Which PSU would you buy?

  • Thread starter Thread starter toronado455
  • Start date Start date
Bob said:
Neither of these would make it onto my short list.

Except for their 80mm fans, which could be loud, what's wrong with
them? What do you suggest instead?
 
This 620W Enermax has only a 12cm fan and appears to perform well:

http://www.extremeoverclocking.com/reviews/cases/Enermax_ELT620AWT_6.html


For years Enermax has overrated PSU to a point I would
consider fraud. IMO, they should have all products pulled
from the market and be prosecuted. Their products may not
be as bad as "some" generics, but if we were to draw a line
in the sand, it would have to at least intersect between PSU
that can sustain their rated wattage long-term in real use,
and those that can't. Enermax can't.

It's quite likely their so-called 620W PSU is capable of no
more wattage than the other 400W PSU I'd mentioned, rather
it only costs a lot more. That is a shame, they still think
they can benefit from a few positive reviews from the
Pentium 3 days when systems used under 200W, when just about
any generic PSU without defective parts would've sufficied.

I'll even go so far as to say it's absolutely ludicrous that
they're charging so much for that PSU. The only thing
almost as bad is when a review tests the PSU for a few days
or less then proclaims it's "good". Would you conside a PSU
"good" if it ran ok for 10 days then failed? Perhaps 10
days is short-sighted, but most psu that test ok at full
load in reviews could not run like that for several months,
let alone the several year life expectancy of a system.

Bottom line- ALL of Enermax PSU, every last single one, is a
bad choice compared to other alternatives at same price
point.
 
kony said:
On 17 May 2006 01:11:55 -0700, (e-mail address removed)

For years Enermax has overrated PSU to a point I would
consider fraud. IMO, they should have all products pulled
from the market and be prosecuted. Their products may not
be as bad as "some" generics, but if we were to draw a line
in the sand, it would have to at least intersect between PSU
that can sustain their rated wattage long-term in real use,
and those that can't. Enermax can't.
It's quite likely their so-called 620W PSU is capable of no
more wattage than the other 400W PSU I'd mentioned, rather
it only costs a lot more. That is a shame, they still think
they can benefit from a few positive reviews from the
Pentium 3 days when systems used under 200W, when just about
any generic PSU without defective parts would've sufficied.

I'll even go so far as to say it's absolutely ludicrous that
they're charging so much for that PSU.

But the extra cost buys extra style, an important factor to a former
nude White House shill reporter like me, and Enermax includes this
fashionable pouch for the power cables:

http://tinyurl.com/r5ewo
http://tinyurl.com/rsn9f
The only thing almost as bad is when a review tests the PSU for a few days
or less then proclaims it's "good". Would you conside a PSU
"good" if it ran ok for 10 days then failed? Perhaps 10
days is short-sighted, but most psu that test ok at full
load in reviews could not run like that for several months,
let alone the several year life expectancy of a system.

Bottom line- ALL of Enermax PSU, every last single one, is a
bad choice compared to other alternatives at same price
point.

Here are results for a 600W Enermax that passed a 6-week full power
test:

www.tomshardware.com/2005/07/11/stress_test/page15.html

I would love to see more long-term durability tests, but it's hard
enough to find reviews that run the PSes at full rated power for even
the briefest period.
 
Here are results for a 600W Enermax that passed a 6-week full power
test:

www.tomshardware.com/2005/07/11/stress_test/page15.html

I would love to see more long-term durability tests, but it's hard
enough to find reviews that run the PSes at full rated power for even
the briefest period.


6 weeks is something, in that many overrated PSU wouldn't
last that long. However, it is not a dynamic load. Modern
systems not only use a lot of 12V amps, they have very large
changes in 12V current, constantly.

There is another testing flaw though, that they are not
constrained in air intake by a chassis so the rate is higher
(very few cases/systems have more intake than exhaust
potential), and that intake air is pre-heated. We know for
certain than any psu must be derated at higher ambient or
lower airflow so a test for X watts must validate in a more
realistic environment.

That doesn't necessarily mean the Enermax ~600W wouldn't be
fine in a power hungry system, but when one is paying extra
for "600" W they should be getting that, in use long term
which has to be the context as it is the known lone use of a
PSU as marketed.
 
larry said:
Those are two of the best brands, but Fortron-Source is probably a bit
better, and because it's $18 cheaper I'd definitely go with it.

