Are branded PSUs worth it?

  • Thread starter Thread starter Random Person
  • Start date Start date
R

Random Person

Hi guys. As suggested I'm shopping around for a new PSU, probably 500W.
Is it worth getting unbranded PSUs over say an Antec one? The Antec
PSUs seem to be 5x more expensive than an equivalent rated unbranded
one.

Can you risk damaging your hardware with a bad PSU, or is the danger
exaggerated?
 
Random Person said:
Hi guys. As suggested I'm shopping around for a new PSU, probably 500W.
Is it worth getting unbranded PSUs over say an Antec one? The Antec
PSUs seem to be 5x more expensive than an equivalent rated unbranded
one.

Can you risk damaging your hardware with a bad PSU, or is the danger
exaggerated?
Click to expand...

a $5M Q ............I went branded

Just read a few Google reports on their electrical capabilities and looked
at them physically.

I have to say the branded units did LOOK like they were worth the extra

but who knows............time can only tell
 
Random Person said:
Hi guys. As suggested I'm shopping around for a new PSU, probably 500W.
Is it worth getting unbranded PSUs over say an Antec one? The Antec
PSUs seem to be 5x more expensive than an equivalent rated unbranded
one.

Can you risk damaging your hardware with a bad PSU, or is the danger
exaggerated?
Click to expand...

There's very little risk of damaging hardware by using a inadequate PSU.
The likely problem is that you'll get instability problems and random
reboots.
I went for a generic 400w PSU and it did okay but only lasted a little over
12 months.
It was driving 2 HDDs, a DVD ROM and CDRW plus the usual stuff plugged into
the motherboard.
I splashed out on a branded (but not Antec) 450w PSU and its twice as heavy
and delivers a higher ampage - its lasted
as long as the generic one and is still going strong. The generic cost £20 -
the branded cost £45 - it was worth it.
 
Random said:
Hi guys. As suggested I'm shopping around for a new PSU, probably 500W.
Is it worth getting unbranded PSUs over say an Antec one? The Antec
PSUs seem to be 5x more expensive than an equivalent rated unbranded
one.

Can you risk damaging your hardware with a bad PSU, or is the danger
exaggerated?
Click to expand...
Usually, a cheap PSU of 500 Watt is worse than a brand PSU of 350. The
cheap one often has voltages which way too far off what they should be
(so the 12 volt line at 11 or 13), which is either bad for your
hardware, or causes reboots.

So yes, it is worth it.

Marc
 
Hi Sleepy. Thanks for your input.

I'm curious, what happens when a generic PSU dies on you?

When you say there is "...little risk of damaging hardware by using a
inadequate PSU.", do you mean inadequate in terms of wattage or
quality?
 
Hiya Marc. If I remember the electrical part of my college Physics,
isn't it just a matter of getting the number of coils on the
transformers correct? How hard can it be for a PSU to supply a correct
output voltage?

What other issues am I missing?
 
Hi guys. As suggested I'm shopping around for a new PSU, probably 500W.
Is it worth getting unbranded PSUs over say an Antec one?
Click to expand...

Not if you need 500W. "Watts" do not mean same thing with
generics as they tend to be rated on some imaginary
instantaneous current rather than sustainable current. If
the 500W name-brand were rated the same way, it might be
called a 700W or more.
The Antec
PSUs seem to be 5x more expensive than an equivalent rated unbranded
one.
Click to expand...

Then you're only looking at some pretty bad generics, even
the mid-grade generics are no less than 40% of the cost of
the Antec.
Can you risk damaging your hardware with a bad PSU, or is the danger
exaggerated?
Click to expand...

Yes, they may lack tight safety shutdown circuitry and
insufficient capacitors that cause rest of system to bear
the burder of what the power supply was "supposed" to be
doing. That wears out parts, usually not immediately but
progressively.

Why did you need to ask? Clearly if a $15-25 generic was
suitable then nobody else, including OEMs, would be using
better PSU.
 
Little relationship exists between output power and
quality. As Kony noted, the cheap power supplies will even
claim power ratings that they don't really output. They did
not lie. They measured something different from what you
needed to know. But then they are selling to customers who
look only at dollars and watts; who think they are computer
savvy because they assembled a machine. A power supply must
perform many functions.

Find the Intel ATX power supply specs. Power supplies must
contain and perform numerous functions. For example, if a
function is missing, then a failed power supply can damage the
entire computer. No power supply even 30 years ago would do
that if properly designed. However overseas manufacturers
have discovered a ripe market for higher profits. They forget
to include these many defacto standard functions, forget to
provide the long list of specs, and sell the $35 supply to
North Americans who have no technical knowledge; ripe for a
scam.

