Jump to content


 


Register a free account to unlock additional features at BleepingComputer.com
Welcome to BleepingComputer, a free community where people like yourself come together to discuss and learn how to use their computers. Using the site is easy and fun. As a guest, you can browse and view the various discussions in the forums, but can not create a new topic or reply to an existing one unless you are logged in. Other benefits of registering an account are subscribing to topics and forums, creating a blog, and having no ads shown anywhere on the site.


Click here to Register a free account now! or read our Welcome Guide to learn how to use this site.

Photo

Unable to start up PC


  • Please log in to reply
16 replies to this topic

#1 geewin

geewin

  • Members
  • 12 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Brasil
  • Local time:02:53 PM

Posted 22 August 2009 - 11:17 AM

Hi, your advice please.
First the details:
Gigabyte GA-P35-DS3
Intel E8400
2GB Kingston DDR800
evga GeF8800GT
160GB SATA2 Caviar HDD
Satellite 600W PSU
XP Pro SP3
The machine is 18 months old. I leave the PC on most of the time, but I have previously had some `beeps`on cold startup, which could indicate power, memory or graphics problems (according to which manual - English or Portuguese - you read! I live in Brasil by the way.), but these soon stop and the machine works very well. However, yesterday I told the PC to restart, just to free some memory, and instead of restarting it switched on, then off, then on again but no POST or BIOS - nothing! The fans were all running and the drives open and close etc. but no POST/BIOS and the monitor light was blinking. I have re-seated everything, changed Voltage stabilizers, cables, monitors, everything, but no success.
The machine switches on for about two seconds, then off for the same time, and on again once or twice before it stays on but nothing on the monitor. The only conclusion I could come to is that the PSU is faulty; now this would be no surprise being a crap make, but when I tried with a different one (an old 350Watt) I got the same result! Twice during all this testing the machine started up normally, but with a continuous beeping, once with each PSU!
I still think PSU but it also seems a big coincidence that 2 PSUs should exhibit the same symptoms, and I wonder if it could be the M/B. I want to be certain before spending (parts are 100%+ more expensive here), so I would like your thoughts please.

BC AdBot (Login to Remove)

 


#2 hamluis

hamluis

    Moderator


  • Moderator
  • 55,900 posts
  • ONLINE
  •  
  • Gender:Male
  • Location:Killeen, TX
  • Local time:12:53 PM

Posted 22 August 2009 - 12:13 PM

It could be the CMOS battery, I would have replaced that before anything else.

PSU would be 2d suspect.

I guess I have to wonder...if you routinely employ a 600w PSU...why test it with a 350W? Especially so if you think the 600W PSU doesn't cut the mustard?

Louis

#3 geewin

geewin
  • Topic Starter

  • Members
  • 12 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Brasil
  • Local time:02:53 PM

Posted 22 August 2009 - 03:27 PM

Hi Louis,
thanks for the response. I used the 350 PSU because its all I have, and the machine has run OKwith it before. As for the 600W itīs not that it doesnīt "cut the mustard", its that it is a SolyTech product and as you have said before, "not to be trusted". I will buy a Corsair PSU if necessary to replace it, but I donīt want to do that and then find that it was something else, the M/B for instance? Iīll remove and test the battery, but I doubt itīs that as tha M/B is only 18 months old. Also I was told that a PC will still boot even if its dead; is that true?
regards, Geoff

#4 geewin

geewin
  • Topic Starter

  • Members
  • 12 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Brasil
  • Local time:02:53 PM

Posted 22 August 2009 - 03:44 PM

Hi again, just checked CMOS battery its OK 3.3V
Geoff

#5 hamluis

hamluis

    Moderator


  • Moderator
  • 55,900 posts
  • ONLINE
  •  
  • Gender:Male
  • Location:Killeen, TX
  • Local time:12:53 PM

Posted 22 August 2009 - 04:58 PM

Well...I've never heard of testing a CMOS battery, any more than I would test a AA or AAA battery. But I'm here to learn :thumbsup:.

I find that time is no indicator whatever...of how long a battery will last, regardless of type.

As for a "dead PC" booting...that's a new one, also. I guess it's all in how one defines "dead" or whether one believes in returning from the grave.

Louis

#6 geewin

geewin
  • Topic Starter

  • Members
  • 12 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Brasil
  • Local time:02:53 PM

Posted 23 August 2009 - 04:49 PM

Well Louis,
firstly any kind of battery can be tested with a voltmeter. Secondly the word īdeadī in my last sentence refered to the battery, but I guess you knew that and were attempting a joke!
More importantly I have a real problem and was hoping for some helpful advice, like most people who write into these forums. Where I live there are no reliable computer technicians so I have to do what I can myself, and I donīt want to spend money unnecessarily. What I donīt understand is why you feel the need to mock instead of trying to answer my questions.
Thanks

#7 AustrAlien

AustrAlien

    Inquisitor


  • Members
  • 6,772 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Cowra NSW Australia
  • Local time:03:53 AM

