I need your advice - new build: fans and lights, "no signal", nopost beep

  • Thread starter Thread starter ToolPackinMama
  • Start date Start date
Matt said:
May be a dumb question, but how are you getting this "FF" code if the
monitor is not getting a signal and you are getting no beeps?
Click to expand...

There is a display on the motherboard
 
ToolPackinMama said:
OK I did install the CPU double triple extra carefully, because I don't
ever want to be guilty of not seating/installing it properly. I used
the included heatsink/fan which comes with heatsink goo already on it.

The CPU was firmly, properly in place and latched down... I CHECKED, I'M
VERY CAREFUL...the heatsink/fan installed without any complications.

Later, after I read some comments here, I pulled the machine apart and
when I went to lift the heatsink off, the CPU came with it. It came
away easily, I wasn't violent. Stuck fast together - and the ZIF latch
is still down in locked position.

I thought "Oh! That can't be good!"
Click to expand...


Nothing wrong with that unless you bent some pins. But if you do want
to remove the HSF from the CPU, maybe you should try twisting it off
instead of pulling straight up.
 
westom1 said:
ToolPackinMama <philnbl... comcast.net> wrote:

Does not matter. It is a power supply 'system' - not just a power
supply.
Click to expand...

We refer to it as a "power supply".
Get the meter.
Click to expand...

Hello w_tom.











See also:
Google Groups
w_tom
westom1 gmail.com
 
Later, after I read some comments here, I pulled the machine apart and
when I went to lift the heatsink off, the CPU came with it. It came away
easily, I wasn't violent. Stuck fast together - and the ZIF latch is still
down in locked position.

I thought "Oh! That can't be good!"
Click to expand...

Defective ZIF latch? Well that would explain it. You can expect the CPU
and heatsink to get kind of "glued" together. But if the CPU came up with
little effort, that's not normal.

When you installed the CPU, did you have to put pressure on the ZIF latch to
push it down? Not a lot of pressure, but there should have been some
resistance. -Dave
 
Matt said:
Gosh, isn't there someplace to look up the meaning of the code?
Click to expand...

A good manual will include the codes.

If you don't have codes, go here for generic ones.

(BIOS PostCodes, upper left)
http://www.bioscentral.com/

Very few documentation efforts capture all the codes.
Which is why when some people catch a code on the
display, looking it up returns "reserved" for their
troubles. It means someone didn't bother to document
it.

At least one code value has to be reserved as an
initial value for the display, so you can tell
when the processor is dead. 0xFF is an example.

Paul
 
Dave said:
When you installed the CPU, did you have to put pressure on the ZIF
latch to push it down? Not a lot of pressure, but there should have
been some resistance.
Click to expand...

Yeah everything seemed normal at the time.
 
Just keep in mind (and this is why I advise people NOT to test with a
multimeter/voltmeter) that 90% or more of DEFECTIVE power supplies will test
GOOD with a multimeter. That's because a (GOOD!) power supply doesn't just
provide stable DC voltage for a fraction of a second...or a few seconds, or
even a few MINUTES. So any reading you take with a multimeter will only
tell you that the power supply is not dead.
Click to expand...

Meanwhile an engineer who even designed power supplies says
otherwise. Dave’s only recommendation is to keep replacing things
until something works. Circuits such as power supply filters and
other specifications mean his mythical power drop cannot (and does
not) happen. But then Dave has no solution other than to keep
replacing parts on “it could be this or could be that” speculations.
Dave fears what he does not understand such as multimeters. We
sometimes get a tech who does that same thing. Either he learns how
to solve problems or he is quickly unemployed (as Dave would be).

A multimeter is the only tool (other than $thousand equipment) that
can report a power supply - definitively. How many others only post
speculation or personal attacks rather than a solution? "It might be
this or might be that" is called shotgunning. If does not work as
demonstrated by the next sentence.

