Power Supply

[Dave Zass]

I just replaced a 465w PSU with a 520w PSU. However, the grid on the side
of the PSU would lead me to believe that I just "downgraded" the PSU.

The original, a 465W Enermax, shows the following:

+3.3V 36A
+5V 44A
+12V 20A
-5V 2A
-12V 1A
+5VSb 2.2A

The replacement, a 520W Aspire, shows:

+3.3V 28A
+5V 30A
+12V 34A
-5V 0.3A
-12V 0.8A
+5VSb 2.0A

I thought my original PSU was starting to die, so I bought a replacement.
Did I just go backwards?

 

[w_tom]

Did the new supply come with a long list of technical
numbers making essential spec claims? Did it specifically
state that it has overvoltage protection, overcurrent
protection, EMI & RFI compliance, safety codes, and a long
list of other numbers? If not, then you probably got
something worse - something intentionally dumped into the
market at higher profit and lower retail price because so
many 'computer assemblers' don't have basic electrical
knowledge.

The 465 watt power supply was probably significantly more
power than you would need. So why did you replace it?

Another damning fact. Why is the current on the 'higher
power' supply listed as less? This occurs when power supplies
are bought using a classic bean counter mentality (price and
watts) rather than first learning how and why things work
(demand in writing a long list of numerical specifications).

Another characteristic of power supplies marketed to bean
counters. Those power supplies can also cause other computer
hardware damage. One essential function found in a minimally
sufficient supply - power supply failure will not cause other
hardware failure. Notice how many 'computer assemblers'
assume a power supply failure also caused other component
failure. They consider that acceptable?

From few numbers provided, I would suspect the Aspire could
also contribute to other component damage.

One does not suspect a power supply is failing. One first
sees it in numbers using a 3.5 digit multimeter. Getting
those numbers takes only minutes.

 

[Dave Zass]

The 465 watt power supply was probably significantly more

power than you would need. So why did you replace it?

Maybe I wanted a different color PSU. Maybe I wanted more power because I
added more devices. Maybe my old PSU has survived a dirty, dusty
environment and I decided I wanted a new one. Maybe....... Other than
suiting your purpose to make me look foolish, why do you care why I replaced
it?

And "probably significantly" more than I need? How much do you recommend,
since you've got me pegged.

Another damning fact. Why is the current on the 'higher
power' supply listed as less? This occurs when power supplies
are bought using a classic bean counter mentality (price and
watts) rather than first learning how and why things work
(demand in writing a long list of numerical specifications).

I trusted recomendations and bought from a respected source. The PSU was
deemed suitable and comparable.

One does not suspect a power supply is failing. One first
sees it in numbers using a 3.5 digit multimeter. Getting
those numbers takes only minutes.

One uses a bios hardware monitoring program to check voltages. One noticed
that the voltage was not remaining stable under load on the +12v. One
prefers to eliminate possible sources of a problem, in order to isolate the
problem. One asked a question which was ignored and used by Another to make
themself feel good by getting on their soapbox and fingerpointing instead of
offering simple advice. One is accustomed to seeing replies from Another.
One realizes that he gets what he pays for on usenet.

 

[Chuck F.]

** Repaired the accumulated effects of top-posting and snipping of
attributions **

Maybe I wanted a different color PSU. Maybe I wanted more
power because I added more devices. Maybe my old PSU has
survived a dirty, dusty environment and I decided I wanted a new
one. Maybe....... Other than suiting your purpose to make me
look foolish, why do you care why I replaced it?

And "probably significantly" more than I need? How much do you
recommend, since you've got me pegged.


I trusted recomendations and bought from a respected source.
The PSU was deemed suitable and comparable.


One uses a bios hardware monitoring program to check voltages.
One noticed that the voltage was not remaining stable under load
on the +12v. One prefers to eliminate possible sources of a
problem, in order to isolate the problem. One asked a question
which was ignored and used by Another to make themself feel good
by getting on their soapbox and fingerpointing instead of
offering simple advice. One is accustomed to seeing replies
from Another. One realizes that he gets what he pays for on
usenet.