If the computer power supply is not selling for about $65
retail, then you know it must be missing essential functions.
That does not say a $65 power supply does contain those
functions. But far more happens inside a power supply than
the OP realizes. The Intel ATX Power Supply spec lists
numerous required functions that $25 power supplies must
forget to include.

Provide no long list of specifications to sell inferior
supplies. Perfect way to market to bean counter types who
fear simple technology concepts. Many North Americans have
bean counter training and therefore declare themselves as
'computer literate'. Ironic. The term 'computer literate'
was coined by an executive who could not even use a computer.
 
Hi guys. As suggested I'm shopping around for a new PSU, probably 500W.
Is it worth getting unbranded PSUs over say an Antec one? The Antec
PSUs seem to be 5x more expensive than an equivalent rated unbranded
one.

Can you risk damaging your hardware with a bad PSU, or is the danger
exaggerated?
Click to expand...

a bad psu can fry a system, an underpowered one can make it unstable,
unstable is preffered, but only till it becomes annoying. if the generic
supply manufacturer has good reviews from enough people who know what
they're talking about it may be good enough. i had myself a supply, rock
solid for about 2 and a half years, then all of a sudden starting the
system and keeping it running became an adventure. i pulled the supply
open(which could have killed me if i didn't know what i was doing), it was
nice and clean, which surprised the hell outta me, but a pair of
regulators had desoldered themselves due to thermal overload. from that i
learned that i should get reputable oem when i want a good supply.
maybe there's an antec solutions supply that you can dig up which will do
well enough, it's their oem, and lots cheaper than the more well know of
their supplies. although i've had troubles with my antec supply, i was
able to fix it by soldering on a new connector.
the supply is important not just to keep things running, but to keep them
running well, even somewhat flaky hardware will keep haulin if it's got
real clean power.
 
Hiya Marc. If I remember the electrical part of my college Physics,
isn't it just a matter of getting the number of coils on the
transformers correct? How hard can it be for a PSU to supply a correct
output voltage?

What other issues am I missing?
Click to expand...


You are missing that "number of coils" is only valid at a
narrow range of power output, it varies quite a lot
depending on the load. For example, grab a cheap wall-wart
(brick AC-DC adapter) of the typical
transformer->rectifier->capacitor type design (rather than a
switching type design) and measure it's output while under
different loads- and compare to it's rated output.
 
Random said:
Hiya Marc. If I remember the electrical part of my college Physics,
isn't it just a matter of getting the number of coils on the
transformers correct? How hard can it be for a PSU to supply a correct
output voltage?

What other issues am I missing?
Click to expand...
I have a branded one, and even that one never delivers the exact voltage
it should, and from stories at tweakers.net (dutch ICT site), I hear,
every time someone asks, that generic ones are even worse.
 
Random Person said:
Hi guys. As suggested I'm shopping around for a new PSU, probably 500W.
Is it worth getting unbranded PSUs over say an Antec one?
Click to expand...

Is the sky blue? Is the pope a Catholic? You get the picture.
The Antec PSUs seem to be 5x more expensive than an equivalent rated
unbranded
one.
Click to expand...

What are the actual prices? 5x X is not significant if X is small.
 
w_tom wrote:

If the computer power supply is not selling for about $65
retail, then you know it must be missing essential functions.
That does not say a $65 power supply does contain those
functions. But far more happens inside a power supply than
the OP realizes. The Intel ATX Power Supply spec lists
numerous required functions that $25 power supplies must
forget to include.
Click to expand...

Apparently Hiper are one make that sell good quality power supplies,
but are still cheap. They get good reviews.
This one's only 350w, but it is still cheap (£20) for 350w

http://www.hexus.net/content/reviews/review.php?dXJsX3Jldmlld19JRD03NzQmdXJsX3BhZ2U9Ng==

Though Hiper seems to pull up only uk sites. May be a uk make / uk
retailers only at the moment!

Though it's questionable whether reviewers know what they're doing when
they test PSUs.

Electronics people I speak to tell me that even using a multimeter to
back-probe the PSU's connectors while the PSU is under load, will not
give accurate results. And it's impossible to test a PSU's voltage
accurately and adequately, without expensive equipment. Though that
back-probing with a multimeter method can pick out a a really bad psu.

For those that want to know what back-probing is / where I heard of it.
The back-probing method is in Scott Mueller's popular well known PC
Tech book "upgrading and repairing PCs" now in its 16th edition.

you poke a probes of the multimeter into the back of - say - the
connector that is pluged into the HDD. And you check that the red wire
is about 5V. yellow is 12v .

