4 Bad Motherboards in a row?

[Samson]

This one has me stumped. I build systems all the time, and have 4 orders
for Core 2 Duo systems with the new Gigabyte GA-965P-DQ6 motherboard. I got
2 of the motherboards from Newegg and the other 2 from ZipZoomFly. I have
an E6300, E6400 and 2 x E6600 processors to go with them, along with
DDR2-533 1.8v sticks from Mushkin for all of them. Two of the systems have
Mushkin 550Watt PSUs, and the other two have Antec TruePower2 550s. All 4
have different cases, graphics, optical, hard drives, etc.

The problem:
All four builds are giving me the same exact issue. When powered up for the
first time, all of them each power on for maybe 2-3 seconds, then power off,
then power on again for another 2 seconds, then power off again...and again
forever and ever, until I flip the power switch on the PSU. The only
activity seen when they briefly power on is that the CPU and case fans spin
for a moment. That's it...no graphic output, no BIOS beeps...nada.
Initially, these symptoms made me think that the motherboard was shorting
out or a faulty PSU.

What I've done to verify what's wrong:
1. Tried each motherboard outside of the cases as well as installed
properly in the cases, making sure that no shorts were happening. When
trying them outside of the cases, no case pinouts were connected. I just
shorted the power button pinouts briefly on each board to start them up, in
order to rule out faulty case power button connections.
2. Tried with no peripheral except CPU, heatsink/fan, 1 stick memory and
graphics card. Made sure that the CPU fan was plugged in properly on all
four boards.
3. Tried different combinations of known-good memory (single/dual channel)
in all slots in all boards, as well as different known-good power supplies
and graphics cards on all four boards. These are standard spec 1.8 volt
DDR2-533 memory sticks, chosen to run synchronously with the motherboard/CPU
1066 MHz Front Side Bus.
4. Checked all PSU connections to all boards numerous times, thinking I was
going insane. Each motherboard has a 24-pin ATX connector, 8-pin 12V EPS
connector and a 4-pin standard Molex connector for providing extra 12V to
PCI-Express graphics cards. All 4 power supplies (different reputable
models) have the proper 24-pin, 8-pin and Molex connectors, and they're all
plugged in securely and correctly in each motherboard. All graphics cards
have the 6-pin PCI-Express power cable attached to each card.
5. I even double-checked the UPS power sources that the power supplies were
connected to, to rule out the possibility of something being wrong with the
building's wiring. Other working systems, when plugged into them, work just
fine.

The only common denominator between all four builds is the motherboard and
the brand/type of memory, but each memory stick has been tested in other
systems...so I'm out of ideas, except to come to the conclusion that all 4
motherboards are DOA.

Anyone have any clues? I did see a couple mentions of the same problem with
the same board by Googling forum posts. There is this "CrazyCool heatsink
that Gigabyte puts on the back of the boards, possibly shorting something
else out on each board, but I'm not going to rip them off to find out

I'm just thinking: one board, even maybe two could be a possibility with a
new motherboard release...but 4 in a row from different distributors? I'd
very much appreciate any input anyone could provide.

 

[Dan]

Sorry if I missed this info, but are you making sure to plug in
the 4-pin CPU power plugs on these motherboards?

 

[johns]

I've seen this on one model of Gigabyte mobo.
The problem was incompatibility between the
mobo BIOS and the ATI 9800 graphics card.
The mobo had a VIA chipset, and from that
day on I made sure I only used GA boards
with nVidia chipsets. One thing for sure ..
you should only get bundles from Mwave.com
if you are building for profit and reliability. The
way you are going, you only have warranty
on each part from its respective manufacturer.
That is a lot of hassle to RMA ram to one,
cpu to another, and mobo to another. Nuts
to that. That way, Mwave gets to crank it
up before they ship it, and you won't have
that problem again.

johns

 

[johns]

Forgot. I also ran into another brand of Gigabyte mobo
that I could not seat Kingston ram in to save my life.
It would just not go, even after the "click". I could flip
the board over, and the ram stick would fall out at
one corner.

johns

 

[Samson]

Sorry if I missed this info, but are you making sure to plug in
the 4-pin CPU power plugs on these motherboards?

As stated, these motherboards use the newer EPS 4x4-pin CPU power plugs,
like 2 of the 4-pin plugs put together, and yes, they're plugged in.

