A8V Deluxe keeps rebooting and shutting down

[hitch]

Brand new system setup:

A8V Deluxe
Athlon 64 3400+
PSU Codegen 400W
XFX Geforce 6600GT
HD 80G SATA

Problem:

Keeps shutting down all the time.

History:

Worked fine for the whole day. Then I went into BIOS to check the RAM
speed. Didn’t find the setting and didn’t change anything, but after
that the message "Overclocking failed!" appeared at BIOS startup,
and the system started shutting down. I left it in standby overnight,
and woke up to find it off. It wouldn’t react to the front panel power
button, only to the power cord getting uplugged and plugged back again
- then it would start for like 3 secs, clink, and turn off again.

After about 12 hours of being left alone it did start almost with no
problem - almost, because it blanked out a lamp in the room during the
bootup (a floor lamp with a relay in it). Worked for half an hour,
then Windows shut itself down. Rebooted itself a couple of times,
worked for another 15 minutes, shut down again.

I stripped it down to mobo, CPU, PSU, and RAM (even removed videocard
from its slot) - it worked for an hour and a half as a BIOS machine
and then shut down.

I went to BIOS to AI Overclocking, changed it from Auto to Standard,
then it shut down half a minute later.

I flashed the BIOS with the latest version only to get a "Bad CMOS
checkup" and a shutdown in five minutes.

What could it possibly be and how to fix it?

[HardwareForumz editor: this issue was also posted in response to a
thread "Voltage and temp issues", but without the details provided
here]

 

[Rob Hemmings]

Brand new system setup:

A8V Deluxe
Athlon 64 3400+
PSU Codegen 400W
XFX Geforce 6600GT
HD 80G SATA

Problem:

Keeps shutting down all the time.

History:

Worked fine for the whole day. Then I went into BIOS to check the RAM
speed. Didn't find the setting and didn't change anything, but after
that the message "Overclocking failed!" appeared at BIOS startup,
and the system started shutting down. I left it in standby overnight,
and woke up to find it off. It wouldn't react to the front panel power
button, only to the power cord getting uplugged and plugged back again
- then it would start for like 3 secs, clink, and turn off again.

After about 12 hours of being left alone it did start almost with no
problem - almost, because it blanked out a lamp in the room during the
bootup (a floor lamp with a relay in it). Worked for half an hour,
then Windows shut itself down. Rebooted itself a couple of times,
worked for another 15 minutes, shut down again.

I stripped it down to mobo, CPU, PSU, and RAM (even removed videocard
from its slot) - it worked for an hour and a half as a BIOS machine
and then shut down.

I went to BIOS to AI Overclocking, changed it from Auto to Standard,
then it shut down half a minute later.

I flashed the BIOS with the latest version only to get a "Bad CMOS
checkup" and a shutdown in five minutes.

What could it possibly be and how to fix it?

Sounds like the PSU might not up to the job to me. Any chance
you can test it with another?
HTH

 

[Mike]

I'd look at the power supply

 

[Peter van der Goes]

Brand new system setup:

A8V Deluxe
Athlon 64 3400+
PSU Codegen 400W
XFX Geforce 6600GT
HD 80G SATA

Problem:

Keeps shutting down all the time.

<snip>
Hmmm...
I seem to recall this motherboard/CPU combination as one offered for sale by
Tiger Direct which generated a lot of hate, discontent and returns. The
issue was that specific model of OEM CPU that was sold by many vendors as
available only when purchased with a few motherboard/BIOS revision
combinations known not to cause problems. NewEgg, in particular restricted
sales in this manner. Tiger Direct ran a "special" on the A8V - 3400+ 939
pin combination knowing (or, they *should* have known) the combination was
problematical at best. IIRC, to make it work correctly, you have to flash to
a later BIOS revision than that supplied with the A8V at the time, but of
course, you cannot do that unless you can get it running stably.

If you want to know more about the A8V - 3400+ saga, search in Google Groups
for "A8V and 3400+ CPU" and have a good long read.

You'll note that if you go to stores like NewEgg, MultiWave, etc. you won't
find a 939 pin 3400+ for sale at all. AAMF, if you try PriceWatch, you won't
find it there either.

You can try removing all but one RAM module to get it running long enough
the update the BIOS. First, clear your CMOS to go back to default settings.

 

[hitch]

Sounds like the PSU might not up to the job to me. Any chance
you can test it with another?
HTH

I had it run with another PSU over this past night. First it ran fine,
but by the morning is started rebooting again.