It's all DC outside the PSU box, 'cos the PSU is also an AC DC
converter. so it's safe to probe connectors.
 
Random said:
Is it worth getting unbranded PSUs over say an Antec one? The Antec
PSUs seem to be 5x more expensive than an equivalent rated unbranded
one.

Can you risk damaging your hardware with a bad PSU, or is the danger
exaggerated?
Click to expand...

Look up "inductor saturation". When that happens, the transformer or
choke can generate a voltage surge that can damage the computer, and
because of where it occurs in the PSU, the overvoltage protection can't
prevent it. Also the output filter in a cheapo PSU won't block it well
enough, and the surge is more likely to be generated in the first place
in a cheap PSU becaust the transformer and chokes are smaller in
capacities.

I'm a very cheap person, but one thing I won't skimp on is PSU quality.
OTOH in the past 3+ years I haven't spent more than $25 on one, and
that was a 300W Antec with a case. There's just no need to spend a lot
for the best quality, especially because Fortron-Source Power PSUs
costing only $20-30 for 300-350W and $40-50 for 400-450W are available
from sources like www.newegg.com and www.directron.com. These PSUs
tend to be very conservatively rated, meaning a 350W Fortron is
probably as strong as the average 400-450W model of another brand. So
why take a risk with crap? You won't save much money.

The best PSU reviews are probably at www.xbitlabs.com (includes very
detailed description of their test methods), www.tomshardware.com,
www.slcentral.com, www.3dvelocity.com, and www.silentpcreivew.com.
Unlike most other review sites, their full load tests apply an actual
full load, not the usual 250W.
 
Now that you've heard from everyone who says buy the branded PSU. In
my experience, I've not noticed any real difference.

That's not to say there isn't one, but; I've got somewhere around 12
PC's running out there and not one has failed because of a PSU. In
fact, not one has failed at all. Most of these PSUs came with the
cases.

I will say this; I'm running a MSG 500 watt no name that came with
this case. It runs 4 hard drives, an overclocked 6600 GT, XP-3200,
sound card, TV capture card and all the extras including about 20
lights and 10 fans. According to my mobo software, the voltages are
right on spec and it's been running for about 7 months now. In fact,
the voltage specs are closer then the Antec 430 "True Power" it
replaced. No computer will be damaged by a PSU that's within 5 to 7%
of it's voltage specs.

I find the vast majority, check that; EVERY failure I've seen, has
been due to dirt and filth. If the computer is kept clean, and the
inside of the PSU is peridically blown out, you shouldn't have any
problems. Just make sure the cooling fans are working. That's true
with any power supply.

Remember; if there were vast differences in PSUs; as some suggest.
The cheapo's wouldn't even be out here.

Buy the way, OEM's use the cheapest components they can find anywhere.
Don't let anyone fool you, that includes the PSU.
 
They also said the household electricity was just fine
because the lights work. Fortunately no one was home when the
house exploded due to a missing safety ground and another
problem.

Dannysdailys is using same logic to justify a good power
supply. By his own admission, he does not know what a good
power supply must do. But 'since the lights work' then
everything must be OK. His logic was also sufficient to
murder seven Challenger astronauts. Granted, a power supply
will not cause seven deaths. But the lessons from that
exploded house AND from the Challenger are how to not approach
any problem. Everyone should see right through the failed
reasonings by dannysdailys.

My experience is that those who buy cheap supplies then have
computer damage that they blame on mythical events such as
surges. No wonder they've "not noticed any real difference".

Every failure cited due to 'dirt and filth', well, we then
know that person has insufficient experience and usually does
not have basic electrical knowledge. When he can cite
specific parts that have failed AND fix a power supply work by
soldering parts, only then can we suspect minimal technical
knowledge. Any power supply damaged by 'dirt and filth' was
defective when it was designed.

Furthermore, if using motherboard voltage monitors to
measure power supply voltage, then he failed to learn why it
is only a monitor - why that monitor must first be
calibrated. But again, too many computer experts don't even
have basic electronics experience; instead promote myths.

A vast difference exists in power supplies which is why cheap
ones do exist - in direct contradiction to what dannysdailys
posted. Bean counters among us routinely pretend to be
computer experts only because the power supply worked for him
this week or this year. Bean counters promote power supplies
that are missing essential functions. Overseas suppliers have
found a so lucrative market by dumping supplies missing
essential functions. Notice those supplies never come with a
full page of numerical specs - since they are selling to
dannysdailys. So many *computer experts* don't even know how
electricity works. dannysdailys has demonstrated same by
promoting business school speculations as if fact. He even -
and this is embarrassing - claims that dirt and filth caused
power supply failures. Clearly bean counter reasoning had
nothing to do with those failures.
 
