Enermax PSU

[Graeme]

Just wondering whether the above (Enermax Noisetaker 480W) would be
sufficient for;
Gigabyte GA8IPE1000
P4 3GHz
1gig DDR
120gig IDE x2
200gig SATA
DVD ROM
DVD RW
Gforce FX5700Ultra
Floppy
Case fans x2
56k Modem (not used much)
No over-clocking.

I have the above (except the SATA drive) running using a real good quality
(not) Winpower PSU. Just want to make sure that the Enermax will cope (with
the SATA too).
TIA

 

[w_tom]

480 watts is probably just under double the wattage of what
is required.

 

[Graeme]

Well, that's good news. How do you calculate power consumption(s)? Is there
any accurate web-based info?

 

[kony]

Well, that's good news. How do you calculate power consumption(s)? Is there
any accurate web-based info?

Keep in mind that Enermax are generally over-rated so it
starts to balance out when you consider the uneven rail
distribution, that your system will draw most power from 12V
rail. Unfortunately it's very rare for a modern PSU to
have such a large percentage of it's power distributed from
12V rail, generally anyone with a modern P4 will need at
least a 350W PSU else one with atypical rail bias towards
12V power production.

Realistically the power supply would have about 20A on 12V
rail, as a P4 can use 8-12, figure another 4 for video (not
sure about that particular card but that's ballpark for
modern high-end cards), a couple amps for HDD spinup and a
couple for the fans and optical drives. You'd not ever
realize that much current drawn in use, but it's a good
margin. In the past Enermax was rating their units for peak
rather than sustained power, factoring at about 70%, which
is a bit troubling... as 70% of 480W is only 336W.
Fortunately your system could not approach actually using
336W as configured but the 12V capacity bears further
scrutiny. For the $$ one of the better values is a
Sparkle/Fortron in the 500W range, as some have a dual 12V
rail which isolates the CPU supply from the rest of the
system... it's not that you'd need that in a 500W supply but
as with other things you get some desirable features only
with the higher-end models.

 

[Graeme]

Keep in mind that Enermax are generally over-rated so it
starts to balance out when you consider the uneven rail
distribution, that your system will draw most power from 12V
rail. Unfortunately it's very rare for a modern PSU to
have such a large percentage of it's power distributed from
12V rail, generally anyone with a modern P4 will need at
least a 350W PSU else one with atypical rail bias towards
12V power production.

Realistically the power supply would have about 20A on 12V
rail, as a P4 can use 8-12, figure another 4 for video (not
sure about that particular card but that's ballpark for
modern high-end cards), a couple amps for HDD spinup and a
couple for the fans and optical drives. You'd not ever
realize that much current drawn in use, but it's a good
margin. In the past Enermax was rating their units for peak
rather than sustained power, factoring at about 70%, which
is a bit troubling... as 70% of 480W is only 336W.
Fortunately your system could not approach actually using
336W as configured but the 12V capacity bears further
scrutiny. For the $$ one of the better values is a
Sparkle/Fortron in the 500W range, as some have a dual 12V
rail which isolates the CPU supply from the rest of the
system... it's not that you'd need that in a 500W supply but
as with other things you get some desirable features only
with the higher-end models.

The Enermax web site states that their Noisetaker PSUs do have dual 12V
rails for CPU and the other stuff, which was one reason I looked at this
PSU. Max of 32A for 12V, and a max of 18A per 12V rail.
Of course it has SATA power cables, dual fan (80mm and 90mm, hopefully
quieter than what I have at the moment), fan run-on (if that's really
important), and active PFC.
And it costs about £60, so unless it's a rip-off, seems to be reasonably
priced.

 

[w_tom]

Take notice what Kony has posted AND how that differs from
what most people post. He cited specific power requirements
for each voltage. That is where your calculations begin -
numbers for each voltage.