Normal is for a computer to boot with a defective power supply. And
normal is for the meter to identify that same supply as defective.
How would I know that? Our techs routinely demonstrated reality using
a multimeter. Sometimes a complete system failure is averted by
identifying a slowly getting worse power supply. Only the meter could
identify the defect. After all, a computer may still boot with a
defective supply.

How important is a meter? It is a first tool that even an auto
mechanic students get and use. Even car mechanics need a meter to
find and fix simple problems. Even auto mechanics would quickly be
unemployed if they used Dave’s diagnostic recommendations.

Until the many components of the power supply ‘system’ are verified
(a minute of labor), then no other labor will be productive – generate
‘definitive’ answers. Yes, other good parts can act defective if a
power supply system is not first identified. Ie. Paul’s ‘pull the
ram’ beep test.

Ignore the insults and claims based only in 'it could be this'
logic. Get the meter for replies from the minority who actually know
how computers work.
 
Later, after I read some comments here, I pulled the machine apart and
when I went to lift the heatsink off, the CPU came with it. It came
away easily, I wasn't violent. Stuck fast together - and the ZIF latch
is still down in locked position.
Click to expand...

Appreciate what a heatsink and thermal compound do. Thermal
compound does not hold a heatsink to an IC (ie CPU). The best thermal
transfer (also called low thermal resistance) is a one medium
transfer. That means direct contact between CPU and heatsink. That
means pressure must hold a heatsink to CPU.

Minimally acceptable heatsinks are properly machined (also not
perfectly flat) so that the majority of two surfaces - CPU and
heatsink - make direct contact. This is more than sufficient for
cooling any processor. Lower that thermal resistance by filling
microscopic gaps with a thermal compound. Resistivity from CPU
through thermal compound to heatsink is higher - a two medium
transfer. But that higher resistance in parallel with direct 'CPU to
heatsink' conductive paths means an overall slightly lower thermal
resistance.

Too much thermal compound only increases thermal resistance - makes
a heatsink less efficient. Thermal compound is applied so sparingly
(a small pea) that only the center half of a 'CPU to heatsink'
interface has compound. Compound must only fill microscopic gaps. So
little thermal compound means more direct 'CPU the heatsink' contact.

If any thermal compound appears at the 'CPU to heatsink' edge, you
have applied far too much and may create additional electrical
problems.

Meanwhile that heatsink interface has no relevance to your problem.
A common myth often promoted by many who otherwise do not know what to
do next. Many will also promote another myth about 'dried' thermal
compound. Any minimally acceptable heatsink with no thermal compound
would provide sufficient cooling even for a decade..

Removing a heatsink can only complicate a challenging problem and
will do nothing to solve it. This is not popular among the many due to
widespread technical naivety. If heat was creating your problem, the
computer would boot and run for many seconds before just shutting
down. That is not your symptom.

Explained is how to minimize thermal compound and why for the best
heatsink interface - for a lowest thermal resistance.

Until the power 'system' (not just a power supply) is known good,
then no other tasks will be productive. The only productive solution
leaves you with a 'definitively good' or 'definitively bad' answer.
Anything less is wasted labor. Without removing or disconnecting
anything, verify all power 'system' components are functional. A
minute of labor. Only then move on to other suspects. Heatsink was
never on that list of suspects.

Paul's discussion of beep code by "pulling the ram” is one of the
few posts that provides useful information. No beep code with memory
removed? However your FF already told us what Paul's test will
report. His test only confirms it. . Again, until many components of
the power supply system are verified (one minute of labor) then all
other solutions are called 'pissing in the wind' complete with adverse
consequences. No beep code still leaves the same list of suspects -
which obviously never included a CPU heatsink.
 
westom1 said:
"Dave" <no... nohow.not> wrote:

Meanwhile an engineer who even designed power supplies says
otherwise.
Click to expand...

You misspelled "a blabbering idiot like Tom who doesn't even know
how surge suppressors work but pretends to be a design engineer".
Dave's only recommendation is to keep replacing things until
something works.
Click to expand...

