Is it my RAM? Please help

  • Thread starter Thread starter Kevin M
  • Start date Start date
K

Kevin M

Hi, just recently my computer started acting up a year after I completed the
build. I does not want to turn on for me, the power light works, but the HD
light does nothing and no signal is sent to the monitor. The one thing that
fixes it is "unseat" the RAM and then click it back in, and then the
computer fires up. I don't pull the RAM out all of the way, just unseat one
corner of em all(4x512mb) and secure them back in to their slot. What could
possibly be the problem? I'm getting tired of doing this every time.
Thanks in advance for your help.
 
Hi, just recently my computer started acting up a year after I completed the
build. I does not want to turn on for me, the power light works, but the HD
light does nothing and no signal is sent to the monitor. The one thing that
fixes it is "unseat" the RAM and then click it back in, and then the
computer fires up. I don't pull the RAM out all of the way, just unseat one
corner of em all(4x512mb) and secure them back in to their slot. What could
possibly be the problem? I'm getting tired of doing this every time.
Thanks in advance for your help.
Click to expand...

This is a very unusual problem you have. Is the environment
very dirty, dusty, humid, etc, etc? Do the contacts on the
memory and board look good?

Does the memory fit good, firmly but fully seating in the
slot? Is it possible the board has different length
standoffs installed or the case is deformed a bit (the
motherboard tray that is) such that the board itself isn't
perfectly planar?

What else do you do when going through the routine of
pulling the memory? Is it possible that (presumably you do)
turn off the AC power to the system prior to pulling the
memory, and that this alone is what's causing system to
start properly the next time rather than the memory being
reseated?

Check for extra standoffs behind the board, in case any
might be where they shouldn't and are shorting against the
back of the board. Is it possible the board has been under
some stress prior to installation and might have physical
damage like cracks in the traces?

What if you pull out, leave out all but one module? Would
it then continue having problems or start properly?

How about your power supply, does it have plenty of 5VSB
power? Is it a good brand or generic? Do you have a lot of
USB or PS devices connected and they're jumpered
(motherboard jumpers- see board manual) to use 5VSB rather
than 5V for power?
 
This can be from a faulty connector, or cold solder connection on the board
nearby, or under the RAM socket, or something intermittent on the RAM card
itself.

Try running the computer on one stick at a time, if you have two sticks of
RAM, and see if it still does the fault.

I had one mother board, that had a cold solder connection under one of the
RAM sockets causing it to be intermittent. This is rare, but can happen. If
you strongly suspect that there is a cold solder under the RAM socket, and
you have a proper solder station, you can try to re-sweat the solder
connections under the socket. You better know what you are doing, if you
decide to try this.


--

JANA
_____


Hi, just recently my computer started acting up a year after I completed the
build. I does not want to turn on for me, the power light works, but the HD
light does nothing and no signal is sent to the monitor. The one thing that
fixes it is "unseat" the RAM and then click it back in, and then the
computer fires up. I don't pull the RAM out all of the way, just unseat one
corner of em all(4x512mb) and secure them back in to their slot. What could
possibly be the problem? I'm getting tired of doing this every time.
Thanks in advance for your help.
 
Kevin M said:
Hi, just recently my computer started acting up a year after I completed the
build. I does not want to turn on for me, the power light works, but the HD
light does nothing and no signal is sent to the monitor. The one thing that
fixes it is "unseat" the RAM and then click it back in, and then the
computer fires up. I don't pull the RAM out all of the way, just unseat one
corner of em all(4x512mb) and secure them back in to their slot. What could
possibly be the problem? I'm getting tired of doing this every time.
Thanks in advance for your help.
Click to expand...

That IS weird. It sounds like the dimm(socket) is damaged. What if you take
out all but one ram stick? Then add the 2nd then three etc. If you leave
that one dimm empty does the computer fire up constantly?
 
Just out of curiosity, how did you determine this?
Click to expand...

Usually it would be through trial and error... you pull the
board and examine it, then if the solder looks iffy you get
out an iron and touch it up, retry it... a job only for
those with experience and steady hands since these pins are
tightly spaced.
 
doS said:
Gary, JANA's crap is copy and paste from other sources, beware the troll
wanna be
Click to expand...

