Discussions

Can't Restart for an Hour After ShutDown

+
1 Votes
Locked

Can't Restart for an Hour After ShutDown

brian.catt
I am kinda technical but have a life. My problem is a my son's ex gaming dual core many GHz RAM stuffed PC (which has a serious graphics card from NVIDIA) he made himself with an ASUS motherboard. He doesn't have a life. It used to work OK. My wife now uses it. I just fixed a long running boot.ini invalid problem (rebuilt the file) hoping the below problem would finally be solved as well. Its still there. PROBLEM

It works OK - once its started. If its shut down its often impossible to restart it for about 1 hour, and it has to be switched off at the power switch after shut down for that to work, not just left quiescent. If its left the hour plus switched off and then switched back on at the power switch it will boot first time every time.

It will also will restart when running no problem.

If an attempt is made before the hour or so , when you switch mains power on the little lights in the KB flash and the Logitech WIreless Mouse driver green light illuminates, but nothing happens when you push the start button. No lights inside, no whirring things. No boot. This failed start also seems to reset the 1 hour time delay in failing to start the PC.

Flakier behaviour: If you catch it fast on the button as the fan dies after shutdown it may restart (when their is residual charge and some back emf in the heavier drain components?). Also If one is very lucky by constant pushing of the start button while switching the power on and off it MAY fire up after a recent shut down. But not when you really need it to.

Sounds like a PSU problem but maybe something going on with too much drain? I would like to know what controls the PSU actually booting itself when the button is pushed? I really need someone who has had this problem to solve.

Any ideas? I can answer the inevitable questions, BTW can I get config info from a software tool in XP somewhere? Can rip it apart but it ain't broke/as in unusable.

I'm more of a Mac user so not used to such problems and the idea of a hobbyist building a high technology device from a bag of vaguely compatible bits with the odd fatal incompatibility sounds like an unnatural act with sub systems to me :-)

Thanks for any inspired or even vaguely relevant ideas.

Brian +44 1932 772731 Skype Catthouse
  • +
    1 Votes
    Slayer_

    And once its hot it won't turn back on. Its probably throwing funky voltages at the mother board and the board is preventing it from starting.

    +
    0 Votes
    brian.catt

    Thanks. Do you have experience of measuring out of limits voltages from PSU to Motherboard in these circumstances? I have none and don't know the failure modes at all, which is why I'm here. However as an electrical design engineer for many years I don't see why a PSU should get "weak"or spontaneously lose its regulation. Especially as it works fine when it has started.

    Failure modes in such a circuit as I recall them were generally either work or fail, as voltage regulation is a very strongly fed back control mechanism. But it could be a load related problem, yes.

    What controls does the Motherboard have to switch off the PSU?

    Thanks,

    +
    0 Votes
    NickNielsen Moderator

    If you have a multimeter, you can test the PSU voltages at the mainboard connector. Use the chassis as ground.

    The following two tests will usually isolate the problem to either the mainboard or power supply:

    1. In standby mode, measure on the purple or violet wire. You should see +5VDC±.05.

    2. Place the test lead on the Power OK line (grey wire) and press the power switch. This line should go to +5 VDC. If this signal is missing, the processor will not boot.

    Other wires and voltages are <a href=http://pinouts.ru/Power/atxpower_pinout.shtml>listed here</a>.

    +
    0 Votes
    brian.catt

    Thanks. Do you have experience of measuring out of limits voltages from PSU to Motherboard in these circumstances? I have none and don't know the failure modes at all, which is why I'm here. However as an electrical design engineer for many years I don't see why a PSU should get "weak"or spontaneously lose its regulation. Especially as it works fine when it has started.

    Failure modes in such a circuit as I recall them were generally either work or fail, as voltage regulation is a very strongly fed back control mechanism. But it could be a load related problem, yes.

    Just in case it was marginal when hot I tried booting from cold, then immediately shutting down before it got to operating temperature and rebooting after 20 sec or so. No difference.

    What controls does the Motherboard have to switch off the PSU?

    Thanks,

    +
    0 Votes
    HAL 9000 Moderator

    Turn off unplug the Power Lead and press the On Button for 10 Seconds.

    After the 10 Seconds is up plug the Power Lead back in and try to turn on.

    If it works now the Capacitors in the PS are faulty.

    By unplugging it and pressing the Power Switch you are draining the Capacitors in the Power Supply and on the M'Board so when you plug it back in it is like a fresh start from a system not run for a logn time.

    With ATX PS's they always supply a 5 V DC Voltage and the On Switch on the front causes a Ripple on the Rail which turns on the remainder of the Power Supply and it runs till either it receives a Signal from the M'Board to tell it to keep running or it turns off if it doesn't receive that signal.

    I would have to know which ASUS M'Board is involved here but most have a LED on them so is this Lit when the unit is powered not necessarily running?

    I would first be looking at replacing the PS with a Good Brand and not a No Name which has very low Specifications and a Peek Ratting. These do not filter the Mains properly and allow Spikes through inside the case degrading the Computers Internal Electronics.