Another problem with saying "XXX watt power supply" is that
many lesser brands so often promoted by naive computer
assemblers do not provide accurate wattages. For example, a
power supply should even be shorted and not be damaged. This
web site tried to measure the actual wattage from power
supplies. Before some could even achieve their rated wattage,
instead, they self destructed. You tell me what the real
wattage is when the supply is so pathetic as to even self
destruct under load:
http://www6.tomshardware.com/howto/02q4/021021/index.html

When selling to those who don't learn electrical basics,
then many power supplies are instead recommended on wattage
and price. As Kony notes, calculate amperage for each
specific output to obtain a workable number.

Now it gets more complex. How do you determine the current
consumed by each voltage? Accurate numbers (the
specifications) are often difficult to obtain. For example,
how many amperes does your motherboard with a CPU require for
each voltage - no other peripherals attached? Do they provide
numbers? Many will simply avoid numbers to sell to the
naive. You may be left to test; taking measurements after the
system is constructed.

Active PFC is nice to have. But it is provided more often
because PFC is required in Europe. Europe has more stringent
power requirements on appliances.

 

[kony]

The Enermax web site states that their Noisetaker PSUs do have dual 12V
rails for CPU and the other stuff, which was one reason I looked at this
PSU. Max of 32A for 12V, and a max of 18A per 12V rail.

Those would be peak values, it cannot supply 18A sustained
on both rails. Perhaps in translation it means 70% of
(22.4A) @ 25C.

Of course it has SATA power cables, dual fan (80mm and 90mm, hopefully
quieter than what I have at the moment), fan run-on (if that's really
important), and active PFC.
And it costs about £60, so unless it's a rip-off, seems to be reasonably
priced.

£60 is $113 USD? Or is the market higher there?
It's somewhat on the high side for value per $ but if you
want an Enermax then it seems appropriate.

 

[Graeme]

Those would be peak values, it cannot supply 18A sustained
on both rails. Perhaps in translation it means 70% of
(22.4A) @ 25C.

I took the above spec to mean that it could supply 32A sustained over both
12V rails, with a max of 18 for any one rail.

£60 is $113 USD? Or is the market higher there?
It's somewhat on the high side for value per $ but if you
want an Enermax then it seems appropriate.

I looked at this PSU because it had dual 12V rails. My current PSU suffers
from a slight voltage drop on 12V when the CPU is under load, and then there
is a noticeable speed drop of the case fans (which makes an audible change
to their tone). This is annoying. The voltage drop is about 300mV, but the
rail still stays with 12V +/-2%.

Are Enermax no good then? What about Thermaltake? I thought they only made
fans. I need something quiet.

Oh, and thanks for all of the comments.

 

[kony]

I took the above spec to mean that it could supply 32A sustained over both
12V rails, with a max of 18 for any one rail.

No that goes back to what I wrote about "peak" ratings for
Enermax. It might supply a peak of 32A for a brief moment,
or 18A peak on one rail.

I looked at this PSU because it had dual 12V rails. My current PSU suffers
from a slight voltage drop on 12V when the CPU is under load, and then there
is a noticeable speed drop of the case fans (which makes an audible change
to their tone). This is annoying. The voltage drop is about 300mV, but the
rail still stays with 12V +/-2%.

Fans plugged into motherboard or power supply leads?
It is typical for a motherboard's 12V rail to drop unless
that CPU power is isolated on the board... some are, maybe
even most are by now. A 300mV drop could be a sign of
nearing full capacity load of the current supply but on 12V
rail it's not necessary so, because the typical PSU still
regulates based on 5V too, so if 5V current remained
constant instead of rising proportionately with 12V, then
12V will indeed dip. Even on those PSU that have dual 12V
rails, it's not actually two independant 12V circuits on any
of them, AFAIK. Rather the voltage produced is a bit above
12V and the rails split after a bridge of two low-ohm
resistors. Regardless, it's most likely better than the
winpower psu, though did you mention it's wattage?
Regardless there probably isn't *any* PC winpower PSU that's
as good as the Enermax... not that the Enermax is "great"
but certainly a step above generics.

Are Enermax no good then?

My primary grip is their overstated capacity... and that
they're priced similar to other makes that DON'T overstate
the capacity.

What about Thermaltake? I thought they only made
fans. I need something quiet.