I've understand that Fortron is good if it has FSP in the model number.
The one I posted has ATX instead of FSP in the model number. Are those
just as good?

But
80mm fans tend to be louder than 120mm ones, so you may want to
consider this FSP 120mm model for $41.50:

www.newegg.com/Product/Product.asp?Item=N82E16817104953

I don't want to put 12cm fan PSUs in some of my systems because of the
close proximity to the CPU HSF at the top of the MB less than 1/2" away
from the bottom of the PSU which will obstruct airflow to a good
portion of the PSU. However I'm thinking about building a system with a
taller case which will have a larger gap between the CPU and the PSU in
which I could use a 12cm fan PSU. But for the system in question right
now, I need one with straight-through airflow from front to back with
no fan or vents at all in the bottom. (Also because I want to be able
to make a duct for air intake from the front of the case using the top
two 5.25" bays in the future.)

Are all the 80mm fan PSUs loud? I thought some of them were quiet. I
know the NeoHE is 80mm. But I'd like to spend a little less than that.
I don't want a loud PSU. It doesn't have to be as quiet as a 120mm but
it should still be fairly quiet.
 
toronado455 said:
I've understand that Fortron is good if it has FSP in the model number.
The one I posted has ATX instead of FSP in the model number. Are those
just as good?

The couple of 300W and 350W ATX-xxxGT and FSP-xxxnn models I checked
were identical inside, except one of the FSPs had unmarked grey
transistors that I couldn't look up.
Are all the 80mm fan PSUs loud? I thought some of them were quiet. I
know the NeoHE is 80mm. But I'd like to spend a little less than that.
I don't want a loud PSU. It doesn't have to be as quiet as a 120mm but
it should still be fairly quiet.

My 80mm PSUs have ranged from very quiet (Antec 300W) to loud (350W
Enermax), but the NeoHEs are reputed to be very quiet. But if you buy
one, be sure it's revision A3 or later because earlier designs had
problems with the power-saving of certain mobos and CPUs.
 
The couple of 300W and 350W ATX-xxxGT and FSP-xxxnn models I checked
were identical inside, except one of the FSPs had unmarked grey
transistors that I couldn't look up.


?

Transistors where?
Some are kinda important, like in the chopper. Some are
kinda unimportant, like the small-signal TO92 that might be
scattered about.


This is a large issue that depends a lot on the user.

IF your system actually needs a fairly high wattage output,
you should not want a "quiet" PSU. Why? Because although
PSu efficiency can change, some, overall the temps of all
decent units directly correspond to the airflow, and airflow
corresponds to noise. A 120mm fan can reduce the noise:flow
ratio, but it also inherantly reduces the size of crucial
components like the heatsinks, transformer, and causes a
re-arrangement of several other parts, if not complete
abandoment of more sophisticated subcircuits. One can build
a PSU that is more or less clean, more or less long-lasting,
and in a short/brief test, the difference may not be
realized.

Occasionally I write "If your (motherboard, etc) fails, will
you blame the right part?". I do so because there is a
horrible disparity between cost and PSU fitness. I've
pulled aux PSU out of copy machines that were rated for
100W, that make PC 500W PSU look like crap.

I can't make sweeping indictments of all PSU manufacturers
though, because there is a real need to expand the PSU
beyond the PS2 form-factor. It was spec'd when systems used
under 100W, when even today's really really poor low-end
PSU, would've sufficed.




Keep in mind that noise and lifespan are a tradeoff. Which
do you prefer? Are you capable of replacing parts? A
hardcore guru could make mods to a power supply to improve
it, but is it worth the time and cost? Up until now I'd
felt 10 years was a reasonable lifespan for a PSU, but
current demands keep increasing, and everyone wants quiet,
so a compromise might have to be made, what $$ you are
willing to pay to have lower noise, or time, given enough
time and knowledge to modify parts, many things can be
accomplished but they; 1) Require more than PS2 form
factor 2) Require extensive attention to detail in
modifying the only component in a system that has lethal
voltages and will definitely POP if done "wrong".