The critical item is that you observed problems on the +12 line.
This is often highly stressed at turn-on time, because it is used
to spin up all the rotating memory. The cure is often simpler,
i.e. you can usually arrange to delay drive spin-up, either
specifically in the bios, or by ensuring that time is spend in
memory testing during the post.

You didn't specify this observation in your original, without which
w_toms caustic observations are justified (but not the top-posting).

You probably now have a lower quality PS, but that is not certain.
At least you have been shown various criteria.

 

[w_tom]

BIOS (motherboard) monitor is just that - a monitor.
Intended to detect voltage changes. To perform voltage
measuring, that monitor must first be calibrated - which is
why the 3.5 digit multimeter is necessary. Quite probable
that the voltage monitor detected a low +12 volts - now that
this important detail was finally provided. But we still don't
know without accurate numbers; we can only speculate.

Meanwhile, I detect some emotion in your response. If I
provide technical facts, then it is only a technical post -
with absolutely no personal opinions attached. If I want to
insult you, you will have zero doubt about it. I imply
nothing. I don't waste time with silly emotion nor waste time
making a post 'politically correct'. What was posted was
technically accurate - as any human wants in a reply.

Essential to understanding those label numbers are written
and numerical specifications. Anyone can cite a short term
surge current instead of a constant current on a label - to
claim power a supply really cannot really output. Would they
be lying? Legally no. Technically yes. But most 'computer
assemblers' would never know the difference. Generally,
distorting power and current output numbers is common among
manufacturers who also don't provide a written list of
numerical specifications.

I am sorry if you took personal insult to that fact. No
insult existed in that previous post. Facts were posted as
all facts should be - blunt - without any attempt to be
politically correct nor worry about personal feelings. Many
power supplies are dumped into the clone computer market
because so many 'computer assemblers' don't have basic
electrical knowledge. One need not even have technical
knowledge to quickly suspect inferior power supplies - no
numerical specs. That is a fact that does not even imply
personal insult.

Kony has posted same in another thread:

To be concise, there is little point reading the label on a
PSU unless you already have reason to specifically, trust
the label.

I further suggested how one might 'trust' the label - long
list of written and numerical specs that inferior supplies
(that are a 'step backward') don't provide so that you also
don't ask about missing and essential functions.

I pegged you for nothing. I don't care is your face is an
anus. That is completely and technically irrelevant. You
asked a technical question. I replied technically accurate -
including reasons why those answers exist. You were also
properly warned of 'computer assemblers' masking as
electrically literate. They are but one reason why selling
inferior supplies is so profitable. But again, that is not an
insult. Just blunt facts.

 

[kony]

One uses a bios hardware monitoring program to check voltages. One noticed
that the voltage was not remaining stable under load on the +12v.

It is not possible to get an accurate 12V reading from
onboard voltage sensor and bios or software readout when the
board has any fair to high current 12V parts running on it.
Even when there aren't any, there is significant deviation
between what any one point on the board sees as the "12V
rail" voltage level, and an actual 12.0 voltage.

Point being it is inherant that the voltage drops on the
motherboard. Manufacturers may "pad" the reported numbers
in attempt to offset this, but it cannot account for current
changes. The power supply voltage is only reliably read at
the one point it is "supposed" to output within +-5% of the
rail rating- at the physical connector connected to the
load. In this case, it means inserting a multimeter probe
into the connector plastic housing, for the 4-pin CPU power
connector, the single 12V lead in the 20+ pin ATX connector,
and at all significant consumers of 12V current, for example
hard drives during system power-on due to their higher
spinup current requirements.

It is quite common and to be expected that a motherboard
reading through bios will read lower than 12.0V volts. IN
fact, if the bios reports 12.0V, it is to be expected that
the power supply itself is outputting too high, above 12.0V.
Higher voltage isn't necessarily better either, the
transformer has a certain capacity and a higher voltage
output will simply reduce the current it can support, or
lengthen the recovery time.

I'm not trying to suggest there was nothing wrong with the
Enermax, that is unknown- only that there are a lot of
misconceptions on the web about suitability of certain
features for troubleshooting purposes.