It was a 250W power supply, the one I just tested it with, so maybe
because it’s insufficiently powerful? However, I worked on it for like
an hour after connecting it, and it was fine, so I guess it shouldn’t
have acted up when no one was touching it...

 

[Axl Myk]

CLUE!

almost, because it blanked out a lamp in the room during the
bootup (a floor lamp with a relay in it).

If that doesn't hit you in the face, then you had better take the system
to someone that knows what they're doing..

 

[Oldish sod]

I had it run with another PSU over this past night. First it ran fine,
but by the morning is started rebooting again.

It was a 250W power supply, the one I just tested it with, so maybe
because it's insufficiently powerful? However, I worked on it for like
an hour after connecting it, and it was fine, so I guess it shouldn't
have acted up when no one was touching it...

A "good" 350W PSU would probably be more than fine for your
system, but even the "best" 250W would struggle a bit..
HTH

 

[Paul]

CLUE!

almost, because it blanked out a lamp in the room during the
bootup (a floor lamp with a relay in it).

If that doesn't hit you in the face, then you had better take the system
to someone that knows what they're doing..

The reason that happened, is he had the floor lamp and the
computer on the same circuit. When you flip the switch on
the back of a standard ATX supply, a huge slug of current
goes flying into the power supply (for a short time). This
is the charging of the main caps in the power supply, to
370V or so. This will cause your lights to blink (mine do).
The relay in the floor lamp will drop out, when the AC is
removed for only a short time. If the floor lamp is put on
another circuit, maybe there will be enough separation
between the computer and lamp, for the lamp to be able to
stay on. (In North America, there are two phases, and all you
need to do, is find a plug in the room that is running off
the other phase.)

Based on the symptoms that the computer worked better, with
an underpowered (250W) supply, chances are, a more capable
supply is the solution to keeping the computer running.

Paul

 

[Bucky]

I had it run with another PSU over this past night. First it ran fine,
but by the morning is started rebooting again.

It was a 250W power supply, the one I just tested it with, so maybe
because it’s insufficiently powerful? However, I worked on it for like
an hour after connecting it, and it was fine, so I guess it shouldn’t
have acted up when no one was touching it...

Uh, I don't think that you can use a 250watt power supply, that should
be common sense. You want at least 400watts of a GOOD quality power
supply that has quality ripple current control and over current
protect worth a flip and so on. Get an Enermax Power supply. No
excuse spending a lot of money on a computer and nothing on a power
supply.

You will be lucky if you didn't screw up the motherboard and
components now. You should work with someone that knows what they are
doing a little bit more so that you don't have any more accidents, no
offense intended.

 

[hitch]

I had it run with another PSU over this past night. First it
ran fine, but by the morning is started rebooting again.

It was a 250W power supply, the one I just tested it with, so
maybe because it's insufficiently powerful? However, I worked
on it for like an hour after connecting it, and it was fine,
so I guess it shouldn't have acted up when no one was touching
it...

Thanks Peter van der Goes, very interesting, I didn’t know this issue
existed - now I’m reading a bunch of topics on that...

 

[hitch]