Posted 23 August 2009 - 08:53 PM

geewin

Please do not take offense: I see no "mocking" by hamluis. I am quite sure there is no such "mocking" intent in hamluis' comments.
I certainly did not understand what you meant by "Also I was told that a PC will still boot even if its dead; is that true?", and I am guessing the same applied to hamluis. As for the comment by hamluis "Well...I've never heard of testing a CMOS battery", I suspect you have enriched his knowledge in this respect: Dealing with a good number of old machines, I routinely test CMOS batteries with a voltmeter, discarding those showing less than 3.0 volt.
------------
May I suggest, for the purpose of trouble-shooting, that you disconnect all external devices completely. Inside the box, do likewise. Remove extra RAM cards, and leave only the one. I am inclined to suspect the graphics card, but I gather that your mobo does not have on-board graphics, so am wondering if it is possible for you find an old PCI graphics card to try instead of the PCI-e card.

Do these steps one at a time, testing your system after each step.

Let us know what happens.
AustrAlien
Google is my friend. Make Google your friend too.

Posted Image

#8 geewin

geewin
  • Topic Starter

  • Members
  • 12 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Brasil
  • Local time:02:53 PM

Posted 24 August 2009 - 09:10 PM

hello AustrAlien,
thanks for the response.
I may have been a bit abrasive with Hamluis, but I found his reply less than helpful and I thought he was taking the mickey with the remark about testing batteries! The question re. booting with a `deadī (or no) CMOS battery was because I was told that it was possible by a local PC `technician`. I haven`t tried it myself.
One of my problems with trying to fix my PCs is that, thanks to the continually changing standards, the stuff in my older PCs are not compatible with the newer, so I donīt have other parts to test with, particularly other graphics cards; I only have AGP. I am purchasing another PSU (OCZ 500W StealthXStream) through īmercardo livreī (our e-bay) but wonīt get it for at least a week (it is costing approx. $US136!!! and thats the cheapest in Brasil). Letīs hope it proves to be the culprit and I havenīt spent the money for nothing.
If youīre interested in the result, watch this space.
Thanks
Geoff

#9 AustrAlien

AustrAlien

    Inquisitor


  • Members
  • 6,772 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Cowra NSW Australia
  • Local time:03:53 AM

Posted 25 August 2009 - 02:27 AM

Geoff

When a PC with a flat/dead CMOS battery is started, it presents the Bios Setup Menu to the user. Having lost battery power, the BIOS will have been "cleared", which means it will have been re-set to it's default original settings, including the date & time. At this stage the user can choose to set the correct date, or not, and then continue to load Windows.

This is the same as "clearing" CMOS by removing the battery and using the "clear CMOS" mobo jumper settings, often done to restore BIOS to default settings and/or to remove the BIOS password.

So, in answer to your question, yes, a computer can be started with a "dead" CMOS battery. I have not tried with no battery at all, but would guess it to be the same as with a "dead" battery.
------------------
You have tested the PSU by using a replacement, and I am relatively comfortable with that, and therefore think it unlikely that the problem lies with the PSU. (I know you did not want to hear that .... but I could be wrong?)

I have reviewed your posts/information supplied, and what I said in my previous post still seems to me the most likely procedure for you to resolve the issue. I would be first checking graphics card or possible ram failure or incompatibility (remove all but one card).

Geoffrey
AustrAlien
Google is my friend. Make Google your friend too.

Posted Image

#10 geewin

geewin
  • Topic Starter

  • Members
  • 12 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Brasil
  • Local time:02:53 PM

Posted 29 August 2009 - 01:29 PM

Hi Geoffrey,
well, you were right it isnīt the PSU!
While I was trying with different configurations e.g. after removing 1 RAM card etc. the machine started once, but again with the continuous beeping. I tried to restart with the intention of checking the BIOS info, but it has never started again with any combination.
What I donīt understand is what turns the machine off and on again after a couple of seconds, sometimes 2 or 3 times by itself, and why there is no POST or BIOS. Also, surely it is unusual for a Graphics card to fail completely in this way, particularly a quality card like the evga?
Iīm going to try to borrow a graphics card from a friend but itīs a pci-e GeF 8600GT, so thereīs still no way to test the pci-e slot.
So the saga continues,
Geoff

#11 AustrAlien

AustrAlien

    Inquisitor


  • Members
  • 6,772 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Cowra NSW Australia
  • Local time:03:53 AM

Posted 29 August 2009 - 02:54 PM

What I donīt understand is what turns the machine off and on again after a couple of seconds, sometimes 2 or 3 times by itself, and why there is no POST or BIOS. Also, surely it is unusual for a Graphics card to fail completely in this way


Based on past experience, these symptoms are a very strong (100% ?) indicator of graphics card failure/motherboard not happy with graphics card. If you were to plug in an old working PCI graphics card, I would be expecting the computer to work without a problem.

Considering that in my mind, the graphics card and the PSU are the top two trouble-making suspects, when first facing a computer with startup problems, my answer would be "No, I don't think it unusual at all."