 

[Chris Hill]

As stated, these motherboards use the newer EPS 4x4-pin CPU power plugs,
like 2 of the 4-pin plugs put together, and yes, they're plugged in.

Any need to build another machine? If so, buy a different brand mb
and see what happens with it. This kind of thing really takes the fun
out of building systems.

 

[Rod Speed]

This one has me stumped. I build systems all the time, and have 4
orders for Core 2 Duo systems with the new Gigabyte GA-965P-DQ6
motherboard. I got 2 of the motherboards from Newegg and the other 2
from ZipZoomFly. I have an E6300, E6400 and 2 x E6600 processors to
go with them, along with DDR2-533 1.8v sticks from Mushkin for all of
them. Two of the systems have Mushkin 550Watt PSUs, and the other
two have Antec TruePower2 550s. All 4 have different cases,
graphics, optical, hard drives, etc.
The problem:
All four builds are giving me the same exact issue. When powered up
for the first time, all of them each power on for maybe 2-3 seconds,
then power off, then power on again for another 2 seconds, then power
off again...and again forever and ever, until I flip the power switch
on the PSU. The only activity seen when they briefly power on is
that the CPU and case fans spin for a moment. That's it...no graphic
output, no BIOS beeps...nada. Initially, these symptoms made me think
that the motherboard was shorting out or a faulty PSU.

What I've done to verify what's wrong:
1. Tried each motherboard outside of the cases as well as installed
properly in the cases, making sure that no shorts were happening. When trying
them outside of the cases, no case pinouts were
connected. I just shorted the power button pinouts briefly on each
board to start them up, in order to rule out faulty case power button
connections. 2. Tried with no peripheral except CPU, heatsink/fan, 1 stick
memory
and graphics card. Made sure that the CPU fan was plugged in
properly on all four boards.
3. Tried different combinations of known-good memory (single/dual
channel) in all slots in all boards, as well as different known-good
power supplies and graphics cards on all four boards. These are
standard spec 1.8 volt DDR2-533 memory sticks, chosen to run
synchronously with the motherboard/CPU 1066 MHz Front Side Bus.
4. Checked all PSU connections to all boards numerous times,
thinking I was going insane. Each motherboard has a 24-pin ATX
connector, 8-pin 12V EPS connector and a 4-pin standard Molex
connector for providing extra 12V to PCI-Express graphics cards. All
4 power supplies (different reputable models) have the proper 24-pin,
8-pin and Molex connectors, and they're all plugged in securely and
correctly in each motherboard. All graphics cards have the 6-pin
PCI-Express power cable attached to each card. 5. I even double-checked the
UPS power sources that the power
supplies were connected to, to rule out the possibility of something
being wrong with the building's wiring. Other working systems, when
plugged into them, work just fine.

The only common denominator between all four builds is the motherboard and the
brand/type of memory, but each memory stick has been tested in other
systems...

You can eliminate the memory too, try with no memory.
It should keep running and whine about no memory.

Worth trying with no video card too in case its that thats the problem.

Thats the most likely given you dont see too many howling about that
motherboard.

so I'm out of ideas, except to come to the conclusion that all 4 motherboards
are DOA.

If they are that bad, you'd expect the sellers
to be being deluged with pissed off customers.

Anyone have any clues?

You should reset the cmos setting with a jumper etc and make sure
that motherboard doesnt come with the jumper in the reset position.

I did see a couple mentions of the same problem with the same board by
Googling forum posts.

You'd expect more than a couple if there is some fundamental problem.

Can you try it with an older cpu to check if its some Gigabyte stupidity ?
They do tend to shove stuff out the door too quickly and hope they can
fix the remaining design flaws with a new bios. Quite often they end up
having to redesign the motherboard and you can see that in the rev levels.

There is this "CrazyCool heatsink that Gigabyte puts on the back of the
boards, possibly shorting something else out on each board, but I'm not going
to rip them off to find out

Could certainly be that.

I'm just thinking: one board, even maybe two could be a possibility with a
new motherboard release...but 4 in a row from different distributors?

Could obviously be gigabyte not supporting the Core 2 Duo properly yet, which
is why I'd try an older cpu to see if you get the same symptoms with that.

If you dont, try the latest bios.

 

[Samson]

That's for sure, Chris. I do have another order for a system with a
MicroATX 945G-based Core 2 Duo system (Gigabyte GA-945GM-S2 motherboard) and
am expecting shipment of that motherboard in a day or two.

