Hello,

My computer will not power on. It is a GA-MA770 DS3 rev 1 board. I
have disconnected everything (ethernet, keyboard, mouse, etc) and all
drives but it still will not power up. The cpu fan does not turn.

Do I assume the board is dead? Is there anything I can do? If not,
what replacement board do you recommend?

Thanks.

Stephen wrote:
> Hello,
>
> My computer will not power on. It is a GA-MA770 DS3 rev 1 board. I
> have disconnected everything (ethernet, keyboard, mouse, etc) and all
> drives but it still will not power up. The cpu fan does not turn.
>
> Do I assume the board is dead? Is there anything I can do? If not,
> what replacement board do you recommend?
>
> Thanks.

Debugging this, is not as easy as it seems. Either the motherboard
or the PSU could be at fault. And changing out the PSU might not
fix it - for example, if the motherboard is shorting out the +5VSB
rail on the supply, it'll give the impression that the supply
is dead, when in fact the motherboard is causing the problem.

To test the supply, you can disconnect it from the system, then
connect PS_ON# (green) to COM (black). PS_ON# has to be a low
level for the ATX supply to start. For details on pinout, grab
the appropriate ATX spec. It helps of the supply has a dummy
load on it (some supplies have a minimum loading spec).

(20 pin standard, back when the supply offered -5V as well. You
likely don't have one of these.)
http://web.archive.org/web/200304240...12V_PS_1_1.pdf

(20 pin modern)
http://www.formfactors.org/developer...X12V_1_3dg.pdf

(24 pin modern)
http://www.formfactors.org/developer...public_br2.pdf

You can even short PS_ON# to COM, while the main connector is
connected to the motherboard. But then, that defeats motherboard
safety features, such as overheat protection via THERMTRIP,
and is not recommended as a permanent fix.

The basic sequence goes like this.

(PSU end) (motherboard end)

1) Plug in AC
Switch on at back PSU delivers +5VSB Motherboard logic receives power.
Asus power LED lights up. Or, look
for any other LEDs powered by +5VSB.
Not all motherboard brands, have a
power LED to check.

2) Press front
Power Motherboard latches pulse.
<--- Makes a steady level on PS_ON#
Sends it to the PSU.

Supply delivers
3.3, 5, 12V --->

PSU fan is spinning.

Supply sends
Power_Good ---> Motherboard releases Reset and
POST starts.

If the power supply refuses to recognize a valid level on PS_ON#, it
can refuse to switch on. In that case, the motherboard is not at fault.
And if the power supply never delivered +5VSB in the first place,
that can also cause the whole process to come to a halt. The motherboard
logic can't work, without initial power being delivered. +5VSB is
delivered, without the PSU fan spinning, so it is "silent delivery".

Paul

On Mon, 20 Feb 2012 14:35:02 -0500, Paul <> wrote:

>Debugging this, is not as easy as it seems. Either the motherboard
>or the PSU could be at fault. And changing out the PSU might not
>fix it - for example, if the motherboard is shorting out the +5VSB
>rail on the supply, it'll give the impression that the supply
>is dead, when in fact the motherboard is causing the problem.
>
>To test the supply, you can disconnect it from the system, then
>connect PS_ON# (green) to COM (black). PS_ON# has to be a low
>level for the ATX supply to start. For details on pinout, grab
>the appropriate ATX spec. It helps of the supply has a dummy
>load on it (some supplies have a minimum loading spec).
>
>(20 pin standard, back when the supply offered -5V as well. You
> likely don't have one of these.)
>http://web.archive.org/web/200304240...12V_PS_1_1.pdf
>
>(20 pin modern)
>http://www.formfactors.org/developer...X12V_1_3dg.pdf
>
>(24 pin modern)
>http://www.formfactors.org/developer...public_br2.pdf


Thank you for your help and thanks for those links. I had the last
one, the one with 12 pins either side.

I was confused before you replied because I could not understand why
the psu fan was not whirring. Thanks to your reply I now know.

I shorted com and ps_on as you suggested and the psu fan came to life.
I also tested the voltages on a molex at the same time and I got 5 and
12v as there should be.

So I think we can say this is a motherboard rather than a psu fault.

I unscrewed the motherboard and looked underneath but there was no
sign of damage. I refitted the board and with just the cpu (and
heatsink!) fitted tried to power it on. The cpu fan came on!

I switched it off and fitted the ram. I switched it on again and the
cpu fan started again but then it stopped. There was a warm smell
coming from the far side of the motherboard.

I think the capacitors UEC2 and UEC3 next to the clear cmos jumper
seemed warm; should they be?

But worse still, EC40 another elctrolytic capacitor was too hot to
touch. I'm wondering if the board came on for a few seconds and then
shut down when this overheated?

there does not seem to be anything burnt out, no visible damage, but
it does smell hot and not right.

Do you know what these capacitors do or does gigabyte not release this
level of detail publicly?

I notice these capacitors are near connections for front usb sockets.
These were not connected but I did connect a usb cable to the rear usb
sockets and fitted the B-usb plug upside down in my laser printer a
couple of days before the motherboard failed.

