voltages) of
a mobo fairly fixed? You seem to suggest it is not so.

I had always thought that if I a mobo based on chipset X with a given
Athlon processor (say a T'bred B) then the current at various
voltages required to run this by most mobos on the market is pretty
much the same.

Is there a significant variation in voltages/current required?

Is any such variation mainly due to the design of the mobo? Or is it
mainly due to the choice of components (chipset and processor)?

 

[kony]

The CPU doesn't need the 12 VDC, does it?

That is part of the problem with PS choice...the consideration must be
made for amperage needs for all voltages, not just total power.

Yes, modern P4 or Athlon 64 platforms and many (mostly
nForce2) Athlon XP do use 12V for CPU. The CPU power is
derived from 12V input by step down regulation. Any board
using 12V for CPU power will have the "Intel" 4-pin 12V
connector on it. When there is no 4-pin 12V input the odds
are overwhelming on a "PC" that it's using 5V for CPU power,
as did earlier Athlon & Pentium 3 'boards.

 

[Chip]

Yes, modern P4 or Athlon 64 platforms and many (mostly
nForce2) Athlon XP do use 12V for CPU. The CPU power is
derived from 12V input by step down regulation. Any board
using 12V for CPU power will have the "Intel" 4-pin 12V
connector on it. When there is no 4-pin 12V input the odds
are overwhelming on a "PC" that it's using 5V for CPU power,
as did earlier Athlon & Pentium 3 'boards.

I think you may be mistaken there. I have come across a number of boards
that don't have the 4-pin connector, but still run the CPU off the 12v
supply. The Albatron nf2 board is one that springs to mind.

Chip

 

[kony]

I think you may be mistaken there. I have come across a number of boards
that don't have the 4-pin connector, but still run the CPU off the 12v
supply. The Albatron nf2 board is one that springs to mind.

Nope, then it uses 5V for CPU. What made you think it used
12V? Measure it with a multimeter.

If this is the Albatron you're referring to,
http://www.amdboard.com/km18g-pro.jpg
it clearly is only 2 phase VRM, 5V CPU power.

If you have other boards in mind, supply link to a decent
picture. There are definitely NONE from any remotely
recognizable mainstream motherboard manufacturers, except
those not traditionally considered "PC" boards... something
using EPS 24 pin ATX or another deviation from the 20 pin
ATX connector. The reason for this is that no competent
designer will derive CPU power from a single 12V lead on the
ATX 20-pin wiring harness.

 

[Chip]

Nope, then it uses 5V for CPU. What made you think it used
12V? Measure it with a multimeter.

If this is the Albatron you're referring to,
http://www.amdboard.com/km18g-pro.jpg
it clearly is only 2 phase VRM, 5V CPU power.

No, its not. As soon as I posted I thought perhaps on reflection it was not
their nf2 board.

If you have other boards in mind, supply link to a decent
picture.

I will if I can find it. I specifically remember a review of an Albatron
board where they were questioning the sanity of the designers not including
the P4 connector on a 12V board. Perhaps it was the KT600 or KT880 board, I
can't remember.

There are definitely NONE from any remotely
recognizable mainstream motherboard manufacturers, except
those not traditionally considered "PC" boards... something
using EPS 24 pin ATX or another deviation from the 20 pin
ATX connector. The reason for this is that no competent
designer will derive CPU power from a single 12V lead on the
ATX 20-pin wiring harness.

Not so. See my comment above.

Chip

 

[kony]

Kony, aren't the power requirements (current at certain voltages) of
a mobo fairly fixed? You seem to suggest it is not so.

No, the voltage may be constant but just like with a CPU,
higher frequency causes higher current.

I had always thought that if I a mobo based on chipset X with a given
Athlon processor (say a T'bred B) then the current at various
voltages required to run this by most mobos on the market is pretty
much the same.

I don't understand what you're trying to say.

Is there a significant variation in voltages/current required?

Is any such variation mainly due to the design of the mobo? Or is it
mainly due to the choice of components (chipset and processor)?

Any/all of these can vary current. One board may default
memory to different voltage, another may be running at
higher FSB speed, then the obvious things like chipset or
processor count too... these differences could offset each
other or add up to a signficant difference in some cases.
Even so, the difference between one motherboard and another
will often be much less than the difference between one CPU
or another (if large enough frequency or voltage, core
change), or comparing a budget video card to a high-end
model, or number of hard drives.

 

[Chip]

Nope, then it uses 5V for CPU. What made you think it used
12V? Measure it with a multimeter.

If this is the Albatron you're referring to,
http://www.amdboard.com/km18g-pro.jpg
it clearly is only 2 phase VRM, 5V CPU power.

When you say "clearly", how can you be clear that its 5V CPU power? I can
see how you can deduce whether its two or three phase, but how can you tell
its a 5v design?

Chip