&nbsp;> > "hitch" &lt;UseLinkToEmail@HardwareForumz&gt; wrote
in message
&nbsp;> > &nbsp;&nbsp;> > > Brand new system setup:
&nbsp;&nbsp;> > >
&nbsp;&nbsp;> > > A8V Deluxe
&nbsp;&nbsp;> > > Athlon 64 3400+
&nbsp;&nbsp;> > > PSU Codegen 400W
&nbsp;&nbsp;> > > XFX Geforce 6600GT
&nbsp;&nbsp;> > > HD 80G SATA
&nbsp;&nbsp;> > >
&nbsp;&nbsp;> > > Problem:
&nbsp;&nbsp;> > >
&nbsp;&nbsp;> > > Keeps shutting down all the time.
&nbsp;&nbsp;> > >
&nbsp;&nbsp;> > > History:
&nbsp;&nbsp;> > >
&nbsp;&nbsp;> > > Worked fine for the whole day. Then I went
into BIOS to
&nbsp;> > check the RAM
&nbsp;&nbsp;> > > speed. Didn't find the setting and didn't
change anything,
&nbsp;> > but after
&nbsp;&nbsp;> > > that the message "Overclocking failed!"
appeared at BIOS
&nbsp;> > startup,
&nbsp;&nbsp;> > > and the system started shutting down. I left
it in standby
&nbsp;> > overnight,
&nbsp;&nbsp;> > > and woke up to find it off. It wouldn't
react to the front
&nbsp;> > panel power
&nbsp;&nbsp;> > > button, only to the power cord getting
uplugged and plugged
&nbsp;> > back again
&nbsp;&nbsp;> > > - then it would start for like 3 secs,
clink, and turn off
&nbsp;> > again.
&nbsp;&nbsp;> > >
&nbsp;&nbsp;> > > After about 12 hours of being left alone it
did start almost
&nbsp;> > with no
&nbsp;&nbsp;> > > problem - almost, because it blanked out a
lamp in the room
&nbsp;> > during the
&nbsp;&nbsp;> > > bootup (a floor lamp with a relay in it).
Worked for half an
&nbsp;> > hour,
&nbsp;&nbsp;> > > then Windows shut itself down. Rebooted
itself a couple of
&nbsp;> > times,
&nbsp;&nbsp;> > > worked for another 15 minutes, shut down
again.
&nbsp;&nbsp;> > >
&nbsp;&nbsp;> > > I stripped it down to mobo, CPU, PSU, and
RAM (even removed
&nbsp;> > videocard
&nbsp;&nbsp;> > > from its slot) - it worked for an hour and a
half as a BIOS
&nbsp;> > machine
&nbsp;&nbsp;> > > and then shut down.
&nbsp;&nbsp;> > >
&nbsp;&nbsp;> > > I went to BIOS to AI Overclocking, changed
it from Auto to
&nbsp;> > Standard,
&nbsp;&nbsp;> > > then it shut down half a minute later.
&nbsp;&nbsp;> > >
&nbsp;&nbsp;> > > I flashed the BIOS with the latest version
only to get a
&nbsp;> > "Bad CMOS
&nbsp;&nbsp;> > > checkup" and a shutdown in five minutes.
&nbsp;&nbsp;> > >
&nbsp;&nbsp;> > > What could it possibly be and how to fix it?
&nbsp;> >
&nbsp;> > Sounds like the PSU might not up to the job to me.
Any chance
&nbsp;> > you can test it with another?
&nbsp;> > HTH
&nbsp;> > --
&nbsp;> > Rob


Uh, I don't think that you can use a 250watt power supply,
that should
be common sense. You want at least 400watts of a GOOD quality
power
supply that has quality ripple current control and over
current
protect worth a flip and so on. Get an Enermax Power supply.
No
excuse spending a lot of money on a computer and nothing on a
power
supply.

You will be lucky if you didn't screw up the motherboard and
components now. You should work with someone that knows what
they are
doing a little bit more so that you don't have any more
accidents, no
offense intended.

Thanks everyone.

I bought a 430W Antec and it’s been running without incidents since
then, knock on the wood.

I know what I was doing re: 250W PSU I disconnected almost
everything, leaving only the basic components connected.

I thought Codegen was a good brand - turned out problematic.

 

[Peter van der Goes]

:

Thanks Peter van der Goes, very interesting, I didn't know this issue
existed - now I'm reading a bunch of topics on that...

I'm glad for you that you seem to have solved your problem with the new PSU.
You must have a new enough BIOS not to have the compatibility problem.
Enjoy!

 

[Bucky]

The reason that happened, is he had the floor lamp and the
computer on the same circuit. When you flip the switch on
the back of a standard ATX supply, a huge slug of current
goes flying into the power supply (for a short time). This
is the charging of the main caps in the power supply, to
370V or so. This will cause your lights to blink (mine do).
The relay in the floor lamp will drop out, when the AC is
removed for only a short time. If the floor lamp is put on
another circuit, maybe there will be enough separation
between the computer and lamp, for the lamp to be able to
stay on. (In North America, there are two phases, and all you
need to do, is find a plug in the room that is running off
the other phase.)

Based on the symptoms that the computer worked better, with
an underpowered (250W) supply, chances are, a more capable
supply is the solution to keeping the computer running.

Paul

Actually, it isn't 2 phases, it is 2 different legs phased the same.
They are two different legs off of the step down transformer up on the
pole or down on the ground some distance from the house.

 

[formerprof]

In my ignorance I always assumed that they had to be two-phased so that 240
volts was across them. I'd appreciate an explanation if I have it wrong.

formerprof

 

[Paul]

Actually, it isn't 2 phases, it is 2 different legs phased the same.
They are two different legs off of the step down transformer up on the
pole or down on the ground some distance from the house.

On my house, there are three wires. If you measure between
one hot and neutral, you get 110V. If you measure between the
two phases, you get 220V (good for water heater, clothes
dryer, stove). The two "legs" are 180 degrees out of phase
with respect to one another. That is why you get 220V between
them, and yet only 110V to neutral.

