Survey of PC Failure Mechanisms - Power On

I'd like to know the opinions of Experts:

Situation: A PC starts but it takes more than one press of the power button to get it going.

Possibilities:
1) The power supply
2) The power switch
3) The power switch wire/connector at the mainboard
3) The mainboard

If you had this situation and a new power supply didn't fix it, what would be your first notion in diagnosing it?

This isn't homework.  I see this situation occasionally and have one I'm dealing with right now where I did replace the power supply to no avail.
Please assign a percentage to each adding to 100% .. unless you can think of another
LVL 27
Fred MarshallPrincipalAsked:
Who is Participating?

Improve company productivity with a Business Account.Sign Up

x
 
CallandorConnect With a Mentor Commented:
It's a classic cold boot problem, usually caused by failing capacitors.  The bad capacitors can be in the power supply, motherboard, or video card, and if you replaced the power supply, the next culprit would be the motherboard.
0
 
JohnConnect With a Mentor Business Consultant (Owner)Commented:
For the most part, a power switch is a soft connection (signal only) to the motherboard. So unless the power circuit has its own board, the problem is likely motherboard. That is what I see. I cannot assign a probability.
... Thinkpads_User
0
 
Mohammed RahmanCommented:
@fmarshall.
Please note the room temperature when this hapens. If not possible, let us know if you notice any variation in room temperature between perfect start and more than one press start.
0
Improved Protection from Phishing Attacks

WatchGuard DNSWatch reduces malware infections by detecting and blocking malicious DNS requests, improving your ability to protect employees from phishing attacks. Learn more about our newest service included in Total Security Suite today!

 
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
The Power Supply or Motherboard.

You should be able to rule out the switch, and motherboard connector, using a Metal Link.

Are you just connecting the motherboard without connecting any other devices?
0
 
Fred MarshallPrincipalAuthor Commented:
Looks like we pretty much agree.  Any other experience to share?

thinkpads_user:  Yes, I know that it's a signal / switch closure to the MB function.  And I tend to agree that it's likely not at the top of the list.  This particular one lived in a dusty environment and without designed-in switch "wiping" or some higher level of current, I wonder still.

Callandor:  I don't understand "classic cold boot".  I looked it up but the descriptions I saw were a bit different - having to do with gyrations to "fix" it that had to do with power cycling, etc. etc.
The capacitors look OK so at least there aren't any puffy ones as is often the case.
But, indeed I'm concerned about the mainboard capacitors or not....

I would assign these probabilities myself:

88%    The power supply
10%    The mainboard
1.5%   The power switch
0.5%   The power switch wire/connector at the mainboard

Well, not that these numbers are any good to the nearest 1% at all but their relative size seems appropriate.  Said another way:

Power supply 9 of 10
Mainboard 1 of 10 (I've seen about as many with puffy caps as ones without)
Power switch ... never seen since they went from actual power to a switch closure ... have you?
Wiring / connector ... never seen
But the last two can't have a 0% probability of failure can they?
Thus the choices above.....  


0
 
CallandorConnect With a Mentor Commented:
Capacitors can look perfectly normal and still not function - it's the electrolyte inside that is the key.  I would assign probabilities more like 50% power supply, 48% motherboard, 1% power switch, 0.5% wiring, and 0.5% other.  Power switch failures and wiring can and do happen, but much less likely than the others.
0
 
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
If the motherboard is over 2-2.5 years old, Change the Caps for Good Quality Panasonic or Rubycon units, with low ESR, and High Temps. (the caps used on motherboards somestimes can be cheap Foxcon, nasty ones)
0
 
Fred MarshallPrincipalAuthor Commented:
I once considered doing this myself.  I decided that my board soldering skills and available tools were inadequate.  The cost of having it done is about the same as a new MB.  Suggestions?

Oh!  I forgot to mention the CPU!!
What are views on that?
I once knew a tech who kept CPUs around.  But it seems there are too many of them to warrant inventory, etc.
0
 
Andrew Hancock (VMware vExpert / EE MVE^2)Connect With a Mentor VMware and Virtualization ConsultantCommented:
Personally, I would purchase a new PSU and Motherboard.
0
 
CallandorCommented:
You can send a motherboard to www.badcaps.net to have the work done for less than the cost of a new motherboard.

The probability of a cpu failure is extremely low, and you need a working motherboard in order to test one.  They usually require specific motherboard chipsets and they are obsolete in a short time, so it's not worth it.
0
 
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
$85 plus your Insured shipping costs to destination, I'd purchase a new motherboard! (if your motherboard is 2-2.5 years old).
0
 
Fred MarshallPrincipalAuthor Commented:
Thanks for your views on this!
0
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.