Intelligently swapping parts is a good way to troubleshoot a modular
device like a personal computer, Tom.
But then Dave has no solution other than to keep replacing parts
on "it could be this or could be that" speculations. Dave fears
what he does not understand such as multimeters. We sometimes
get a tech who does that same thing. Either he learns how to
solve problems or he is quickly unemployed (as Dave would be).
Click to expand...

Something similar happens here on USENET... Tom... but you keep
coming back.
A multimeter is the only tool (other than $thousand equipment)
that can report a power supply - definitively. How many others
only post speculation or personal attacks rather than a solution?
Click to expand...

That's because you are a hellbent troll who searches USENET for your
pet peeves (example "power supply" and "surge suppressor") and then
jumps into the group to spam your warped ideas, Tom.
"It might be this or might be that" is called shotgunning.
Click to expand...

Intelligently swapping parts is a well accepted troubleshooting
method around here, Tom.
Normal is for a computer to boot with a defective power supply.
And normal is for the meter to identify that same supply as
defective.
Click to expand...

Average users should not be poking around inside of a live computer
with metal probes.
How would I know that?
Click to expand...

How you would know anything is a mystery, Tom.
Our techs routinely demonstrated reality using a multimeter.
Click to expand...

I invented a voltmeter. Besides being a nym-shifting troll
pretending to be an engineer and spreading weird ideas here on
USENET, what exactly are you Tom?
 
Matt said:
ToolPackinMama wrote:


Nothing wrong with that unless you bent some pins. But if you do want
to remove the HSF from the CPU, maybe you should try twisting it off
instead of pulling straight up.
Click to expand...

:)
 
Dave said:
Defective ZIF latch? Well that would explain it. You can expect the
CPU and heatsink to get kind of "glued" together. But if the CPU came
up with little effort, that's not normal.
Click to expand...

OK thank you, that's what I thought.
 
ToolPackinMama said:
Now you tell me! ;)
Click to expand...


Well, I'm expecting you've had a lot of experience (besides common
sense) regarding what could cause harm. Just have some respect for the
thing, and we'll be fine. :-)
 
Matt said:
ToolPackinMama wrote:
....



Nothing wrong with that unless you bent some pins. But if you do
want to remove the HSF from the CPU, maybe you should try twisting
it off instead of pulling straight up.
Click to expand...

In other words, she said that the CPU came out of the ZIF socket
without releasing the ZIF latch. Maybe the CPU was just sitting on
top of the socket without being inserted. Maybe the socket was not
gripping the CPU pins. Whatever the case, something was very wrong.
 
Well, I'm expecting you've had a lot of experience (besides common
sense) regarding what could cause harm.
Click to expand...

Of course, since a CPU heatsink is not even on a list of suspects,
then there was no reason to remove it. Now the problem may be
exponentially complicated by making changes only on speculation.
 
Of course, since a CPU heatsink is not even on a list of suspects,
then there was no reason to remove it.
Click to expand...

Since the heatsink has to be removed to check that the CPU is correctly
seated in the socket, you are yet again wrong.

You must be very, very stupid to believe that the purpose of removing
the heatsink is to check the heatsink itself.

Do you know anything at all about building and diagnosing modern PCs? *
 
Mike Tomlinson said:
googlegroups.com, (e-mail address removed) writes
....

Do you know anything at all about building and diagnosing modern PCs?
Click to expand...

Tom found this thread in a USENET archive (Google Groups) search for
"PSU" "power supply" or whatever. To spread his silly ideas about power
supplies and surge suppressors on USENET is the only reason he's here.
Maybe he's trying to promote something, or maybe there isn't any normal
reason. It's a mystery.
 
FWIW, I RMA'd the motherboard. Defective CPU socket latch. Thing
couldn't securely hold the weight of the CPU/Heatsink when the machine
was upright.
 
Back
Top