I don't know how I managed to do this. I was rushing but didn't use
excessive force but managed to get the plug in upside down. I realise
this would have connected the power to the data pins and vice versa.
I'm wondering whether these capacitors are involved with the usb,
being near that area?

Could this have caused them to fail? But if so, it seems strange that
it was a few days later before they failed. I would have expected them
to fail straight away. Also it seems poor design to allow a usb
failure to destroy a whole board.

I guess I will have to just replace the board but as I said in another
post, I am having great difficulty finding another am2 board, so I may
have to upgrade the processor at the same time.

Thanks again,
Stephen

Stephen wrote:
> On Mon, 20 Feb 2012 14:35:02 -0500, Paul <> wrote:
>
>> Debugging this, is not as easy as it seems. Either the motherboard
>> or the PSU could be at fault. And changing out the PSU might not
>> fix it - for example, if the motherboard is shorting out the +5VSB
>> rail on the supply, it'll give the impression that the supply
>> is dead, when in fact the motherboard is causing the problem.
>>
>> To test the supply, you can disconnect it from the system, then
>> connect PS_ON# (green) to COM (black). PS_ON# has to be a low
>> level for the ATX supply to start. For details on pinout, grab
>> the appropriate ATX spec. It helps of the supply has a dummy
>> load on it (some supplies have a minimum loading spec).
>>
>> (20 pin standard, back when the supply offered -5V as well. You
>> likely don't have one of these.)
>> http://web.archive.org/web/200304240...12V_PS_1_1.pdf
>>
>> (20 pin modern)
>> http://www.formfactors.org/developer...X12V_1_3dg.pdf
>>
>> (24 pin modern)
>> http://www.formfactors.org/developer...public_br2.pdf
>
>
> Thank you for your help and thanks for those links. I had the last
> one, the one with 12 pins either side.
>
> I was confused before you replied because I could not understand why
> the psu fan was not whirring. Thanks to your reply I now know.
>
> I shorted com and ps_on as you suggested and the psu fan came to life.
> I also tested the voltages on a molex at the same time and I got 5 and
> 12v as there should be.
>
> So I think we can say this is a motherboard rather than a psu fault.
>
> I unscrewed the motherboard and looked underneath but there was no
> sign of damage. I refitted the board and with just the cpu (and
> heatsink!) fitted tried to power it on. The cpu fan came on!
>
> I switched it off and fitted the ram. I switched it on again and the
> cpu fan started again but then it stopped. There was a warm smell
> coming from the far side of the motherboard.
>
> I think the capacitors UEC2 and UEC3 next to the clear cmos jumper
> seemed warm; should they be?
>
> But worse still, EC40 another elctrolytic capacitor was too hot to
> touch. I'm wondering if the board came on for a few seconds and then
> shut down when this overheated?
>
> there does not seem to be anything burnt out, no visible damage, but
> it does smell hot and not right.
>
> Do you know what these capacitors do or does gigabyte not release this
> level of detail publicly?
>
> I notice these capacitors are near connections for front usb sockets.
> These were not connected but I did connect a usb cable to the rear usb
> sockets and fitted the B-usb plug upside down in my laser printer a
> couple of days before the motherboard failed.
>
> I don't know how I managed to do this. I was rushing but didn't use
> excessive force but managed to get the plug in upside down. I realise
> this would have connected the power to the data pins and vice versa.
> I'm wondering whether these capacitors are involved with the usb,
> being near that area?
>
> Could this have caused them to fail? But if so, it seems strange that
> it was a few days later before they failed. I would have expected them
> to fail straight away. Also it seems poor design to allow a usb
> failure to destroy a whole board.
>
> I guess I will have to just replace the board but as I said in another
> post, I am having great difficulty finding another am2 board, so I may
> have to upgrade the processor at the same time.
>
> Thanks again,
> Stephen

Electrolytic capacitors should not be hot to the touch.
If an electrolytic were to have reverse polarity applied to it,
it could explode. That used to be a joke in electronics lab,
to reverse an electrolytic in someone's breadboard when they
weren't looking, such that it would pop when they turned the
power on. The air around the area, is filled with filaments
of what looks like carbon deposits (like black confetti). I've
only been present for one of these "jokes", which I don't
consider to be that funny.

Now, the other kind of capacitor, is the polymer kind. The difference
is, the electrolytic has "pressure relief seams" stamped in the top
for safety. The polymer cap doesn't have pressure relief on top,
just a smooth finish. I have no idea how polymer caps behave to
polarity reversal or other insults. I've never tried to make
any blow up :-)

USB ports can be protected, but in a "cheap PC environment", there
isn't much incentive to do "belt and suspenders" protection. On
USB ports, you can add extra protection for ESD (static discharge).
I don't know if you could do good protection against
connecting power or not.

If you were using a USB2 PCI card, and reversed a port,
just the PCI card would be ruined, and then you could
replace it. Which would be another form of protection - a
$10 USB card would be cheaper to replace, than a whole
motherboard.

Whatever has happened here, it sounds like a job for
motherboard replacement. If the Southbridge was blown by
this reversal, that would be expensive to repair at any
place except the factory.

