Computer won't turn on right away

[bigredmoose]

Please help!

All of a sudden, my computer won't turn on - at first. When I hit the
power button, nothing happens right away. Sometimes the computer will
start up within seconds, sometimes not for several hours. There appears
to be absolutely no response from the PC when I hit the power button, I
just have to wait and all of a sudden it will turn itself on.

Some notes:

1)The machine will not turn itself on. It will only turn on after I
have pushed the button, just not right away
2)There are no problems whatsoever once the computer is on
3)There is nothing new about the set up, it's in the same place, with
the same peripherals attached

Any help would be greatly appreciated!

Thanks,

Bigredmoose

 

[SteveH]

Please help!

All of a sudden, my computer won't turn on - at first. When I hit the
power button, nothing happens right away. Sometimes the computer will
start up within seconds, sometimes not for several hours. There appears
to be absolutely no response from the PC when I hit the power button, I
just have to wait and all of a sudden it will turn itself on.

Some notes:

1)The machine will not turn itself on. It will only turn on after I
have pushed the button, just not right away
2)There are no problems whatsoever once the computer is on
3)There is nothing new about the set up, it's in the same place, with
the same peripherals attached

Any help would be greatly appreciated!

Thanks,

Bigredmoose

Can you beg/borrow another power supply to try?
Sounds like you may well have a power supply issue.

SteveH

 

[Guest]

All of a sudden, my computer won't turn on - at first. When I hit the
power button, nothing happens right away. Sometimes the computer will
start up within seconds, sometimes not for several hours. There appears
to be absolutely no response from the PC when I hit the power button, I
just have to wait and all of a sudden it will turn itself on.

Some notes:

1)The machine will not turn itself on. It will only turn on after I
have pushed the button, just not right away
2)There are no problems whatsoever once the computer is on
3)There is nothing new about the set up, it's in the same place, with
the same peripherals attached

Any help would be greatly appreciated!

Thanks,

Bigredmoose

Some on/off buttons run through a "tunnel" before coming out of
the front of the computer. Make sure it's not binding or hanging up
inside one.

 

[w_tom]

It could be this. It could be that. It could be due to anything if
we want to fix it before collecting facts. A power supply 'system'
contains many parts. So facts obtained in but two minutes either say
immediately what is happening OR when posted, result in definitive
answers. With information provided, replies can only contain wild
speculation.

Get a 3.5 digit multimeter which is as essential as a screwdriver.
So ubiquitous as to be sold in K-mart, Radio Shack, Home Depot, Sears,
and Lowes. In your case, voltage is measured on a purple wire between
power supply and motherboard when computer is powered off (but
connected to AC wall receptacle). (Notice why AC power wire is removed
before changing anything inside a computer.) Nothing is changed or
disconnected to get a number that must exceed 4.87 (and posted for
further information).

Next, measure green wire both before and when power switch is
pressed. It should be well over 2 volts and then drop to less than 0.8.

And finally, gray wire voltage must be near zero; then rise to well
above 2.4 volts within but seconds after power switch is pressed.

What does your system do? Better information obtained in but minutes
and then posted (especially numbers) means useful responses.

 

[SteveH]

Top posting fixed again

It could be this. It could be that. It could be due to anything if
we want to fix it before collecting facts. A power supply 'system'
contains many parts. So facts obtained in but two minutes either say
immediately what is happening OR when posted, result in definitive
answers. With information provided, replies can only contain wild
speculation.

Get a 3.5 digit multimeter which is as essential as a screwdriver.
So ubiquitous as to be sold in K-mart, Radio Shack, Home Depot, Sears,
and Lowes. In your case, voltage is measured on a purple wire between
power supply and motherboard when computer is powered off (but
connected to AC wall receptacle). (Notice why AC power wire is removed
before changing anything inside a computer.) Nothing is changed or
disconnected to get a number that must exceed 4.87 (and posted for
further information).

Next, measure green wire both before and when power switch is
pressed. It should be well over 2 volts and then drop to less than 0.8.

And finally, gray wire voltage must be near zero; then rise to well
above 2.4 volts within but seconds after power switch is pressed.

What does your system do? Better information obtained in but minutes
and then posted (especially numbers) means useful responses.

Or, if the o/p has a spare PSU to hand, he could spend two minutes plugging
it in and see if it makes any difference.

SteveH

 

[w_tom]

For every minute with a meter, the spare power supply takes far
longer to replace - and only to suspect one part of a 'system'.
Worse, a good power supply in one system may sometimes appear
defective in another system. Was the power supply defective or are
mismatched parameters causing a problem? Ah, shotgunning complete with
its myths and confusions. Shotgunning which completely ignores the
rest of a 'system'.

