replacing power supply?
Warren
supply just to be safe. Currently I've got a "700 WATT OEM ATX PSU REV2"
anyonhe know if I can assume it's OK to buy a basic 700W power supply
from Newegg or what do I need to keep in mind when finding a replace
ment that will asuredly fit my tower? It's an Alienware AREA-51 7500-R3
(don't ask, worst money I've ever thrown away in my life and I've
learned a *valuable* lesson) and I would like to be sure that a new PS
will fit. Any advice is most welcome. Thanks.
--
HTTP://www.sushifaq.com/ The Sushi FAQ ...><((((º>
HTTP://www.sushifaq.com/sushiotaku/ The Sushi Otaku Blog
HTTP://www.sushifaq.com/sushiyapedia/ Sushi-Ya-Pedia Restaurant Finder
HTTP://www.theteafaq.com/ The Tea FAQ
HTTP://www.jerkyfaq.com/ The Jerky FAQ
HTTP://www.omega3faq.com/ The Omega 3 Fatty Acids FAQ
Dave
"Warren" <usenet1--...@ransom.us> wrote in message
news:13tan72...@news.supernews.com...
>I keep getting weird hardware failures and thought i'd replace my power
>supply just to be safe. Currently I've got a "700 WATT OEM ATX PSU REV2"
> anyonhe know if I can assume it's OK to buy a basic 700W power supply from
> Newegg or what do I need to keep in mind when finding a replace ment that
> will asuredly fit my tower? It's an Alienware AREA-51 7500-R3 (don't ask,
> worst money I've ever thrown away in my life and I've learned a *valuable*
> lesson) and I would like to be sure that a new PS will fit. Any advice is
> most welcome. Thanks.
>
I'm not familiar with alienware, but googling the 7500 I found a open
chassis view of the base model of your system and it appears to contain a
standard ATX format power supply. If so, that's a good thing. To be
absolutely sure, you'd have to post the exact make and model number of the
power supply so that we could verify the specifications of it.
But assuming for a moment that you need a new power supply that is ATX
format, which specific power supply will work for you depends also on how
your system is equipped. That particular model is SLI capable, but from
what I read, two video cards is an option. In either case, you shouldn't
need a 700W power supply, assuming that the power supply you buy is not
crap. For one video card, 500W or more should do. For two video cards,
you'll probably want 560-600W or a little more. Keeping that in mind, I'd
recommend you try one of the following, if it must be from newegg:
One Video card:
http://www.newegg.com/Product/Product.aspx?Item=N82E16817151027
Two Video Cards:
http://www.newegg.com/Product/Product.aspx?Item=N82E16817812005
Actually the second one would work for either.
Again, you do NOT need 700W to power your rig regardless of how it's
equipped. So I suspect the original power supply is probably junk quality.
That is, it probably outputs as much actual current as a good quality 550W
or so. -Dave
Warren
spent 1.5 years getting fixed by the manufacturer and I'm so sick of
them that it's all on me now. I'll verify the PS, but the second may be
a good choice. I use a GF 7950 GX2 (2 cards in SLI) and with a DUal COre
Extreme X6800 CPU and a few other addins (Ageia PhysX card, etc) perhaps
they though that 700W made sure I would always have enough power, but
your assessment make be correct. that's still a pretty high rating for a
PC.
I appreciate the help, if it's been mu power supply all along I'll be
happy to stop blowing out parts, and it's worth the $$$ to replace it
just to be safe.
philo
> supply just to be safe. Currently I've got a "700 WATT OEM ATX PSU REV2"
> anyonhe know if I can assume it's OK to buy a basic 700W power supply
> from Newegg or what do I need to keep in mind when finding a replace
> ment that will asuredly fit my tower? It's an Alienware AREA-51 7500-R3
> (don't ask, worst money I've ever thrown away in my life and I've
> learned a *valuable* lesson) and I would like to be sure that a new PS
> will fit. Any advice is most welcome. Thanks.
>
>
I'd run diagnostics such as a RAM test
and the HD mfg's
I find it hard to fathom that a 700w supply would be underpowered
Dave
>
> I'd run diagnostics such as a RAM test
> and the HD mfg's
> I find it hard to fathom that a 700w supply would be underpowered
>
>
A 700W no-name power supply could easily be under-powered. -Dave
philo
"Dave" <no...@nohow.not> wrote in message
news:fr4bf7$tcd$1...@registered.motzarella.org...
Anything is possible of course...
but why buy a new PSU unless one has first at least tested the RAM and HD
Warren
numerous hardware failures over the past year (the manufacturer has
replaced the motherboard twice, the memory, and basically replaced every
component at least once so far) I am thinking my issue must be the power
supply. I have no record of them replacing the PS, though I did mention
to them that it might be the source of the issues (Alienware, in my
experience, has the least knowledgeable or caring technicians I have
met, they even sent a guy to my house to replace the MoBo once, then
called me saying that he shouldn't have replaced it and that to maintain
my warranty I would have to send the system back for them to check his
work).