 

[Samson]

Thanks for responding, Rod. I didn't mention it in my original post, but
part of the troubleshooting also included removing all memory and graphics
card to just try powering up with only CPU and heatsink installed to see
whether I'd at least get a beep out of the board. I tried this on 2 of the
four boards (after resetting CMOS by shorting out CMOS jumpers.) Same
problem (just continual power on/power off after 2-3 seconds with no beeps.)
These boards have CMOS jumper pins just underneath the battery, but they
ship without a jumper attached, in order to eliminate the possibility of
someone trying to start up the system for the first time with CMOS clear
jumpers shorted.

I posted a similar message to the Toms Hardware Guide Gigabyte Motherboard
Forum, and soon thereafter, several people replied, reported staying up all
night with the same problem with the same motherboard...so I at least know
I'm not alone.

 

[Rod Speed]

Thanks for responding, Rod. I didn't mention it in my original post,
but part of the troubleshooting also included removing all memory and
graphics card to just try powering up with only CPU and heatsink
installed to see whether I'd at least get a beep out of the board. I
tried this on 2 of the four boards (after resetting CMOS by shorting
out CMOS jumpers.) Same problem (just continual power on/power off
after 2-3 seconds with no beeps.) These boards have CMOS jumper pins
just underneath the battery, but they ship without a jumper attached,
in order to eliminate the possibility of someone trying to start up
the system for the first time with CMOS clear jumpers shorted.

I posted a similar message to the Toms Hardware Guide Gigabyte
Motherboard Forum, and soon thereafter, several people replied,
reported staying up all night with the same problem with the same
motherboard...so I at least know I'm not alone.

OK, I'd see what Gigabyte has to say then.

I'd also use an older cpu and update the bios too if there is a later one.

 

[Samson]

I had a Pentium D 930 lying around, so I installed and tried that on one of
the boards and...still the same.

I'd love to update the BIOS if the could get the $%#@ things to power up
long enough to have a POST screen. I'm waiting on a callback from Mushkin
to try out different memory, but at this point, it looks like actually all 4
motherboard are just DOA.

Incredible.

 

[johns]

My vote is kludged BIOS. I know Gigabyte does that,
and will even have an incompatible setting in that
BIOS which can be the default setting ... example:
GA-K8NS first version out for the AMD 64 Athlon
3000+. I had to put that setting to "high performance",
or the mobo would not boot at all. The next version
of the board ran fine with the default settings. Also,
I discovered that there is no way to upgrade the
original board BIOS without dragging the mobo to
its knees. I had to restore the original BIOS to get
it to work ... and it worked fine like that. Right now,
I have a whole lab full of machines running the latest
version of that board. They all run fine except if I
try to run an image utility from the hard drive, the
utility writes over the boot sector, and I have to sys
the drives to recover. Again, that is clearly a BIOS
problem, and I can get around it by running the
image utillity from a boot floppy ... no problems
at all ... except, if I have to RE-image the drive, I
can do it from the boot floppy, but afterwards,
the D-partition has been deleted ????????? and
I have to restore that partition in Disk Manager,
and run the image utility again. Eh, big deal!
The boards run great otherwise.

johns

 

[Rod Speed]

I had a Pentium D 930 lying around, so I installed and tried that on one of
the boards and...still the same.

Then I'd be kicking Gigabyte.

I'd love to update the BIOS if the could get the $%#@ things to power up long
enough to have a POST screen.

Is there a later bios available ?

I'm waiting on a callback from Mushkin to try out different memory,

Cant be that because you get the same
effect with no memory or video card.

but at this point, it looks like actually all 4 motherboard are just DOA.

Incredible.

Unbelievable, actually. Its gotta be a dud bios that
gigabyte was actually stupid enough to ship it with.

 

[Rod Speed]

My vote is kludged BIOS.

Yeah, mine too.

I know Gigabyte does that, and will even have an incompatible
setting in that BIOS which can be the default setting ... example:
GA-K8NS first version out for the AMD 64 Athlon 3000+. I had to
put that setting to "high performance", or the mobo would not boot
at all. The next version of the board ran fine with the default settings.

Yeah, I avoid Gigabyte for that reason, they ship stuff
too early presumably to get a jump on their competitors.

Also, I discovered that there is no way to upgrade
the original board BIOS without dragging the mobo
to its knees. I had to restore the original BIOS to
get it to work ... and it worked fine like that.

