Computer won't turn on

[brady.sondreal]

My home-built PC has worked fine for a year, but now it won't power on.
There is a red light on the motherboard which i assume means it is
getting power.
I bought a brand new power supply and still the thing won't turn on.
The fans would spin and the hard drive would churn the first few times
i tried turning it on, but now nothing happens, other than the red
light on the motherboard.

I'm using an 24 pin connector with the additional 4 pin connector and
both are plugged in.
I reset my CMOS with the jumper and pulled the battery, but still
nothing powers up.
I tried wiring the reset button to the power on the mobo because i
thought maybe the wires to the power button were faulty, but that
didn't help.

NOTE: My old power supply has things rattling around in it and has a
burnt smell.

Could my orginal PSU have fried my mobo?
Would it have fried the Chip too?
Could the battery on my mobo be dead? I ran a mutli-meter on it and it
gave a reading of 3 volts.

Thanks for your help.

 

[philo]

NOTE: My old power supply has things rattling around in it and has a
burnt smell.

Could my orginal PSU have fried my mobo?
Would it have fried the Chip too?
Could the battery on my mobo be dead? I ran a mutli-meter on it and it
gave a reading of 3 volts.

Well...
your old supply is bad for sure...and it sure sounds like it took out your
motherboard too.

I've seen that happen after lightning storms.

In the machines that I;ve repaired that had been hit by a voltage spike...
the good news is that the CPUs and harddrives were still ok...
but of course you could have more damage than just the psu and mobo

 

[Rod Speed]

(e-mail address removed) wrote

My home-built PC has worked fine for a year, but now it won't power
on. There is a red light on the motherboard which i assume means it is
getting power.

I bought a brand new power supply and still the thing won't turn on.
The fans would spin and the hard drive would churn the first few times
i tried turning it on, but now nothing happens, other than the red
light on the motherboard.

I'm using an 24 pin connector with the additional 4 pin connector and
both are plugged in.
I reset my CMOS with the jumper and pulled the battery, but still
nothing powers up.
I tried wiring the reset button to the power on the mobo because i
thought maybe the wires to the power button were faulty, but that
didn't help.

NOTE: My old power supply has things rattling around in it and has a
burnt smell.

Thats the important bit.

Could my orginal PSU have fried my mobo?
Yes.

Would it have fried the Chip too?
Unlikely.

Could the battery on my mobo be dead?
Nope.

I ran a mutli-meter on it and it gave a reading of 3 volts.

Thats fine.

 

[w_tom]

Power supplies are required to contain circuits so that it cannot
harm a motherboard and so that motherboard cannot harm power supply.
But when selling power supplies only on dollars and watts, then many
clone power supplies are dumped into the market missing required
functions.

Trying to fix the problem only by swapping parts is called
shotgunning. Informed repair starts with less than 2 minutes and your
3.5 digit multimeter.

In your case, measurements start with voltage on purple wire (from
power supply to motherboard). That voltage must exists whenever power
supply is connected to AC mains. More important is value of that number
- must exceed 4.87 volts.

Then move on to green and gray wire. What happens to green wire
before and during press of power switch? That voltage must exceed 2.0
volts and then drop well below 0.8 volts. What does gray wire do as
power switch is pressed? Starts at near zero volts and rises to what?

Numbers from those reading go a long way to getting a useful reply.
Or you could just keep replacing parts until everything is replaced or
until something finally starts working. Use the meter so that your
posts include useful numbers; so that replies will be useful. Light
can illuminate and still power is 100% defective. We don't know until
those numbers are provided.

Due to design, CPUs are rarely damaged. But at this point, you are
only looking at the power supply 'system'. Power supply is only one
component of that 'system'.

 

[brady.sondreal]

Thanks for all of your help. I really appreciate it.

I have tried a brand new power supply and still the issue remains.
I will try taking readings of the wires suggested in the previous post
just to make sure i don't have 2 bad power supplies.

Should i take these readings with the power supply connected to the
mobo or should i disconnect the power supply from everything, put a
saftety pin in green and black, and check the readings from there?

If the readings are fine, should I swap out my motherboard with another
Socket 939 from my friend?

 

[JamesG]

I just fixed a friend's computer that seems to have had a similar
situation. The motherboard was dead. There was power getting to the
MB and the cpu fan would run but it wouldn't POST or do anything else.
I replaced the power supply and the motherboard and it booted up no
problem. I used the old CPU and it still seems to be good.

Good Luck,
James

 

[w_tom]

Take those reading with everything connected. Don't disconnect
anything to find a suspect. That was a major point in the earlier
post. Remember that reference to a "power supply system"? Described
previously was but less than a minute of a two minute analysis.

Disconnecting to collect facts is called shotgunning. Don't shotgun.
Don't start changing things to somehow (miraculously) make it work.
That can exponentially complicate a problem. Get numbers. Do not
disconnect anything until, first, facts are collected and suspects
identified. Do not even ask what to replace until those first numbers
say what specifically to disconnect or change.