Well ..... coping and pasting something doesn't make it
wrong, or RIGHT, either :-0

Memory retention and comprehension is key,
so I'll not address the trolling part
-since I am a fisherman and *DO* troll :-)

My point was, determining a "cold solder joint" takes
quite an expensive piece of equipment.

I'd be willing to bet that major motherboard manufactures,
i.e. Abit, Asus, etc. etc., do not manufacture the actual PCB,
but rather has a company such as Jabil
http://finance.yahoo.com/q/pr?s=JBL make them to their specs,
the same way Sears does with a wash machine. Thus, Abit
or whoever ... they would not have the necessary equipment
to check for a cold solder joint, let alone JANA!

I few RMA's here and there, Abit would just toss them.
After a significant number of RMA's, they would THEN contact the
actual manufacture and say, hey buddy, something's wrong here,
you want to check these out? The same as Sears would do when
X model of wash machine shows a problem with the electric motor.
 
Well ..... coping and pasting something doesn't make it
wrong, or RIGHT, either :-0

Memory retention and comprehension is key,
so I'll not address the trolling part
-since I am a fisherman and *DO* troll :-)

My point was, determining a "cold solder joint" takes
quite an expensive piece of equipment.
Click to expand...

No, but thanks for guessing.
When you have plated through-board holes with a metal pin
sticking through it usually does not take any fancy
equipment. Sure there might be some rare hypothetical
situation where only special equipement might find it, but
the cold joints on memory banks would be rare enough as it
is, there's really no point in driving even further down the
"rarest of rare" roads.
I'd be willing to bet that major motherboard manufactures,
i.e. Abit, Asus, etc. etc., do not manufacture the actual PCB,
but rather has a company such as Jabil
http://finance.yahoo.com/q/pr?s=JBL make them to their specs,
the same way Sears does with a wash machine. Thus, Abit
or whoever ... they would not have the necessary equipment
to check for a cold solder joint, let alone JANA!
Click to expand...

Considering that they design the boards in the first place,
I'd be quite surprised if they DIDN'T have plenty of test
equipment to cover this and many other faults. Did you
think they just snap together kits without any ability to
troubleshoot? How would they ever prototype and tune their
manufacturing processes?


I few RMA's here and there, Abit would just toss them.
After a significant number of RMA's, they would THEN contact the
actual manufacture and say, hey buddy, something's wrong here,
you want to check these out? The same as Sears would do when
X model of wash machine shows a problem with the electric motor.
Click to expand...

Sounds like a limited case scenario. Board manufacturers do
recertify RMA boards and they are resold... they do not all
get thrown away.
 
How?
Click to expand...


Look, examine the solder. Cold joints usually look
different, not uniform or shiney or low solder volume.
If you have a swing-arm fluorescent desk lamp with a big
magnifying glass in it, IMO it's easier. For example,
http://www.keysan.com/pictures/big/lwhw1010.jpg


"iffy" is determined how, by you?
Click to expand...

I think it would be easier for the person who has the board
to determine it rather than someone who doesn't, yes?
There are a lot of things someone *could* check if they're
inclined to, and maybe they'll find the problem or maybe
not... but while we're on the topic and someone is facing
this particularly rare problem...

Let alone, maybe >200 joints :-)

What if it is ONLY (hehehehhe) the trace?
What now Batman?
Click to expand...

Be specific.

I have to wonder if you're just trying to troll with no idea
what you're talking about. Maybe you never try repairs and
just think "that's too hard" or "that's impossible".
Sometimes it is, others it isn't.

I've repaired traces myself... they weren't memory bus
though, at least not that I *recall*. It's a PITA and I
have made some worse (or rather) completely lifted and
damaged the area in the attempt... but other times it turned
out fine. Such is the case with very small traces in
cramped areas, with someone who only does it on occasion and
doesn't value the equipment (boards) enough to buy the very
best reworking station, tools possible to increase the odds
of success... like myself. In other words the value of the
board, the time, the repair expenses all have to be weighed
but that doesn't make it impossible or even difficult to do
some repairs.