Once things are changed, than new failures or mysterious 'cures'
appear. More examples of complexity and confusion created by
shotgunning.

A marginal supply can work today; fail tomorrow. Just another reason
why a meter would have found that marginal 'new' supply before it
failed later. Just another reason why the responsible tech confirms
the power supply system with a meter even if computer 'appears' to be
working.

How do we identify a scam auto mechanic? He also shotguns.
Computers are so easily restored a first time with minimal knowledge,
equipment, and time. Waste time swapping a power supply - or first
learn if the power supply 'system' is even problematic. SteveH
assumes the entire power supply 'system' is only a power supply.
But then fixing things based only on speculation is a common mistake;
also used by that scam auto mechanic.

Seen it too often. He replaces the power supply. It works only to
return with problems later. Even worse, many techs try to fix symptoms
of a defective supply including memory and disk replacement. More time
lost because basic numbers were no first obtained - also called
shotgunning. The informed repairman identifies a problem before fixing
it - and in far less time. Shotgunning is performed often by those who
never learned basic technology.

 

[SteveH]

Top posting fixed AGAIN

For every minute with a meter, the spare power supply takes far
longer to replace - and only to suspect one part of a 'system'.
Worse, a good power supply in one system may sometimes appear
defective in another system. Was the power supply defective or are
mismatched parameters causing a problem? Ah, shotgunning complete with
its myths and confusions. Shotgunning which completely ignores the
rest of a 'system'.

Once things are changed, than new failures or mysterious 'cures'
appear. More examples of complexity and confusion created by
shotgunning.

A marginal supply can work today; fail tomorrow. Just another reason
why a meter would have found that marginal 'new' supply before it
failed later. Just another reason why the responsible tech confirms
the power supply system with a meter even if computer 'appears' to be
working.

How do we identify a scam auto mechanic? He also shotguns.
Computers are so easily restored a first time with minimal knowledge,
equipment, and time. Waste time swapping a power supply - or first
learn if the power supply 'system' is even problematic. SteveH
assumes the entire power supply 'system' is only a power supply.
But then fixing things based only on speculation is a common mistake;
also used by that scam auto mechanic.

Seen it too often. He replaces the power supply. It works only to
return with problems later. Even worse, many techs try to fix symptoms
of a defective supply including memory and disk replacement. More time
lost because basic numbers were no first obtained - also called
shotgunning. The informed repairman identifies a problem before fixing
it - and in far less time. Shotgunning is performed often by those who
never learned basic technology.

I tend to find years of experience helps just as much as a multimeter
thanks. And if that doesn't work THEN I'll get the meter out.

SteveH

 

[OSbandito]

w_tom wrote: <...>

t, measure green wire both before and when power switch is

pressed. It should be well over 2 volts and then drop to less than 0.8.

And finally, gray wire voltage must be near zero; then rise to well
above 2.4 volts within but seconds after power switch is pressed.

What does your system do? Better information obtained in but minutes
and then posted (especially numbers) means useful responses.

wtom--you gave me a good idea. rather than depend on running puter
software to tell me var voltages, I might permanently hookup a nice big
LED standalone readout and mount it on top of the tower. A small
gold-contact rotary switch can allow me to select the particular PS output.

You indicate, I think, a 3% tolerance on the DC out. Do you know if this
is typical for the various mb's?

 

[w_tom]

wtom--you gave me a good idea. rather than depend on running puter
software to tell me var voltages, I might permanently hookup a nice big
LED standalone readout and mount it on top of the tower. A small
gold-contact rotary switch can allow me to select the particular PS output.

You indicate, I think, a 3% tolerance on the DC out. Do you know if this
is typical for the various mb's?

Numbers are selected first from industry standard limits as even
defined in Intel's ATX standards. Numbers are above those Intel values
because of how meters work during high ripple voltage.

As voltage reduces, some ICs slowly become intermittent - also called
a computer crash. Other peripherals (memory, disk drive, video
controller) then act defective. Some techs fix these problems by
shotgunning those components.

Some motherboards contain voltage monitor hardware. A monitor will
detect voltage variations - changes. But to obtain a useful voltage
number, a meter is required to calibrate that monitor. Then monitor
alarm levels are set to warn when voltages drop excessively - and
before low voltage causes a crash.

In high reliability operations, voltage displays are not used.
Instead, each voltage is providedd at test sockets. A tech confirms
voltages by simply touching his meter to each socket - thereby
obtaining a better number than the display might provide.

 

[Rod Speed]

OSbandito wrote

Numbers are selected first from industry standard limits as even
defined in Intel's ATX standards. Numbers are above those Intel
values because of how meters work during high ripple voltage.