So, since they think they know better than someone who has normally
built his own machines (I'm no master, but I know enough to do this), it
seems that they may have overlooked the problem all this time. At least
I hope that this is the issue and I can finally have a working machine.
Every time it gets fixed, it works for a few weeks and then dies. So I
am hoping that a faulty power supply is frying my components and I can
finally get this all behind me and have a rig that works.
philo
Well I guess you may as well replace the supply then
hope the machine stays working !
Except for a machine that I built for my GF a few years ago.
my machines are all pretty much made out of discarded or surplus parts.
One day I bought a *new* power supply.
it lasted about 24 hours then went **POOF*** <G>
w_tom
> I've tested both and they diagnostics report all is well. Due to
> numerous hardware failures over the past year (the manufacturer has
> replaced the motherboard twice, the memory, and basically replaced every
> component at least once so far) I am thinking my issue must be thepowersupply.
> I have no record of them replacing the PS, though I did mention
> to them that it might be the source of the issues
Replaced how many parts on speculation and fixed nothing? At what
point do you 'follow the evidence' rather than replace parts on wild
speculation. After all that labor, what was accomplished? Even the
replaced parts may still be defective. Worse, nobody has yet to
identify the actual failure. It's called shotgunning.
Two minutes with a multimeter would have identified the power supply
'system' (more than just a power supply) as 'definitively good' or
'definitively bad'. Currently your existing 'system' and a new supply
both are 'unknown'. Two minute procedure is posted in "When your
computer dies without warning....." starting 6 Feb 2007 in the
newsgroup alt.windows-xp at:
http://tinyurl.com/yvf9vh
Connector chart to locate each color:
http://www.hardwarebook.net/connector/power/atxpower.html
In your case, relevant part of the test occurs by maximizing load
(multitasking to all peripherals) to measure voltage on any one of
purple, orange, red, and yellow wires. Voltages must remain above
3.23, 4.87, and 11.7. If so, supply 'system' would not explain
intermittent failures (that are not specifically defined).
It fails? How? Specifically what is the failure during when and
with what numbers and error messages reported?
Of course other important information from system (event) logs and
Device Manager were also verified and posted?
What diagnostics were performed? Obviously the most useful
diagnostics execute without Windows.
What happens where parts are taken from cool (room temperature) to
ideal operating temperature (when heated with a hairdryer on highest
heat setting)? No I am being bluntly and technically honest. Any
computer is happy when at those higher temperatures from a hairdryer
on high. But defective (intermittent) computer hardware makes its
defect obvious at those ideal temperatures.
Most computers consume less than 150 watts. A 300 watt supply is
sufficient (if properly sized for each voltage). But many power
supplies measure something different to declare that 300 watt supply
as 450 watts. 500 watts would be more than sufficient. But when a
computer assembler has no idea how electricity works, then assembler
speculates: more watts is better - 700 watts. To dump supplies that
are missing essential functions for higher profits into a market of
the technically naive, the manufacturer proclaims even more watts
while forgetting to provide some essential functions. IOW they
maximize profits by selling inferior supplies with more watts to the
naive. 700 watts is to be sure? Sure of what? More watts do not
mean a better supply. 700 watts for an average load of less than 150
watts?
Where is the long list of functions required in all power supplies?
That spec list gets forgotten when the large watt power supply sells
at a lower price for higher profit. IOW the 'more watt' supply is the
inferior supply - when designed to sell to the electrically naïve.
Is power supply sufficient? The above 'less than two minute'
procedure answers that - and more. This post has numerous questions.
Every one requires an answer since answers are also interrelated.
They used shotgunning. Everything gets replaced and system is still
defective. A classic example of why trained techs don't use
shotgunning; but untrained A+ Certified tech do.
War
I'm assuming you are right. No one at Alienware ever gave me a straight
answer about the machine, and I think they just kept swapping out parts
until they got it to work and they sent it back to me hoping it wouldn't
break again. IMHO I got pretty much the worst service I've ever seen
from their technicians, even the higher ups that they supposedly
escalated the issue to (yeah, right).
Personally, I'm not as skilled as you in terms of testing, so I'm forced
to shotgun it. I'm swapping out the MoBo, which is definitely fried, and
the PS. Hopefully that will do it, if not... I dunno.... But my drives
and mem work fine, were tested, so that leaves pretty much the CPU
and/or PS (or possible defective cards like vid, etc). I'm just willing
to go this far monetarily and see if it helps. we'll see....
--
HTTP://www.sushifaq.com/ The Sushi FAQ
War
http://www.newegg.com/Product/Product.aspx?Item=N82E16817174028
but it's discontinued. the form factor looks different than the ones you
suggested, any ideas if they might fit or how I can determine which PS I
should get to replace this one if those don't? (BTW, I have 2 vid cards
in SLI mode, so I'd go for the higher output one).