Most often though, one looks at the circumstances... when it
is a manufacturing defect it's usually evident right away
and the board can be returned... is not necessary nor
prudent to attempt repair at all. Afterwards it depends a
lot on the problem and/or if it's found... which isn't as
hard as you'd like us to believe, "sometimes".

Often I get boxes with problems and I pull out the problem
part(s) replacing them, so the box is up again in shortest
time possible. Then later when i get a chance a more
careful examination is undertaken. In other words it's a
matter of salvage, sometime the whole board is useful and
others it's just a candidate for cannibalization.
 
kony said:
On Wed, 20 Apr 2005 02:03:45 GMT, "Gary C"


No, but thanks for guessing.
When you have plated through-board holes with a metal pin
sticking through it usually does not take any fancy
equipment.
Click to expand...

Your reply does not state what it does "take", in YOUR *opinion*,
in a large mass production manufacturing plant.

{snip}
the cold joints on memory banks would be rare enough as it
Click to expand...

Rare? In mass production? You're too funny!
Kony, what is your experience inside a mass production, electronics plant,
that you can so surely state "would be rare" ?

SNAFU's and FUBAR's happen all the time in mass production.

Considering that they design the boards in the first place,
Click to expand...

On paper.
I'd be quite surprised if they DIDN'T have plenty of test
equipment to cover this and many other faults. Did you
think they just snap together kits without any ability to
troubleshoot? How would they ever prototype and tune their
manufacturing processes?
Click to expand...

Sure, two guys named Wong and Wang with a VOM.
Wong is head of R. Wang is head of D. Together, with a Fluke,
they make up R&D :-)
Sounds like a limited case scenario. Board manufacturers do
recertify RMA boards and they are resold...
Click to expand...

Wong and Wang, again, eh?
they do not all
get thrown away.
Click to expand...

Who said "all" ?
Did I say all?

I did make a typo. I typed *I* instead of _A_ few RMA's here
and there, but notice, no use of the word "all" :-)
 
Your reply does not state what it does "take", in YOUR *opinion*,
in a large mass production manufacturing plant.
Click to expand...


What are you trying to get at?
This is not a manufacturing plant scenario.
I didn't suggest that ALL boards have every solder joint
tested either, so what's your point?
{snip}


Rare? In mass production? You're too funny!
Click to expand...

So you are just trolling right?
Fact is, most boards that hit retail shelves DO work fine.

Kony, what is your experience inside a mass production, electronics plant,
that you can so surely state "would be rare" ?
Click to expand...

What is your experience with "anything every made in an
automated fashion" if you think otherwise?

That manufacturing problems do occur, and yet that most
equipment does work as intended, is in itself contrary to
your theory that they can't detect these problems.


SNAFU's and FUBAR's happen all the time in mass production.
Click to expand...

Yes they do but we're talking about a board that has shipped
out. Again, what's your point? Driting off-topic isn't
useful unless it's within the context of the thread.

On paper.
Click to expand...

Would you prefer they used whipped cream?


Sure, two guys named Wong and Wang with a VOM.
Wong is head of R. Wang is head of D. Together, with a Fluke,
they make up R&D :-)
Click to expand...

Two guys? LOL.
I can see it's pointless to take your post seriously.


Wong and Wang, again, eh?
Click to expand...

Why do you even bother posting?
Who said "all" ?
Did I say all?
Click to expand...

No but you did only mention them being thrown away. The
other alternative is what you were essentially disputing,
that they have the equipment and personnel to catch
manufacturing problems, correct mistakes, and repair boards.
I suppose you want us to believe it's a lottery if a board
turns out ok? With the number of components on a modern
board NONE of them would work if it were so unprofessional
as you imply.
I did make a typo. I typed *I* instead of _A_ few RMA's here
and there, but notice, no use of the word "all" :-)
Click to expand...

As significant is that it is what you chose to post... being
biased towards (or at least focusing on) only one side of an
argument.

Your goal was to make us believe it's "necessary" to have
some elaborate equipment, that neither someone experienced
in doing repairs nor the board manufacturer themselves could
find such faults. That is clearly wrong.
 