Here is a poorly drawn figure of the two phases and neutral,
present on the three wires entering my house. Some of my 110V
outlets are wired to one phase, some to the other, to roughly
balance the load on the phases. (To verify this, I looked in
my breaker box, and there is a figure glued to the door, that
says half the breakers are on one phase, and half on the
other.)

___
/ \
/ \ / ------ 0V
\ /
\__/

-------------- ------ 0V

___
/ \
\ / \ ------ 0V
\ /
\__/

As for terminology, I don't know if I'm using the right terms
to describe this or not. In three phase power, the waveforms are
120 degrees apart, and the connection types are delta or wye.
So, I assume it is still valid to call these phases. Maybe we
just have a terminology problem.

Paul

 

[signmeuptoo]

On my house, there are three wires. If you measure between
one hot and neutral, you get 110V. If you measure between the
two phases, you get 220V (good for water heater, clothes
dryer, stove). The two "legs" are 180 degrees out of phase
with respect to one another. That is why you get 220V between
them, and yet only 110V to neutral.

Here is a poorly drawn figure of the two phases and neutral,
present on the three wires entering my house. Some of my 110V
outlets are wired to one phase, some to the other, to roughly
balance the load on the phases. (To verify this, I looked in
my breaker box, and there is a figure glued to the door, that
says half the breakers are on one phase, and half on the
other.)

___
/ \
/ \ / ------ 0V
\ /
\__/

-------------- ------ 0V

___
/ \
\ / \ ------ 0V
\ /
\__/

As for terminology, I don't know if I'm using the right terms
to describe this or not. In three phase power, the waveforms are
120 degrees apart, and the connection types are delta or wye.
So, I assume it is still valid to call these phases. Maybe we
just have a terminology problem.

Paul

No, Paul, you have it half right. Phasing isn't what increases voltage,
in fact, what you just drew, if you combined them would add to zero.
Take it from an electroncis technician and former electrician.

If you have three wires coming down your service drop then what you have
have is 2 independant 117 volt **leggs** from a step down transformer.
The transformer takes power from a single line to do this and steps down
to a couple of seperate windings usually. Sometimes there can be
exceptions. This is, at least what I recall from my days of being a
romex jocky. (I didn't mess with main panels except to add breakers,
change them out, run new branches in the house, work the the power
company, rewire homes, install fixtures, prevent fire damage, refit
against aluminum wiring, consult, etc. I have also been a lighting
consultant. I am a certified Laser Safety/Biomedical equipment
technologiest, a Laser technologist, and electronics technician, and an
optics and vacuums science specialist technician.)

When you combine two equal voltages that are out of phase in
relationship to each other, you ***mitigate*** NOT accentuate, the
cumulative voltage out.

Also, your drawing is deceptive in that you show 0volts AC, as a
baseline, in three different positions. Such a thing cannot exist.
There is always only ONE baseline, be it at a charge above earth, below
it, or at it. With that said, baseline is almost always at ground,
which you do have right. A baseline not at ground is also known as a
"floating ground" and is a very very dangerous thing in home electrical
systems, it will kill.

The third line running down from your service drop is known as the
"common" line. It is common on the transformer to pole earth, house
earth, and power line ground. All tied in together.

Now, with all that explained, here is what happens:

When you have a given AC voltage, say 117, at a given phase, and you add
another voltage, say another 117 volts AC (of course, it MUST be AC or a
catastrophy occurs!) at the same phase, then they add up together just
like standard mathematics. It is quite simple really:

We can always see this on an oscilloscope, but you would need a special
probe so that you wouldn't fry the scope as they only operate in very
low voltages, so the probe would be a negative 1k x X or 100 x X probe.

Here is what you would see:

+117VAC-------------------------------------------------------
/\------------------------------

/ \

/ \
+58V /\-------------------------
/\---------------------------/ \--------------------------
/ \ /
\ / \
/ \ /
\ / \
1st --------\---------+--2nd----------\--------- equals =
/----------------\------240Vttl+/-AC
\ / \
/ \ /
\ / \
/ \ /
-58VAC------ \/-------------------------
\/----------------------------------- \ /--------


\ /

\ /
-117VAC------------------------------------------------------------------------
\ /-----------


The voltages are additive, and as they are superimposed on each other,
they add up together. Phasing is NOT required and would REDUCE over all
voltage.