My 80mm PSUs have ranged from very quiet (Antec 300W) to loud (350W
Enermax), but the NeoHEs are reputed to be very quiet. But if you buy
one, be sure it's revision A3 or later because earlier designs had
problems with the power-saving of certain mobos and CPUs.

I have nothing against Antec, they are a lot better than
most generics. Even so, unless one is bundled in a case
(for the cost savings of bundling when on sale), I would
never seek an Antec PSU. Their name-recognition simply
results in too much of a premium price if buying one lone
unit at retail pricing.
 
kony said:
?

Transistors where?
Some are kinda important, like in the chopper. Some are
kinda unimportant, like the small-signal TO92 that might be
scattered about.



This is a large issue that depends a lot on the user.

IF your system actually needs a fairly high wattage output,
you should not want a "quiet" PSU. Why? Because although
PSu efficiency can change, some, overall the temps of all
decent units directly correspond to the airflow, and airflow
corresponds to noise. A 120mm fan can reduce the noise:flow
ratio, but it also inherantly reduces the size of crucial
components like the heatsinks, transformer, and causes a
re-arrangement of several other parts, if not complete
abandoment of more sophisticated subcircuits. One can build
a PSU that is more or less clean, more or less long-lasting,
and in a short/brief test, the difference may not be
realized.

Occasionally I write "If your (motherboard, etc) fails, will
you blame the right part?". I do so because there is a
horrible disparity between cost and PSU fitness. I've
pulled aux PSU out of copy machines that were rated for
100W, that make PC 500W PSU look like crap.

I can't make sweeping indictments of all PSU manufacturers
though, because there is a real need to expand the PSU
beyond the PS2 form-factor. It was spec'd when systems used
under 100W, when even today's really really poor low-end
PSU, would've sufficed.





Keep in mind that noise and lifespan are a tradeoff. Which
do you prefer? Are you capable of replacing parts? A
hardcore guru could make mods to a power supply to improve
it, but is it worth the time and cost? Up until now I'd
felt 10 years was a reasonable lifespan for a PSU, but
current demands keep increasing, and everyone wants quiet,
so a compromise might have to be made, what $$ you are
willing to pay to have lower noise, or time, given enough
time and knowledge to modify parts, many things can be
accomplished but they; 1) Require more than PS2 form
factor 2) Require extensive attention to detail in
modifying the only component in a system that has lethal
voltages and will definitely POP if done "wrong".



I have nothing against Antec, they are a lot better than
most generics. Even so, unless one is bundled in a case
(for the cost savings of bundling when on sale), I would
never seek an Antec PSU. Their name-recognition simply
results in too much of a premium price if buying one lone
unit at retail pricing.


Let me try to clarify what I am in the market for in a PSU...

1. It doesn't need to be ultra-quiet. I would like it to be
*reasonably* quiet. If it makes a little white noise that's ok. But it
should not have a discernable pitch.

2. It needs to have straight-through airflow. Meaning air comes in the
front and goes out the back. No vents or fans on the bottom of the PSU
case.

3. Needs to be of reasonably good quality.

4. Needs to cost under $65 (before tax and shipping) but preferably
under $45.

5. Needs to have enough power for a basic, non-overclocked P4 2.4, P4PE
based system. I'm thinking something in the range of 300W-400W rating
is enough so long as it is made by a good quality company that doesn't
have false and misleading ratings like some of them do.
 
larry said:
The couple of 300W and 350W ATX-xxxGT and FSP-xxxnn models I checked
were identical inside, except one of the FSPs had unmarked grey
transistors that I couldn't look up.

The ones with the ATX-xxx model numbers I've seen seem to either have
ATX-xxxPA or ATX-xxxPAF. I've not seen any with "ATX-xxxGT".
 
kony" wrote and wrote and said:
This is a large issue that depends a lot on the user.

IF your system actually needs a fairly high wattage output,
you should not want a "quiet" PSU. Why? Because although
PSu efficiency can change, some, overall the temps of all
decent units directly correspond to the airflow, and airflow
corresponds to noise. A 120mm fan can reduce the noise:flow
ratio, but it also inherantly reduces the size of crucial
components like the heatsinks, transformer, and causes a
re-arrangement of several other parts, if not complete
abandoment of more sophisticated subcircuits. One can build
a PSU that is more or less clean, more or less long-lasting,
and in a short/brief test, the difference may not be
realized.