Right now, I have a whole lab full of machines
running the latest version of that board.

Thats the other thing I dont like about Gigabyte, too
many versions of a particular board, evidence that
they ship it too early before its been properly tested.

They all run fine except if I try to run an image utility from
the hard drive, the utility writes over the boot sector,

It wont if you use a decent modern imager.

and I have to sys the drives to recover.
Again, that is clearly a BIOS problem,
Dunno.

and I can get around it by running the image
utillity from a boot floppy ... no problems at all ...
except, if I have to RE-image the drive, I can
do it from the boot floppy, but afterwards,
the D-partition has been deleted ????????? and
I have to restore that partition in Disk Manager,
and run the image utility again. Eh, big deal!

Wota mess.

 

[Samson]

Many others have received this board and have it running fine. Lots of
online reviews singing its praises. I don't know what the shipping BIOS
version on these particular boards I have right now is, probably F1 (if I
could get a POST screen to tell me) but F3 has been out for 3 weeks or so,
which is supposed to improve voltage/memory compatibility.

I'd more suspect just a bad run of hardware, not the BIOS file, due to
others being able to get this board model to work out of the box.

I've used Gigabyte boards for years with no issues to speak of...many more
problems with boards I've used from Asus, Abit and DFI. It looks like
general all-around quality control at all the big board manufacturing plants
in Taiwan is going down, but what do I know.

 

[johns]

Me too. I'm a good customer of the Gigabyte boards.
I just know that they are not perfect, and sometimes
I've had to stand my ground with them about RMA.
Also, I just read one site that the guy was having
problems until he upgraded to the F3 BIOS. Who
knows? Sooner or later Gigabyte comes around
and you get a fix.

johns

 

[Rod Speed]

Many others have received this board and have it running fine.

Not necessarily the same bios as yours tho.

Lots of online reviews singing its praises. I don't know what the shipping
BIOS version on these particular boards I have right now is, probably F1 (if I
could get a POST screen to tell me) but F3 has been out for 3 weeks or so,
which is supposed to improve voltage/memory compatibility.

I'd more suspect just a bad run of hardware, not the BIOS file,

Very bloody unlikely indeed with 4 bads in a row.

due to others being able to get this board model to work out of the box.

Most likely a different bios.

Why dont you ask gigabyte ?

I've used Gigabyte boards for years with no issues to speak of...
many more problems with boards I've used from Asus, Abit and DFI. It looks
like general all-around quality control at all the big board manufacturing
plants in Taiwan is going down,

Havent seen any evidence of that.

 

[Jan Alter]

Not necessarily the same bios as yours tho.



Very bloody unlikely indeed with 4 bads in a row.


Most likely a different bios.

Why dont you ask gigabyte ?


Havent seen any evidence of that.

Have you been using the same make PS on all these builds?

 

[Samson]

Not necessarily the same bios as yours tho.

Well, the board's only been out in the USA for a month or so, and I just
received all of these mobos in when they became in-stock again. It's not
like it was old revision product sitting on Newegg's stock shelves for
months.

Very bloody unlikely indeed with 4 bads in a row.

Really, then why did a Newegg rep tell me yesterday that they had a pallete
of over 200 defective Gigabyte motherboards (DS3 version) that had to be
returned to Taiwan for replacement, per Gigabyte's instructions? That would
indicate to me some sort of batch run hardware issue, would it not?

Most likely a different bios.

Why dont you ask gigabyte ?

That's the first place I contacted. I've heard no reply to 2 online service
request forms filled out yet, but I'll post back what they have to say when
and if I hear from them.

Havent seen any evidence of that.


Indeed.

Thanks for your support. Don't bother to follow this particular thread any
longer if you think the posts on it are coming from drooling idiots with
somewhere in the neighborhood of 3,500+ system builds under their belt; it
will waste your time. I'm just trying to get to the bottom of a motherboard
quality/compatibility issue that is causing problems for myself and other
owners of this board.

 

[Samson]

Have you been using the same make PS on all these builds?


--
Jan Alter
(e-mail address removed)
or
(e-mail address removed)12.pa.us

No Jan. As stated in the original post, I've tried 2 separate Antec
TruePower2 550W PSUs and 2 separate Mushkin HP-550 550W PSUs. I am going to
Mushkin's office tomorrow to have them help me try out alternate memory
using different chips, mainly to finally eliminate the faulty memory
possibility before returning all these motherboards to Gigabyte for RMA.