Thanks.
--
HTTP://www.sushifaq.com/ The Sushi FAQ
w_tom
> Personally, I'm not as skilled as you in terms of testing, so I'm forced
> to shotgun it. I'm swapping out the MoBo, which is definitely fried, and
> the PS. Hopefully that will do it, if not... I dunno.... But my drives
> and mem work fine, were tested, so that leaves pretty much the CPU
> and/or PS (or possible defective cards like vid, etc). I'm just willing
> to go this far monetarily and see if it helps. we'll see....
What is this skill? The entire 'skill' comes from newsgroup
replies if and when you perform those simple tasks that result in
facts and numbers. Tasks that result in useful replies. Repair
things to learn. However this post again teaches nothing because the
two minute procedure was not performed AND you are still using shotgun
reasoning.
I see no reason to conclude a motherboard is fried. That conclusion
apparently comes from defective reasoning. The point. You have
concluded without any fact to justify that conclusion. Your replies
(and therefore what you learn) will only be as useful as facts
(numbers) provided. So far, my reply is again unhelpful because facts
and numbers were not provided. Is the motherboard fried? Those who
can say will post no useful replies until facts are apparent.
Get the meter. A multimeter is so complex that only K-mart shoppers
are intelligent enough to use one. Yes, black humor. Do what the two
minute procedure defines. Then post those numbers. Posting takes
longer than getting those numbers. Resulting reply is facts based
upon numbers - no wild speculation - no money wasted on shotgunning.
What is defined repeatedly in CSI - "follow the evidence". How many
hear that in CSI and still never understand? We learn by doing.
Currently are zero reasons to believe motherboard is defective - which
is obvious if when that CSI expression is understood.
Just because something works one place does not mean it works OK.
Even a defective power supply can boot one computer but not work in
another becauase the supply was always defective in both. You are
thinking in binary terms - something is either good or bad. Things
are either good, bad, or unknown. How do we know this? Why does
every military academy graduate everyone with engineering training?
Military also needs people who can solve problems - people who learn
by making things work.
Disk drive typically does not cause computer failures. Just one of a
long list of things to be learned by doing now what that previous post
recommended - and if you do not shotgun..
Well, this post could have moved forward. But your last post did
not answer simple questions. You probably will not get a better
trained source. Again I do nothing for you because you provided
nothing to work with. Answer obvious questions and perform the simple
two minute procedure to reap fruit in the next reply - and also to
learn.
It fails? How? Specifically what is the failure during when and
with what numbers and error messages reported?
Of course other important information from system (event) logs and
Device Manager were also verified and posted?
What diagnostics were performed? Obviously the most useful
diagnostics execute without Windows.
Zero reasons say motherboard is fried. What do we know?
Everything is still at various degrees of 'unknown' meaning labor has
still accomplished nothing useful. Nothing says anything is yet
'definitively good'. Nothing says a motherboard is fried.
War
Well, for now, I can say that channel 1 & 3 on the SATA controller will
not recognize any drives whatsoever. channels 2& 4 worked for a while,
but have started to occasionally not recognize drives too, so that leads
me to conclude that at least that part of the MoBo is dead/dying. the
same drives/memory work just fine in another machine with no quirks and
I've run tests/diagnostics on those items which confirm that they are
working as expected, at least I got no errors from the runs. Data keeps
getting corrupted on any drives connected in this machine (I've tried a
few), but seem to work fine in other machines as well.
As for the meter, perhaps I'll have some time this weekend to go find
one. In understand the point you are trying to make, and I appreciate
your help, but you don't have to sound so hostile simply because you are
talking to someone without your depth of experience. I'll keep digging
though, thanks.
w_tom
> Well, for now, I can say that channel 1 & 3 on the SATA controller will
> not recognize any drives whatsoever. channels 2& 4 worked for a while,
> but have started to occasionally not recognize drives too, so that leads
> me to conclude that at least that part of the MoBo is dead/dying. the
> same drives/memory work just fine in another machine with no quirks and
These are symptoms typical of a defective power supply system.
Things work here an not there: can be a power supply 'system' problem
or can be due to other reasons. Without numbers, nobody can say
anything 'definitively'. Are channels 1 & 3 defective? That
conclusion can only be wild speculation without the numbers.
Nonthing in that post is hostile. Described is how technical
problems are addressed - blunt, factual, with complete disregard to
emotion, and ignoring how another's emotions might perceive it.. If
it sounds hostile, then your preceptions are being applied to the
text. IOW 'assuming' rather than sticking strictly to the facts.
Does it specifically state something hostile? Again the same point.
Assumptions are resulting in error. Relevant is only text which
describes how to solve problems with contempt only for failures.
Provides was the best solution. How many others designed power
supplies even 20 years ago? Get the meter so that one with knowledge
can provide assistance.