Occasionally I write "If your (motherboard, etc) fails, will
you blame the right part?". I do so because there is a
horrible disparity between cost and PSU fitness. I've
pulled aux PSU out of copy machines that were rated for
100W, that make PC 500W PSU look like crap.

I can't make sweeping indictments of all PSU manufacturers
though, because there is a real need to expand the PSU
beyond the PS2 form-factor. It was spec'd when systems
used under 100W, when even today's really really poor
low-end PSU, would've sufficed.





Keep in mind that noise and lifespan are a tradeoff. Which
do you prefer? Are you capable of replacing parts? A
hardcore guru could make mods to a power supply to improve
it, but is it worth the time and cost? Up until now I'd
felt 10 years was a reasonable lifespan for a PSU, but
current demands keep increasing, and everyone wants quiet,
so a compromise might have to be made, what $$ you are
willing to pay to have lower noise, or time, given enough
time and knowledge to modify parts, many things can be
accomplished but they; 1) Require more than PS2 form
factor 2) Require extensive attention to detail in
modifying the only component in a system that has lethal
voltages and will definitely POP if done "wrong".



I have nothing against Antec, they are a lot better than
most generics. Even so, unless one is bundled in a case
(for the cost savings of bundling when on sale), I would
never seek an Antec PSU. Their name-recognition simply
results in too much of a premium price if buying one lone
unit at retail pricing.


OK, cut the crap and make some recommendations.

*TimDaniels*
 
kony said:
On 17 May 2006 09:28:51 -0700, (e-mail address removed)
wrote:

6 weeks is something, in that many overrated PSU wouldn't
last that long. However, it is not a dynamic load. Modern
systems not only use a lot of 12V amps, they have very large
changes in 12V current, constantly.

Do any PC PSes perform well on static tests but poorly on dynamic
tests, other than the first versions of the Antec-Seasonic Neo HE?
There is another testing flaw though, that they are not
constrained in air intake by a chassis so the rate is higher
(very few cases/systems have more intake than exhaust
potential), and that intake air is pre-heated. We know for
certain than any psu must be derated at higher ambient or
lower airflow so a test for X watts must validate in a more
realistic environment.

Do you mean you're one of the rare individuals who doesn't keep the
interior of his computer at a constant 25 Celcius? :)
 
Do any PC PSes perform well on static tests but poorly on dynamic
tests, other than the first versions of the Antec-Seasonic Neo HE?

yes, they all do

even some bottom of barrel generic might be able to put out
peak current in ideal (no heat, no airflow restrictions) for
6 weeks

This kind of testing finds flaws but in itself can't
validate a design.

Do you mean you're one of the rare individuals who doesn't keep the
interior of his computer at a constant 25 Celcius? :)


Yes, I aim for lowest noise possible while checking
individual parts' temps. If the chassis rises 10C but my
HDDs, video, etc are cool enough then I would not raise fan
speeds. Of course it depends on the room ambient temp too,
on occasion I've built systems that run in ambient over 25C.
 
OK, cut the crap and make some recommendations.

Troll boy, I'd already mentioned at least once that any
application would have to be considered, yet conspicuously
missing were details of the target system.

This thread was a bit ridiculous, another case of a poster
expecting they can withold the important details, and expect
the correct answer. Even now, the only details emerging are
"basic, non-overclocked P4 2.4". Since those are the only
details supplied thus far, and that the OP seems to be
completely confused about where or why a PSU will have
bottom intake, I'll link two;


http://www.newegg.com/Product/Product.asp?Item=N82E16817104953
http://www.newegg.com/Product/Product.asp?Item=N82E16817104952

Both of these use relatively LQ yate loon fans though, it
will make them quieter than some but they'll need relubed
every couple years or so for best life.

Though if his implications about basic, 2.4GHz are correct,
the best bang for the buck is probably this,
http://www.newegg.com/Product/Product.asp?Item=N82E16817153023

However, since system is a basic 2.4GHz, it may be rather
spartan and economized, and if the goal were to penny pinch
as much as possible, even a 300W PSU might do the job
providing a newer ATX design so it at least has over 12A
(honest/accurately rated) 12V rail. We have no idea though,
if basic might mean by next year he'll put a couple more
hard drives in, so it all comes back to the original issue-
one buys a PSU applicable for the specific system needs, it
is primary criteria and STILL witheld information.