Edit: btw ... the next on the trouble-making list is RAM ... continuous beeping = ram problem

Edited by AustrAlien, 29 August 2009 - 02:57 PM.

AustrAlien
Google is my friend. Make Google your friend too.

Posted Image

#12 geewin

geewin
  • Topic Starter

  • Members
  • 12 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Brasil
  • Local time:02:53 PM

Posted 05 September 2009 - 07:08 PM

Hi,
well I took the machine to a guy I know who repairs PCs and we changed the graphics card, the memory, and the CPU, and apart from starting once but failing to re-start, it continued to behave the same! Our conclusion is that the power controller on the M/B is at fault. What do you think?
Geoff

#13 AustrAlien

AustrAlien

    Inquisitor


  • Members
  • 6,772 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Cowra NSW Australia
  • Local time:03:53 AM

Posted 05 September 2009 - 07:14 PM

we changed the graphics card

Did you try a PCI graphics card or another PCI-e ?
AustrAlien
Google is my friend. Make Google your friend too.

Posted Image

#14 AustrAlien

AustrAlien

    Inquisitor


  • Members
  • 6,772 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Cowra NSW Australia
  • Local time:03:53 AM

Posted 06 September 2009 - 06:20 AM

Our conclusion is that the power controller on the M/B is at fault. What do you think?

I am not one to draw hasty conclusions, especially when it is not me doing the testing. I am reluctant to make any comment on this, except to say "I would want to do much more testing before outlaying any of my money possibly unnecessarily on another mobo".

For example:
Did you test your PCI-e graphics card in another working machine?
Did you test your RAM in another working machine?
If both the above items checked out as faultless, the next question I would be asking is ...
Did you try with a working PCI graphics card?
What happens when you remove the graphics card altogether, and attempt to start your system?
What happens when you remove the RAM (all of it) and attempt to start your system?
Are any of the characteristics you see then, similar to the symptoms you are seeing when trying to start your system normally?
--------------------
You haven't stated the origin of this computer, whether it was put together by you or otherwise .... but regardless of that .... and before wasting any more money on parts, I would suggest that you do the following (and here I am thinking of possible electrical shorting of the mobo with the box .... could be something as trivial as a loose screw somewhere it shouldn't be, or the mobo just too close to the box .... or something like that.)

Strip your system (everything) from the box and lay the mobo on a table (on a wooden board or cardboard box or some such non-conductive surface. Inspect the mobo carefully for any signs of damage such as cracking or scratching or bent contacts in the slots (PCI-e and RAM slots). Check that there is no sign of swelling or leaking of the capacitors. Connect up the PSU. Connect up the switches from the front of the box so that you can turn it on and off normally (easy to say, maybe not so easy for you to do, and you must be meticulous with recording where the various wires connect to the mobo). Fit your graphics card and ONE stick of RAM, and connect the monitor, mouse and keyboard. Try starting your system now. What happens? "Bench-testing" your system is the best way to go at this time, when things are getting "desperate". Try starting your system without keyboard attached, then without mouse attached.

The problem could indeed be with the mobo, but it is really pretty much "last" on my list of suspects, and really only diagnosable by ruling everything else out as a possibility (unless of course you have a "spare" mobo of the same or similar configuration into which you can connect all the other parts to test with). Again, all this must be done "out" of the box, to ensure there is no problem caused by fitting the system in the box itself. All this type of work is made much easier if you have an abundant supply of "bits and pieces". If you do not have this "luxury" then things are going to be more difficult for you and you may have to at some stage, outlay some money and hope for the best. If your bench-testing, or any of the other steps above, don't help to diagnose the problem, then perhaps you will have to "try a new mobo".

Also, test by replacing all connecting cables where possible.

Please let us know how things are progressing.
Best of luck.
AustrAlien
Google is my friend. Make Google your friend too.

Posted Image

#15 geewin

geewin
  • Topic Starter

  • Members
  • 12 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Brasil
  • Local time:02:53 PM

Posted 12 September 2009 - 07:34 AM

Hello again,
first, in answer to your questions, I (very carefully) assembled the PC myself; the graphics card was PCI-e, also I tested my card and RAM in another PC and it was fine.
Well I've done all the things you suggested and more, and basically the problem is the same (although I no longer get the continuous 'beep' when it does start up!). Now the machine seems to start more often, particularly when I've left it switched off for a while with the power lead unplugged, but it refuses to restart when its been running. I left it on all night running a test utility - 'Hot CPU Tester Pro 4' - which found no fault. The machine is still on and works perfectly even with demanding App.s, with no overheating / noise etc.
I've given a lot of thought to the symptoms, particularly the way that it switches on then off then tries again and I feel sure that the fault is with one (or more) of the 'relays' that switch the power to the main components. I haven't yet found detailed info about the system, but it seems as though the component (maybe like a Thyristor?) fails to 'latch on'. I suspect it's the small black cubes (eg 1R2, R50, 2R0) scattered around the MoBo.
What do you know about how the MoBo power switching system works? Is this a possibility do you think?
Regards,
Geoff




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users