Paul

On Sat, 25 Feb 2012 14:31:47 -0500, Paul <> wrote:

>Electrolytic capacitors should not be hot to the touch.
>If an electrolytic were to have reverse polarity applied to it,
>it could explode. That used to be a joke in electronics lab,
>to reverse an electrolytic in someone's breadboard when they
>weren't looking, such that it would pop when they turned the
>power on. The air around the area, is filled with filaments
>of what looks like carbon deposits (like black confetti). I've
>only been present for one of these "jokes", which I don't
>consider to be that funny.
>
>Now, the other kind of capacitor, is the polymer kind. The difference
>is, the electrolytic has "pressure relief seams" stamped in the top
>for safety. The polymer cap doesn't have pressure relief on top,
>just a smooth finish. I have no idea how polymer caps behave to
>polarity reversal or other insults. I've never tried to make
>any blow up :-)

Hello again,

Sorry for the delay in replying to your posts. I have now bought the
GA-M68M-S2P and as this proves, it is up and running.

It is interesting what you say about capacitors. The one that was hot
had, IIRC, EC40 printed on the board next to it, so I assumed that
stood for electrolytic capacitor number 40. But now I have been
thinking about this and didn't Gigabyte do a marketing campaign where
they said they did not use electrolytic capacitors?

The capacitors on the faulty board are radial metal cylinders but the
tops don't look quite like electrolytics: they look more like metal
transistors if you can imagine what I mean.

However my replacement board has electrolytic capacitors and like you
say, these have stamps on the top (though I never knew before now what
these were for - thanks), so perhaps their budget boards still use
electrolytics or perhaps that marketing campaign has been and gone?

Thanks,
Stephen.

Stephen wrote:
> On Sat, 25 Feb 2012 14:31:47 -0500, Paul <> wrote:
>
>> Electrolytic capacitors should not be hot to the touch.
>> If an electrolytic were to have reverse polarity applied to it,
>> it could explode. That used to be a joke in electronics lab,
>> to reverse an electrolytic in someone's breadboard when they
>> weren't looking, such that it would pop when they turned the
>> power on. The air around the area, is filled with filaments
>> of what looks like carbon deposits (like black confetti). I've
>> only been present for one of these "jokes", which I don't
>> consider to be that funny.
>>
>> Now, the other kind of capacitor, is the polymer kind. The difference
>> is, the electrolytic has "pressure relief seams" stamped in the top
>> for safety. The polymer cap doesn't have pressure relief on top,
>> just a smooth finish. I have no idea how polymer caps behave to
>> polarity reversal or other insults. I've never tried to make
>> any blow up :-)
>
> Hello again,
>
> Sorry for the delay in replying to your posts. I have now bought the
> GA-M68M-S2P and as this proves, it is up and running.
>
> It is interesting what you say about capacitors. The one that was hot
> had, IIRC, EC40 printed on the board next to it, so I assumed that
> stood for electrolytic capacitor number 40. But now I have been
> thinking about this and didn't Gigabyte do a marketing campaign where
> they said they did not use electrolytic capacitors?
>
> The capacitors on the faulty board are radial metal cylinders but the
> tops don't look quite like electrolytics: they look more like metal
> transistors if you can imagine what I mean.
>
> However my replacement board has electrolytic capacitors and like you
> say, these have stamps on the top (though I never knew before now what
> these were for - thanks), so perhaps their budget boards still use
> electrolytics or perhaps that marketing campaign has been and gone?
>
> Thanks,
> Stephen.

There is nothing wrong with electrolytic caps.

The "capacitor plague" was caused by a stolen formula for the electrolyte.
They forgot to stabilize the electrolyte, and a lot of bad capacitors were
made as a result. This caused premature failures, and even failures while
the caps were not being used. I had an Antec power supply, with very few
service hours on it, and two years after purchase, opening the lid showed
leaking capacitors. So the "chemistry error" in the construction of the
caps, can cause their demise, even without bias applied.

http://en.wikipedia.org/wiki/Capacitor_plague

To distance themselves from the capacitor plague, the major motherboard
makers switched to polymer. It was for marketing purposes. It was
a way of saying "see, I didn't buy any electrolytic capacitors of
unknown quality when I made your motherboard".

I have an 11 year old motherboard, and the electrolytics in that
are still good.

One capacitor manufacturer states, that without abusive temperatures
applied, the lifespan can be 15 years or so. Past that point, the
rubber bung in the bottom of the electrolytic, tend to dry and crack,
and then the electrolyte inside dries out. There is a big difference,
between a cap lasting 15 years, and one leaking and oozing after only
two years (stored in a closet).

Gigabyte uses different colors of polymer caps on their motherboards,
but I'm not aware of the significance. I've seen all manner of
construction styles, pure electrolytic, mixed electrolytic/polymer
(polymer where it makes sense, like VCore), as well as pure polymer.
So all the styles have been tried. It's true, that you'd expect
a better lifespan from a pure polymer board, but you can also
get just as much fun from a properly designed electrolytic board.

Paul