    I would look at the Antec PS here

    http://www.antec.com/Believe_it/product.php?Family=MzM5

    As a starting point. Antec PS's are rated to the Power on the Sticker 100% of the time and not to peek values like so many of the No Name units. They have much better Filtering built into them and protect the Internals of the case from all of the muck on the Mains.

    Of course the down side of this is that they stop working more often when they hit something nasty and give the impression of not being as reliable when actually they have sacrificed themselves to protect the Computer proper.

    You should also check against the Antec Power Supply Calculator for the minimum Size PS to use here

    http://www.antec.outervision.com/

    Col

    +
    0 Votes
    NickNielsen Moderator

    Your instinct may be correct and the PSU is getting an overload from somewhere; the power reset and time-out are normally associated with a PSU overload/overtemp fault. You may want to replace the existing unit with one that has 100VA greater load capacity.

    If you are having problems with lock-ups or random shutdowns, check your mainboard for <a href=http://en.wikipedia.org/wiki/Capacitor_plague>swollen capacitors</a>.

  • +
    1 Votes
    Slayer_

    And once its hot it won't turn back on. Its probably throwing funky voltages at the mother board and the board is preventing it from starting.

    +
    0 Votes
    brian.catt

    Thanks. Do you have experience of measuring out of limits voltages from PSU to Motherboard in these circumstances? I have none and don't know the failure modes at all, which is why I'm here. However as an electrical design engineer for many years I don't see why a PSU should get "weak"or spontaneously lose its regulation. Especially as it works fine when it has started.

    Failure modes in such a circuit as I recall them were generally either work or fail, as voltage regulation is a very strongly fed back control mechanism. But it could be a load related problem, yes.

    What controls does the Motherboard have to switch off the PSU?

    Thanks,

    +
    0 Votes
    NickNielsen Moderator

    If you have a multimeter, you can test the PSU voltages at the mainboard connector. Use the chassis as ground.

    The following two tests will usually isolate the problem to either the mainboard or power supply:

    1. In standby mode, measure on the purple or violet wire. You should see +5VDC±.05.

    2. Place the test lead on the Power OK line (grey wire) and press the power switch. This line should go to +5 VDC. If this signal is missing, the processor will not boot.

    Other wires and voltages are <a href=http://pinouts.ru/Power/atxpower_pinout.shtml>listed here</a>.

    +
    0 Votes
    brian.catt

    Thanks. Do you have experience of measuring out of limits voltages from PSU to Motherboard in these circumstances? I have none and don't know the failure modes at all, which is why I'm here. However as an electrical design engineer for many years I don't see why a PSU should get "weak"or spontaneously lose its regulation. Especially as it works fine when it has started.

    Failure modes in such a circuit as I recall them were generally either work or fail, as voltage regulation is a very strongly fed back control mechanism. But it could be a load related problem, yes.

    Just in case it was marginal when hot I tried booting from cold, then immediately shutting down before it got to operating temperature and rebooting after 20 sec or so. No difference.

    What controls does the Motherboard have to switch off the PSU?

    Thanks,

    +
    0 Votes
    HAL 9000 Moderator

    Turn off unplug the Power Lead and press the On Button for 10 Seconds.

    After the 10 Seconds is up plug the Power Lead back in and try to turn on.

    If it works now the Capacitors in the PS are faulty.

    By unplugging it and pressing the Power Switch you are draining the Capacitors in the Power Supply and on the M'Board so when you plug it back in it is like a fresh start from a system not run for a logn time.

    With ATX PS's they always supply a 5 V DC Voltage and the On Switch on the front causes a Ripple on the Rail which turns on the remainder of the Power Supply and it runs till either it receives a Signal from the M'Board to tell it to keep running or it turns off if it doesn't receive that signal.

    I would have to know which ASUS M'Board is involved here but most have a LED on them so is this Lit when the unit is powered not necessarily running?

    I would first be looking at replacing the PS with a Good Brand and not a No Name which has very low Specifications and a Peek Ratting. These do not filter the Mains properly and allow Spikes through inside the case degrading the Computers Internal Electronics.

    I would look at the Antec PS here

    http://www.antec.com/Believe_it/product.php?Family=MzM5

    As a starting point. Antec PS's are rated to the Power on the Sticker 100% of the time and not to peek values like so many of the No Name units. They have much better Filtering built into them and protect the Internals of the case from all of the muck on the Mains.

    Of course the down side of this is that they stop working more often when they hit something nasty and give the impression of not being as reliable when actually they have sacrificed themselves to protect the Computer proper.

    You should also check against the Antec Power Supply Calculator for the minimum Size PS to use here

    http://www.antec.outervision.com/

    Col

    +
    0 Votes
    NickNielsen Moderator

    Your instinct may be correct and the PSU is getting an overload from somewhere; the power reset and time-out are normally associated with a PSU overload/overtemp fault. You may want to replace the existing unit with one that has 100VA greater load capacity.

    If you are having problems with lock-ups or random shutdowns, check your mainboard for <a href=http://en.wikipedia.org/wiki/Capacitor_plague>swollen capacitors</a>.