Gary said:
Your reply does not state what it does "take", in YOUR *opinion*,
in a large mass production manufacturing plant.

{snip}




Rare? In mass production? You're too funny!
Kony, what is your experience inside a mass production, electronics plant,
that you can so surely state "would be rare" ?
Click to expand...

Are you two arguing about what it takes to set up and validate a production
line or what ends up in a users hand?

<snip>
 
kony said:
We need a reason to argue? ;-)
Click to expand...

Yes, after sleeping, it does seem that we were arguing.
That is not my intention.

Enough said, I'll put this thread to bed.
 
Kevin said:
Hi, just recently my computer started acting up a year after I completed the
build. I does not want to turn on for me, the power light works, but the HD
light does nothing and no signal is sent to the monitor. The one thing that
fixes it is "unseat" the RAM and then click it back in, and then the
computer fires up. I don't pull the RAM out all of the way, just unseat one
corner of em all(4x512mb) and secure them back in to their slot. What could
possibly be the problem? I'm getting tired of doing this every time.
Thanks in advance for your help.
Click to expand...

I suspect a power problem or a part of the motherboard that shouldn't be
is grounding against the case. Check to make sure your board isn't
making contact anywhere it shouldn't be. The act of taking out a stick
of ram and setting it back in might just be flexing the board enough to
where it solves the problem until the case gets bumped slightly and it
repeats.

I've seen video problems where the traces leading up to them have very
minor cracks and sometimes you would plug in a monitor and it would
work, and sometimes it wouldn't. Sometimes that can happen with RAM and
PCI card slots when you seat and unseat them forcefully or just
frequently. (Every time you install something you flex the board a
little bit to get it in...) That is another possibility and a little
harder to spot. Those can usually be repaired, but unless it's a nice
board or you know how to do it yourself it's usually cheaper to replace
it. (FYI they sell copper tape that is sticky on one side & you can cut
to the shape you need to repair traces and even solder to it. Of course,
you have to get in there with an exacto knife and scrape a bit to get a
good connection before you lay it, but if you do end up buying a new
board and still have the old one you might try to repair it. I've saved
a couple boards this way.)
 
Timbertea said:
I suspect a power problem or a part of the motherboard that shouldn't be
is grounding against the case. Check to make sure your board isn't
making contact anywhere it shouldn't be. The act of taking out a stick
of ram and setting it back in might just be flexing the board enough to
where it solves the problem until the case gets bumped slightly and it
repeats.

I've seen video problems where the traces leading up to them have very
minor cracks and sometimes you would plug in a monitor and it would
work, and sometimes it wouldn't. Sometimes that can happen with RAM and
PCI card slots when you seat and unseat them forcefully or just
frequently. (Every time you install something you flex the board a
little bit to get it in...) That is another possibility and a little
harder to spot. Those can usually be repaired, but unless it's a nice
board or you know how to do it yourself it's usually cheaper to replace
it. (FYI they sell copper tape that is sticky on one side & you can cut
to the shape you need to repair traces and even solder to it. Of course,
you have to get in there with an exacto knife and scrape a bit to get a
good connection before you lay it, but if you do end up buying a new
board and still have the old one you might try to repair it. I've saved
a couple boards this way.)
Click to expand...
Well, it must have been my RAM. I replaced my Corsair TwinX RAM (1GB) with
2GB of Centon RAM and now my machine has no problems, fires up everytime.
Kinda sucks though cause I spent a good dollar or two on the Corsair
thinking I was getting the good stuff. Oh well, I'm really glad I no longer
have the problem. Thank you to everybody who gave me their input.
 
Kevin M said:
Well, it must have been my RAM. I replaced my Corsair TwinX RAM (1GB) with
2GB of Centon RAM and now my machine has no problems, fires up everytime.
Kinda sucks though cause I spent a good dollar or two on the Corsair
thinking I was getting the good stuff. Oh well, I'm really glad I no longer
have the problem. Thank you to everybody who gave me their input.
Click to expand...

Corsair has a lifetime warranty.
 
Back
Top