When we finally disconnect or swap anything, it is because we have a
specific suspect. If numbers are in spec, we have only started a 'step
by step' procedure to find the problem. Why would you even suggest
swapping a motherboard if those numbers are good? Good numbers say we
have not yet found a problem - therefore have not yet identified the
prime suspect. On CSI, they say, "Follow the evidence". Your 'swap
motherboard' proposed completely violates basic analysis. When we
replace motherboard, a specific number(s) says this is the suspect.
Currently there are too many suspects to disconnect or swap anything
and too many possibilities to ask what to do if this number is good or
that number is bad.

Swap something only because a technical fact says 'this' has failed.

If the readings are fine, should I swap out my motherboard ...

says you are still in shotgun mode. Good numbers from that first less
than 1 minute of readings tells us little, yet. Those numbers tell us
where next to collect facts. Follow the evidence. Never shotgun.

 

[Rod Speed]

Ignore this clown, he has never ever had a clue
about fault finding, or anything else at all, either.

If you find that hard to believe, check the responses to his drivel using groups.google.

 

[brady.sondreal]

w_tom - If the readings on the power supply show that it is working
correctly, why wouldn't i swap out the motherboard? Those readings
would prove that that power system is working.

 

[JamesG]

I just fixed a friend's computer that seems to have had a similar
situation. The motherboard was dead. There was power getting to the
MB and the cpu fan would run but it wouldn't POST or do anything else.
I replaced the power supply and the motherboard and it booted up no
problem. I used the old CPU and it still seems to be good.

Good Luck,
James

Okay maybe I forgot to mention that I also checked that it wasn't
solely the power supply by trying out a new power supply since the old
one was not good quality anyway. The old, dead motherboard still did
nothing. At this point I really didn't see the point in poking around
with a multimeter since it seemed obvious that the motherboard was dead.

 

[Davy]

A blown PSU may not have damaged the mobo, if you get a fault on the
feedback loop, usually a opto-coupler measuring some secondary supply
voltage... ka-pow it'll go big style and could quite easily damage the
mobo since regulation of the chopper device will turn on hard and
harder until destruction elevating the supply voltages....

More than likely the rattling you can hear is whats left of a
capacitor that has exploded, the rattling being caused by the outer
casing of it. What happens is that its internal ESR rises this
creates heat which causes gassing.. if they are not vented... or the
vent cannot cope with the increased pressure.

I would be tempted to try by substitution... a test meter may read the
voltages correctly but it will not show any ripple on the supplies...
ripples produced by the switching frequency (around 18~22Khz) will
upset the applecart usually caused by deteriorating capacitors.

You may have removed the bios battery but is it any good, did you try
a new one..?

Davy

 

[w_tom]

You are assuming the 'system' is only two components. If those three
wires (purple, gray, and green) measure correctly before and during
power switch press, then power supply controller on motherboard is not
defective. And since we did not see a failure, then we still don't
know where the problem is.

However, based upon earlier posts, I do not expect all three wires to
perform properly. IOW 'which wire does what' determines where to look
next. Also, once numbers are provided, then text will follow that
explains what is happening; what is known at this point. Information
that makes future problems solvable even faster.

 

[brady.sondreal]

The problem is fixed. Thanks to everyone who helped.

1. Swapped battery - Problem remained
2. Swapped CPU - Problem remained
3. Swapped Mobo - Problem resolved.

Apparently the power supply took out the motherboard.

 

[w_tom]

Apparently the power supply took out the motherboard.

If power supply took out motherboard, then power supply is defective.
Power supply cannot cause damage to computer peripherals. That
standard requirement existed even 30+ years ago.

It is possible that power supply took out motherboard due to a
computer assembler without basic electrical knowledge. Manufacturers
are dumping inferior supplies at even higher profits into a market of
naive computer assemblers. These supplies do not have functions that,
for example, would make motherboard damage impossible. IOW motherboard
damage would be directly traceable to human ignornace.

If a power supply does not provide a long list of numerical specs,
then suspect the worst. Power supplies that are missing essential
functions (including motherboard protection) are routinely sold to the
naive at higher profits and lower prices.

 

[brady.sondreal]

w_tom

"It is possible that power supply took out motherboard due to computer
assembler without basic electrical knowlegdge".

The computer worked for over a year, so obviously this wasn't the
problem. How much electrical knowledge do you need to plug in a 24 pin
and a 4 pin adapter into a slot on the motherboard?

Furthermore, you indicated that the power supply could have been
faulty. Why then do you attribute this problem to the basic electrical
knowledge of the computer assembler?

Rod Speed - You were sure right about this w_tom guy.

Thanks to everyone who posted, other than w_tom.