Pity the ripple allowed is defined in the ATX standard.

As voltage reduces, some ICs slowly become intermittent
- also called a computer crash. Other peripherals (memory,
disk drive, video controller) then act defective.

Pig ignorant waffle.

Some techs fix these problems by shotgunning those components.

And fools like you wank with their multimeters.

Some motherboards contain voltage monitor hardware.
A monitor will detect voltage variations - changes. But
to obtain a useful voltage number, a meter is required to
calibrate that monitor. Then monitor alarm levels are set
to warn when voltages drop excessively - and
before low voltage causes a crash.

Pathetic, really.

In high reliability operations, voltage displays are not used.
Instead, each voltage is providedd at test sockets. A tech confirms
voltages by simply touching his meter to each socket - thereby
obtaining a better number than the display might provide.

Pathetic, really.

 

[SteveH]

Pity the ripple allowed is defined in the ATX standard.


Pig ignorant waffle.


And fools like you wank with their multimeters.


Pathetic, really.


Pathetic, really.

I suppose we 'should' feel sorry for Mr Tom. He's probably had training in
electronics many years ago, which has probably served him well when fixing
radios etc. However he can't now accept that us young whippersnappers might
just do things a bit differently. And that however differently we do things
it works for us, and I would guess, most of the PC industry.
While I do indeed possess (as I would guess a few others here do) the
ubiquitous 3.5 digit bloody multimeter, I tend to rely on my years of
experience building and repairing PC's and only get the meter out when I
really 'have' to.

SteveH

 

[Rod Speed]

I suppose we 'should' feel sorry for Mr Tom.

The problem is that he misleads so many that dont know
enough to realise that what he spews is so utterly wrong.

He's probably had training in electronics many years ago, which has probably served him well when
fixing radios etc. However he can't now accept that us young whippersnappers might just do things
a bit differently.

The problem appears to be that he's just another
bone head who cant manage to get the basics right.

And I'm not a young wippersnapper, I'm likely older than he is.

And that however differently we do things it works for us, and I would guess, most of the PC
industry.

While I do indeed possess (as I would guess a few others here do) the ubiquitous 3.5 digit bloody
multimeter, I tend to rely on my years of experience building and repairing PC's and only get the
meter out
when I really 'have' to.

Yeah, but that applys in spades with a CRO too.

The modern reality is that trying a spare power supply is quite a bit
more efficient and effective than farting around with a multimeter,
essentially because that whole area is quite complex and its generally
only worth bothering with a multimeter if the results with the spare
power supply makes no sense and you need to prove that there is
a motherboard fault before discarding the system or replacing it.

A multimeter can make a lot of sense with say USB powered
devices where its not clear if the port can supply the power
that its supposed to be able to supply etc.

 

[Guest]

I suppose we 'should' feel sorry for Mr Tom. He's probably had training in
electronics many years ago, which has probably served him well when fixing
radios etc. However he can't now accept that us young whippersnappers might
just do things a bit differently. And that however differently we do things
it works for us, and I would guess, most of the PC industry.
While I do indeed possess (as I would guess a few others here do) the
ubiquitous 3.5 digit bloody multimeter, I tend to rely on my years of
experience building and repairing PC's and only get the meter out when I
really 'have' to.

SteveH

Agreed! Much of what he says has merit, but w_tom is a notorious,
single-minded babblebot that occasionally gets his dick stepped on.
The time needed for the OP to ascertain if the button is sticking, a
well-known cause for the symptoms as described, can be measured
in mere seconds and not much longer for the fix, if needed, of a tiny
shot of WD-40 or something similar. This doesn't require the OP to
purchase and learn to use a meter if he doesn't already have one, or
even to open the case when he may have never been in one before....
sheesh! I do wish more people would post the final resolution of their
problem though.

 

[OSbandito]

Numbers are selected first from industry standard limits as even
defined in Intel's ATX standards. Numbers are above those Intel values
because of how meters work during high ripple voltage.

As voltage reduces, some ICs slowly become intermittent - also called
a computer crash. Other peripherals (memory, disk drive, video
controller) then act defective. Some techs fix these problems by
shotgunning those components.

Some motherboards contain voltage monitor hardware. A monitor will
detect voltage variations - changes. But to obtain a useful voltage
number, a meter is required to calibrate that monitor. Then monitor
alarm levels are set to warn when voltages drop excessively - and
before low voltage causes a crash.

In high reliability operations, voltage displays are not used.
Instead, each voltage is providedd at test sockets. A tech confirms
voltages by simply touching his meter to each socket - thereby
obtaining a better number than the display might provide.

Thx for a useful and interesting answer.