NOTE: You MUST add the positive and the negative phase peaks to add up
to the total voltage.
If you would like me to draw why, I can, but it would take a lot of
drawing, and it is easier for you to just look in an electronics
textbook. This might still seem unclear to many of you, but you don't
have to make it confusing, it is as simple as 1+1=2. With phasing, it
would be something more like 1+1=0 up to but not 2, depending on the
phase relationship.

Also, I drew using the saw tooth paradigm, as there was no good way to
draw a proper curved way, AND, furthermore, voltages are actually not
really measured as Peak to Peak in real world uses, except in basic
calculations, but in .7 x's X, given that X is the Peak, positive or
negative, and that is why we call house voltage 110Volts AC when it
measures usually between 115 and 124 on most averaging
Multimeter/voltmeters. Averaging is NOT accurate in AC. There are many
complex considerations here, and well out of scope of our conversation.

Suffice it to say, that those 2 legs are both in phase with each other.

NOW, with THAT said, many residences and businesses get a third and some
even a fourth non-common leg down from the power drop/step down
transformer, and **those** additional legs **ARE** different phases of
different voltages. That is how we get commercial voltages such as
277Volts AC for commercial lighting.

I hope that I have done an adequate job explaining my argument. I hope
so, or I have spent the last 30 years of my life at waste and great
expense, aha heh.

Sorry Paul, and now I return to bowing to your superior intellect on all
other computer matters, of course!

 

[Galen Atkinson]

Paul is right.... You are wrong....

As an electrical engineer, let me tell you that the two 117VAC legs
coming into your home are each 117VAC relative to ground and are out of
phase with each other, such that when you measure them relative to each
other there is a voltage potential of 234VAC. If they were in phase,
there would be a zero voltage between them. That is why your electric
water heater, oven, etc all have heating elements that connect across
these two legs. If they were in phase, you would be taking cold showers.

The peak voltage of the actual waveform is greater than 117 volts
however, as the 117VAC is actually a measure of the equivalent DC
voltage that could provide the same amount of power at the same current
level. The 117VAC is actually 0.707 of the peak voltage level of the
waveform. 0.707 is one half of the square root of two which if you
full-wave rectified and filtered the AC power, would yield a DC voltage
that is 0.707 times the peak AC voltage. It can all be easily proven
using simple calculus.

 

[Paul]

Paul is right.... You are wrong....

As an electrical engineer, let me tell you that the two 117VAC legs
coming into your home are each 117VAC relative to ground and are out of
phase with each other, such that when you measure them relative to each
other there is a voltage potential of 234VAC. If they were in phase,
there would be a zero voltage between them. That is why your electric
water heater, oven, etc all have heating elements that connect across
these two legs. If they were in phase, you would be taking cold showers.

The peak voltage of the actual waveform is greater than 117 volts
however, as the 117VAC is actually a measure of the equivalent DC
voltage that could provide the same amount of power at the same current
level. The 117VAC is actually 0.707 of the peak voltage level of the
waveform. 0.707 is one half of the square root of two which if you
full-wave rectified and filtered the AC power, would yield a DC voltage
that is 0.707 times the peak AC voltage. It can all be easily proven
using simple calculus.

To corroborate this:

This posting mentions the pole transformer is center tapped.

http://www.handymanusa.com/forum_nine_msg/4186.htm

Or this posting:

http://www.diyaudio.com/forums/showthread/t-19163.html

"the pole transformer which simply has a single ended input at
2kv and a center tapped 220V output with the center being
neutral and grounded at the pole and service panel of each house."

So, the input to the pole transformer is a single phase, and the
output is center tapped, with each leg being 180 degrees out of
phase, like the diagram I originally drew. Measuring from either
leg to the center tap, gives 110V service, and measuring from
leg to leg gives 220V.

There is some more trivia here:
http://www.cogency.com/B_Support/Data_Communications.pdf

"Some attenuation is also experienced due to the
effects of 2-phase wiring. North American
homes typically have two phases of 110 Volt
wiring (the phases are used in tandem to achieve
the 220 Volts required for large appliances such
as a dryer or oven). Powerline data signals are
naturally coupled from one phase to the other
resulting in an attenuation that is generally less
than 10 dB."

Interference placed on one phase, can show up on the other
phase, but it is attenuated by the pole transformer. The Cogency
document addresses power line data communications.

Paul

 

[Dave McCaleb]

Hi All,
This solution is probably too simple,
BUT, a buddy of mine got a Tower with a 2 position(!) power on swith. This
resulted in his compute booting up for 5 seconds, and then turning itself
off again. Absolutely ridiculous, but believe me, true. Have you checked the
switch?
Dave