Thermaltake makes nothing, they just slap their labels on
other products, kinda like Antec does... except that
"almost' all thermaltake fans are crap excepting a few of
the fancy ones on their silent boost heatsink (relabeled
panaflo) and another 'sink whose name escapes me at the
moment. Their PSU are relabeled Sirtecs, and are certainly
a better value but I've not auditioned any of the newer ones
with SATA connectors, the older ~420W series is an
outstanding value at around $40 but it doesn't have such a
bias towards 12V power as some of the others... though
perhaps their higher ~480+ watatge models do, i haven't
bought any Thermaltakes in a while either.

The Fortrons are usually pretty quiet, though any decent
power supply will have thermal fan adjustment such that if
it's getting hot the fan speed will pick up... making it
louder but that's not necessarily a bad thing. In such
situations it could be a sign that the case air intake needs
either higher surface area (open) and/or an intake fan(s).
You might consider models with 12cm fan underneath but that
does take up a bit of room in the PSU, generally the best
~500W power supplies scarsely have room inside for the
components and one fan, let alone multiples. That's another
issue to be aware of, that if the PSU has more than one fan
but isn't physically longer than average, there was some
kind of tradeoff made to allow for the 2nd/3rd/etc fan(s).
Usually it's more than just a shorter/smaller heatsink,
especially on units that have active PFC jammed in there
too. Regular ATX/PS2 form-factors need modernized IMO, the
volume allocated was from a time when systems used 50W.

Frankly I'd go with a Sparkle/Fortron at any price point you
choose, as per $ they seem best built, but most aren't too
pretty if that matters.

 

[Graeme]

No that goes back to what I wrote about "peak" ratings for
Enermax. It might supply a peak of 32A for a brief moment,
or 18A peak on one rail.


Fans plugged into motherboard or power supply leads?
It is typical for a motherboard's 12V rail to drop unless
that CPU power is isolated on the board... some are, maybe
even most are by now. A 300mV drop could be a sign of
nearing full capacity load of the current supply but on 12V
rail it's not necessary so, because the typical PSU still
regulates based on 5V too, so if 5V current remained
constant instead of rising proportionately with 12V, then
12V will indeed dip. Even on those PSU that have dual 12V
rails, it's not actually two independant 12V circuits on any
of them, AFAIK. Rather the voltage produced is a bit above
12V and the rails split after a bridge of two low-ohm
resistors. Regardless, it's most likely better than the
winpower psu, though did you mention it's wattage?
Regardless there probably isn't *any* PC winpower PSU that's
as good as the Enermax... not that the Enermax is "great"
but certainly a step above generics.


My primary grip is their overstated capacity... and that
they're priced similar to other makes that DON'T overstate
the capacity.


Thermaltake makes nothing, they just slap their labels on
other products, kinda like Antec does... except that
"almost' all thermaltake fans are crap excepting a few of
the fancy ones on their silent boost heatsink (relabeled
panaflo) and another 'sink whose name escapes me at the
moment. Their PSU are relabeled Sirtecs, and are certainly
a better value but I've not auditioned any of the newer ones
with SATA connectors, the older ~420W series is an
outstanding value at around $40 but it doesn't have such a
bias towards 12V power as some of the others... though
perhaps their higher ~480+ watatge models do, i haven't
bought any Thermaltakes in a while either.

The Fortrons are usually pretty quiet, though any decent
power supply will have thermal fan adjustment such that if
it's getting hot the fan speed will pick up... making it
louder but that's not necessarily a bad thing. In such
situations it could be a sign that the case air intake needs
either higher surface area (open) and/or an intake fan(s).
You might consider models with 12cm fan underneath but that
does take up a bit of room in the PSU, generally the best
~500W power supplies scarsely have room inside for the
components and one fan, let alone multiples. That's another
issue to be aware of, that if the PSU has more than one fan
but isn't physically longer than average, there was some
kind of tradeoff made to allow for the 2nd/3rd/etc fan(s).
Usually it's more than just a shorter/smaller heatsink,
especially on units that have active PFC jammed in there
too. Regular ATX/PS2 form-factors need modernized IMO, the
volume allocated was from a time when systems used 50W.