The multimeter on a power supply under load will provide
extremely accurate results. The typical region of unknown is
when voltage is in the lowest quarter of those limit. In that
region, an oscilloscope is necessary to confirm what is
probably due to excessive ripple voltage.

Meanwhile, if that electronics person says a multimeter is
not sufficient, then he also said why for numerous reasons.
IOW so many electronics people don't know why ... and promote
common myths.

Been doing this stuff for over 30 years. The multimeter is,
by far, the most effective tester per quid. Once the meter
has confirmed power supply 'system' integrity, then move on;
the power supply 'system' is OK.

I don't care what those reviewers say is and is not a
'quality' supply. Most reviewers could not say what each part
inside the supply is nor what those parts do. A highly
regarded power supplies ... and supply is missing essential
overvoltage protection? Did any of those reviewers even
mention OVP? Why not? Why are they reviewing supplies when
they don't even know of essential functions? OVP is one of
the first essential functions 'forgotten' to sell power
supplies at excessively low price to people such as those
reviewers. AT £20, the supply must be missing essential
functions.

Looking at that Hiper supply, well, where are any specs?
This is how a scam artist sells his product to the naive using
hype. Supply does not meet minimal specs, so they don't even
try to claim such numbers. No numerical specs is a damning
omission. Then when the supply causes more damage, well, they
did not even claim to meet minimal industry standards. That
£20 is provided for the dannysdailys of this world. See the
other post about his recommendations and technical integrity.

When those reviewers test supplies, do they routinely short
out all supply outputs with a wire sized per Intel specs?
Intel even defines the size of that shorting wire. A properly
constructed power supply must have all outputs shorted
together and still the power supply is not damaged. Some 'so
callled' quality supplies fail on this; another required spec.
So why did those reviewers not test this and so many other
functions required even by Intel's ATX specs? Or did
reviewers even read those specs - to confirm each supply meets
each of those specs?

Pictures that demonstrate where to take voltage readings are:
http://techrepublic.com.com/5102-10586-5566528.html
www.ochardware.com/articles/psuvolt/psuvolt.html

Notice the difference. Posted are the whys. To have
credibility, those other 'electronics people' must also
provide whys. Without those whys, then what they have said is
nothing more than rumor.
 
Now that you've heard from everyone who says buy the branded PSU. In
my experience, I've not noticed any real difference.

That's not to say there isn't one, but; I've got somewhere around 12
PC's running out there and not one has failed because of a PSU. In
fact, not one has failed at all. Most of these PSUs came with the
cases.

I will say this; I'm running a MSG 500 watt no name that came with
this case. It runs 4 hard drives, an overclocked 6600 GT, XP-3200,
sound card, TV capture card and all the extras including about 20
lights and 10 fans.
Click to expand...

I'm sure you feel 20 lights and fans is significant- it
isn't. , nor are light load components ike sound or TV card.
According to my mobo software, the voltages are
right on spec and it's been running for about 7 months now.
Click to expand...

Here we see one of the problems. 7 months is not evidence
of anything. A PSU is not meant to run for 7 months, it's
meant to run for the life of the system. You have no reason
to believe the generic will last as long as a good PSU, and
it may easily damage other parts when it fails (or
progressively, during use).
In fact,
the voltage specs are closer then the Antec 430 "True Power" it
replaced. No computer will be damaged by a PSU that's within 5 to 7%
of it's voltage specs.
Click to expand...

If your mobo software shows it right-on-spec, it probably
isn't right-on-spec, rather it's over spec because the
sensors tend to read lower than actual values. It is
essential to measure voltage at the PSU connector to the
powered equipment, as that is the only place that determines
if the PSU is properly outputting the rated voltage per
application.

I find the vast majority, check that; EVERY failure I've seen, has
been due to dirt and filth. If the computer is kept clean, and the
inside of the PSU is peridically blown out, you shouldn't have any
problems. Just make sure the cooling fans are working. That's true
with any power supply.
Click to expand...

Certainly dust can increase temps, but for the most part
generics routinely fail because of design/components, not a
little dust. Fans are definitely a major issue though,
generics usually have poor generic sleeve-bearing fans that
are subject to seizure.


Remember; if there were vast differences in PSUs; as some suggest.
The cheapo's wouldn't even be out here.
Click to expand...

Remember; if there weren't vast differences, the expensive
ones wouldn't even be out there. Nobody would spend 3X as
much.