Since Tim likes to start silly arguments, I'll just let him
go at it till he gets tired. There wasn't a lot of point to
this thread anyway, it was an example of how not suppling
the necessary information turns a simple task into a drawn
out waste of time.
 
kony said:
The couple of 300W and 350W ATX-xxxGT and FSP-xxxnn models I checked
were identical inside, except one of the FSPs had unmarked grey transistors
[/QUOTE]
?

Transistors where?
Some are kinda important, like in the chopper. Some are
kinda unimportant, like the small-signal TO92 that might be
scattered about.

They were the NPN, TO-3P (?) power transistors for the chopper.
This is a large issue that depends a lot on the user.

IF your system actually needs a fairly high wattage output,
you should not want a "quiet" PSU. Why? Because although
PSu efficiency can change, some, overall the temps of all
decent units directly correspond to the airflow, and airflow
corresponds to noise.

Still some fans are a lot quieter than others rated for the same CFM
and running at the same RPMs, and replacing fans can noticeably quiet
down some PSUs.
A 120mm fan can reduce the noise:flow ratio, but it also inherantly
reduces the size of crucial components like the heatsinks, transformer,
and causes a re-arrangement of several other parts, if not complete
abandoment of more sophisticated subcircuits. One can build
a PSU that is more or less clean, more or less long-lasting,
and in a short/brief test, the difference may not be realized.

Here: www.jonnyguru.com/PSU/V-Series/all_inside.jpg

are some PSUs with 120mm fans (left to right: 350W, 400W, 500W) but
heatsinks that aren't chopped down. They seem more substantial than
the heatsinks of 120mm PSUs from Seasonic and Fortron. I don't know
how good a job the manufacturer, Wintech, did with the cooling, but
they did bother to stick a sheet of mylar across about 1/3" of the fan
area at the rear to prevent air coming in from being sucked up into the
fan before it flowed through the heatsinks. Jonny Guru said these PSUs
are OK, but I wonder if the main transformer is a bit puny because in
my 350W Ultra/Wintech it's the same size as the ones in my 300W and
350W Antecs, but it runs at 50 KHz instead of 60 KHz.
Occasionally I write "If your (motherboard, etc) fails, will
you blame the right part?". I do so because there is a
horrible disparity between cost and PSU fitness. I've
pulled aux PSU out of copy machines that were rated for
100W, that make PC 500W PSU look like crap.

If you want to see real quality, look at this PSU found inside an
external USB disk drive enclosure:

http://static.flickr.com/51/128289379_5034f765c4_b.jpg

Naturally it's not UL approved. And sitting right above it is a low
voltage circuit board, separated by just a thin sheet of plastic
pressing against the heatsink that holds the high voltage transistors..
I have two of these enclosures. I'm going to return one, and with the
other I substituted a UL approved external PSU.
I have nothing against Antec, they are a lot better than
most generics. Even so, unless one is bundled in a case
(for the cost savings of bundling when on sale), I would
never seek an Antec PSU. Their name-recognition simply
results in too much of a premium price if buying one lone
unit at retail pricing.

Who pays retail? ;) My most expensive Antec was $15, including the
case, and I can always tack a resistor across the fan control
thermistor for better cooling.
 
toronado455 said:
Let me try to clarify what I am in the market for in a PSU...

1. It doesn't need to be ultra-quiet. I would like it to be
*reasonably* quiet. If it makes a little white noise that's ok. But it
should not have a discernable pitch.
2. It needs to have straight-through airflow. Meaning air comes in the
front and goes out the back. No vents or fans on the bottom of the PSU
case.

3. Needs to be of reasonably good quality.
4. Needs to cost under $65 (before tax and shipping) but preferably
under $45.

5. Needs to have enough power for a basic, non-overclocked P4 2.4, P4PE
based system. I'm thinking something in the range of 300W-400W rating
is enough so long as it is made by a good quality company that doesn't
have false and misleading ratings like some of them do.

One of those 350-400W Fortron/Sparkle/Hi-Q/PowerQ/Trend PSUs with an
80mm fan will do just fine, like that model ATX400-PA. It's one of the
best-made PSUs available at any price.
 
Back
Top