Frankly I'd go with a Sparkle/Fortron at any price point you
choose, as per $ they seem best built, but most aren't too
pretty if that matters.

Thank you very much for the comments. You always seem to take the time to
post a detailed response. I'm sure everyone here is grateful to you for
that.

The review posted earlier
(http://www6.tomshardware.com/howto/02q4/021021/index.html) suggested that
most PSUs tested did meet the stated Wattage (even an Enermax which came 7th
out of 21).

 

[kony]

The review posted earlier
(http://www6.tomshardware.com/howto/02q4/021021/index.html) suggested that
most PSUs tested did meet the stated Wattage (even an Enermax which came 7th
out of 21).

True but the Noise Magic relabled Enermax in that review
also failed. On a different review page the excuse was that
they'd limited power output to 250W and thus it failed...
which is quite wrong. limiting power to 250W would cause it
to shut down, NOT fail. If anything a 350W PSU limited to
250W output would be the last one that could possibly be
induced to fail at any load, IF it were properly designed.
The problem is that the Enermax is overdriven to reach it's
stated wattage, a wattage it probably can't sustain
long-term. It highlights a second problem with the unit in
that the important safety circuitry isn't effective if
existing at all.

 

[Graeme]

True but the Noise Magic relabled Enermax in that review
also failed. On a different review page the excuse was that
they'd limited power output to 250W and thus it failed...
which is quite wrong. limiting power to 250W would cause it
to shut down, NOT fail. If anything a 350W PSU limited to
250W output would be the last one that could possibly be
induced to fail at any load, IF it were properly designed.
The problem is that the Enermax is overdriven to reach it's
stated wattage, a wattage it probably can't sustain
long-term. It highlights a second problem with the unit in
that the important safety circuitry isn't effective if
existing at all.

Yes, I didn't like that excuse for failure either. They also state
(somewhere) that the Noise Magic had a replacement fan; presumably this
wasn't sufficient. Still, a PSU shouldn't fail because the fan stops, should
it?!
Again, someone like myself who only reads the specs (not just the 480W
label, but the finer published detail), can see that this PSU is supposed to
have over-voltage and over-current protection. So the reasons for failure
have to be, unless the spec is wildly innaccurate, that Noise Magic tinkered
with the workings of the PSU.

 

[kony]

Yes, I didn't like that excuse for failure either. They also state
(somewhere) that the Noise Magic had a replacement fan; presumably this
wasn't sufficient. Still, a PSU shouldn't fail because the fan stops, should
it?!

It should shut off. Continually forcing a PSU to overheat
and shut off certainly isn't good for it but just allowing
it to continue running wasn't a Noise Magic decision AFAIK,
rather the base Enermax design.

Again, someone like myself who only reads the specs (not just the 480W
label, but the finer published detail), can see that this PSU is supposed to
have over-voltage and over-current protection. So the reasons for failure
have to be, unless the spec is wildly innaccurate, that Noise Magic tinkered
with the workings of the PSU.

.... but it was over-temp, not voltage or current. As far as
voltage and current go, it should've still been within
margins as it was the same base design. What Enermax did is
overspec their unit such that it was very dependant on a
certain level of airflow, ie- low ambient temp. Even the
Enermax spec for their unit is listing an abnormally low 70%
of max power @ 25C for same MTBF rating as other
name-brands.

Most likely Noise Magic made no alternations at all to the
protection circuitry. It is likely they ony swapped the fan
and perhaps a bias resistor in the fan control circuit, as
it effects the fan RPM.

 

[Graeme]

On Fri, 4 Feb 2005 07:43:48 -0000, "Graeme"

Most likely Noise Magic made no alternations at all to the
protection circuitry. It is likely they ony swapped the fan
and perhaps a bias resistor in the fan control circuit, as
it effects the fan RPM.

You'e probably right. I wonder why the Enermax unit didn't fail, and the
Noise Magic one did.
Anyway, thanks for comments/suggestions. Will probably get a Fortron, if I
can find one in stock that has the big 12cm fan.