The FACT of the matter is that real testing proves the
difference. Put a load on a PSU, see if it pops and how
well it performs. It is certainly possible to run a modern
box from a generic PSU, so long as the actual power
requirement of the system is less than the true power
capability of the generic PSU, _NOT_ the labeled capacity of
the generic PSU. It's rather easy and reproducible to blow
up a generic psu by simply loading it to the output on the
label and leaving it running. IMO, that's fraud.
Buy the way, OEM's use the cheapest components they can find anywhere.
Click to expand...

Completely wrong.
OEMs generally do not cut corners in PSU construction. They
do tend to use PSU with lesser wattage output but comparing
an OEM 300W to a generic 400W, it's no contest, the OEM PSU
is usually better.

OEMs already know that the "cheapest" thing to do is not to
cut so many corners that systems fail under warranty.
Generic PSU manufacturers do not refuse to sell to OEMs, on
the contrary they would LOVE to have high volume OEM
sales... but OEMs qualify parts and choose better PSU.

Don't let anyone fool you, that includes the PSU.
Click to expand...


Agreed. Listen to those who have done testing to the point
of failure... PSUs that seem to work for a few months are
not evidence- PSUs that fail are evidence. Knowing a
failure threshold lets one determine the actual capability
of a PSU and only then can they gauge whether that generic
can be derated enough to be appropriate for any particular
system in the long-term. Otherwise where is the savings, if
you have to buy another PSU when the first one fails?

If you know a 500W generic can output 280W, and that's all
the system needs (would be a reasonable ballpark figure for
the system you described above) then that gets the system
running. Multi-hundred dollar system running off a PSU
proven inferior for a $20 savings.... bad system choice-
anyone can build unreliable boxes. 7 months isn't even
1/10th of the lifespan one should expect from a modern
system IF they chose to run it that long, and indeed, the
average system these days is already a few years old.
 
w_tom said:
The multimeter on a power supply under load will provide
extremely accurate results. The typical region of unknown is
when voltage is in the lowest quarter of those limit. In that
region, an oscilloscope is necessary to confirm what is
probably due to excessive ripple voltage.

Meanwhile, if that electronics person says a multimeter is
not sufficient, then he also said why for numerous reasons.
IOW so many electronics people don't know why ... and promote
common myths.

Been doing this stuff for over 30 years. The multimeter is,
by far, the most effective tester per quid. Once the meter
has confirmed power supply 'system' integrity, then move on;
the power supply 'system' is OK.

I don't care what those reviewers say is and is not a
'quality' supply. Most reviewers could not say what each part
inside the supply is nor what those parts do. A highly
regarded power supplies ... and supply is missing essential
overvoltage protection? Did any of those reviewers even
mention OVP? Why not? Why are they reviewing supplies when
they don't even know of essential functions? OVP is one of
the first essential functions 'forgotten' to sell power
supplies at excessively low price to people such as those
reviewers. AT £20, the supply must be missing essential
functions.

Looking at that Hiper supply, well, where are any specs?
This is how a scam artist sells his product to the naive using
hype. Supply does not meet minimal specs, so they don't even
try to claim such numbers. No numerical specs is a damning
omission. Then when the supply causes more damage, well, they
did not even claim to meet minimal industry standards. That
£20 is provided for the dannysdailys of this world. See the
other post about his recommendations and technical integrity.

When those reviewers test supplies, do they routinely short
out all supply outputs with a wire sized per Intel specs?
Intel even defines the size of that shorting wire. A properly
constructed power supply must have all outputs shorted
together and still the power supply is not damaged. Some 'so
callled' quality supplies fail on this; another required spec.
So why did those reviewers not test this and so many other
functions required even by Intel's ATX specs? Or did
reviewers even read those specs - to confirm each supply meets
each of those specs?

Pictures that demonstrate where to take voltage readings are:
http://techrepublic.com.com/5102-10586-5566528.html
www.ochardware.com/articles/psuvolt/psuvolt.html

Notice the difference. Posted are the whys. To have
credibility, those other 'electronics people' must also
provide whys. Without those whys, then what they have said is
nothing more than rumor.
Click to expand...

Thanks for the info. That article was great.
that techrepublic site used to spam my inbox! They had good articles,
it was good spam. but they don't seem to archive the spam on the site!
And mostly the articles weren't releveant to me at the time.

Regarding the £20 PSU (it was recommended to me, it wasn't just a
random £20 PSU I discovered!)

Here is the article on their site with spec and datasheet
http://www.hipergroup.com/English/products/hpu-3s350.html

They do high end PSUs too, at £70
 
Back
Top