 

[w_tom]

Power supply must contain functions that are missing in discounted
power supplies. "If power supply took out motherboard, then power
supply is defective." Missing function would not be apparent until
one year later when power supply fails. If brady_sondreal's power
supply damaged his motherboard, then it took one year for missing
functions to be obvious and expensive. Apparently brady bought a
supply only by counting 24 connector contacts.

An abridged example of what a minimally acceptable power supply
does - which is more than 24 contacts:
Specification compliance: ATX 2.03 & ATX12V v1.1
Short circuit protection on all outputs
Over voltage protection
Over power protection
100% hi-pot test
PFC harmonics compliance: EN61000-3-2 + A1 + A2
EMI/RFI compliance: CE, CISPR22 & FCC part 15 class B
Safety compliance: VDE, TUV, D, N, S, Fi, UL, C-UL & CB
Hold up time, full load: 16ms. typical
Efficiency; 100-120VAC and full range: >65%
Dielectric withstand, input to frame/ground: 1800VAC, 1sec.
Dielectric withstand, input to output: 1800VAC, 1sec.
Ripple/noise: 1%
MTBF, full load @ 25°C amb.: >100k hrs

Functions even in that short list are simplistic - what a minimally
educated computer assembler would know.

Brady tells us his motherboard was damaged by his power supply.
Therefore functions from that above list were missing in his supply.
Since brady did not know, he posts insults. Others can learn from
brady's mistake and resulting motherboard damage; apparently traceable
to his attitude.

If power supply damages a motherboard, then power supply was
defective when purchased.

 

[Rod Speed]

(e-mail address removed) wrote

w_tom wrote

"It is possible that power supply took out motherboard due
to computer assembler without basic electrical knowlegdge".

The computer worked for over a year, so obviously this wasn't the
problem. How much electrical knowledge do you need to plug in
a 24 pin and a 4 pin adapter into a slot on the motherboard?

 

[Rod Speed]

Power supply must contain functions that are missing in discounted
power supplies. "If power supply took out motherboard, then power
supply is defective." Missing function would not be apparent until
one year later when power supply fails. If brady_sondreal's power
supply damaged his motherboard, then it took one year for missing
functions to be obvious and expensive. Apparently brady bought a
supply only by counting 24 connector contacts.

An abridged example of what a minimally acceptable power supply
does - which is more than 24 contacts:
Specification compliance: ATX 2.03 & ATX12V v1.1
Short circuit protection on all outputs
Over voltage protection
Over power protection
100% hi-pot test
PFC harmonics compliance: EN61000-3-2 + A1 + A2
EMI/RFI compliance: CE, CISPR22 & FCC part 15 class B
Safety compliance: VDE, TUV, D, N, S, Fi, UL, C-UL & CB
Hold up time, full load: 16ms. typical
Efficiency; 100-120VAC and full range: >65%
Dielectric withstand, input to frame/ground: 1800VAC, 1sec.
Dielectric withstand, input to output: 1800VAC, 1sec.
Ripple/noise: 1%
MTBF, full load @ 25°C amb.: >100k hrs

Functions even in that short list are simplistic - what a minimally
educated computer assembler would know.

Brady tells us his motherboard was damaged by his power supply.
Therefore functions from that above list were missing in his supply.
Since brady did not know, he posts insults. Others can learn from
brady's mistake and resulting motherboard damage; apparently traceable
to his attitude.

If power supply damages a motherboard, then power supply was
defective when purchased.

Irrelevant to your stupid claim about him having killed the power
supply by incorrect assembly, AND your even more stupid claim
about how to effectively diagnose what had actually failed.

 

[brady.sondreal]

My power supply was cheap w_tom. I understand more expensive power
supplies may have fewer problems, but I was building a budget-friendly
PC.

My motherboard was damaged as a result of my attitude according to Tom.
I will from now on always suspect my attitude to be the root of any
problem I'm experiencing with a PC. Thanks Tom. I'm enlightened.

If anyone is at work right now and wants to kill some time and have a
good laugh while doing so, please check out w_tom's other posts.

 

[w_tom]

brady - still you remain emotional rather than grasp technical
reality. Is your real name Rumsfeld? Just because a power supply is
expensive does not mean it is better. Minimally acceptable power
supplies claim a long list of functions. A responsible computer
assembler does not buy on price. Insulting others does not correct a
classic 'bean counter' mistake.

One required function means a power supply will not damage
motherboard, disk drives, video card, keyboard, memory, etc. Now you
must replace motherboard. Being cheap saved you how much money? Why
then do you insult? Why do you deny technical reality? If a power
supply damaged a motherboard (and you post that only in speculation),
then the power supply was defective by design. Failure directly
traceable to the computer assembler. Such a defensive attitude implies
that 'bean counter' computer assembler is brady.

How much money did you save on a cheap power supply now that
motherboard must be replaced after only one year? The concept is
called overstress. Other components may also is fail. brady
demonstrates how to not build a computer and how to not learn from
mistakes.