foxconn 600A01-6LRS cooks video cards

Hi Gang.

I was recently given a foxconn 600A01-6LRS (DDR400 Ram, 1.66 gig AMD
CPU, VIA chipset X4/X8 AGP slot) with the backup bios. Apparently it
keeps cooking graphics cards. I put another old PCI video card in it and
away it went. I can't see any bad blue caps, resisters or diodes on the
mobo (but there may be some I didn't find), so I'm wondering if maybe
someone here would be able to point me in the right direction.

-Adrian

--
Power User - Anyone who can format a disk from DOS.
--
NetscapeMozilla SuiteSeamonkey
A blatant plug for the latest offering in a fine tradition.
http://www.mozilla.org/projects/seamonkey/releases
Browser, Mail & Usenet, HTML Editor and IRC Client.
All in one internet application suite.

Xam wrote:
Hi Gang.

I was recently given a foxconn 600A01-6LRS (DDR400 Ram, 1.66 gig AMD
CPU, VIA chipset X4/X8 AGP slot) with the backup bios. Apparently it
keeps cooking graphics cards. I put another old PCI video card in it and
away it went. I can't see any bad blue caps, resisters or diodes on the
mobo (but there may be some I didn't find), so I'm wondering if maybe
someone here would be able to point me in the right direction.

-Adrian


Specs
http://www.foxconnchannel.com/en-us/...D=en-gb0000169

Picture
http://www.foxconnchannel.com/EN-US/...7V600A-FRS.jpg

It is a KT600/VT8237 motherboard for S462 processors.

The AGP slot is keyed for 1.5V, implying it always runs at 1.5V. That
means, as far as the I/O voltage regulator is concerned, it only
has to support one voltage, which is 1.5V. And that means the interface
won't be paying attention to the TYPEDET# pin. (On universal AGP,
TYPEDET# would be grounded by the video card, as a way for the
video card to say "I prefer 1.5V". And that can be tied into the
linear regulator. But I think the slot on this motherboard,
should be 1.5V all the time.)

http://www.playtool.com/pages/agpcompat/agpslots.gif (yours is second one down)

You'd have to get a copy of the AGP spec, and examine the power
pins, to see a reason why it is frying cards. That would be, on
the assumption it is a power problem. An alternate explanation,
is the Northbridge AGP interface has failed, and perhaps
isn't getting power, and all the I/O pins are sitting at
ground potential. And perhaps shorting something out on the
video card, when it is present.

In my limited experience probing around the AGP slot, I
cannot say I really understood what the devices near the
slot were doing. In other words, the voltages I measured,
didn't add up, on MOSFETs and things.

There appears to be a linear regulator near the DIMM slots.
(8 pin DIP opamp plus two MOSFETs.)

Then, a linear regulator and a wide copper plane, feeding
the Northbridge. Perhaps that provides a Northbridge
core voltage or something. (8 pin DIP opamp plus one MOSFET.)

And that leaves the MOSFET near the faceplate end of the
AGP slot. But I don't see anything adjacent that appears
to be controlling it.

For AGP pinout, you can look here. PDF page 50. VDDQ1.5 would
be I/O power.

http://web.archive.org/web/200303140...0_final_10.pdf

For comparison, AGP20 spec is here. PDF page 231 shows
that VDDQ voltage level, reflects the kind of card plugged
in. A 3.3V I/O only card, has the I/O power signals labeled as
VDDQ3.3. But on your slot, the pins will be VDDQ1.5. The VCC3.3
power pins, are available to power core logic. The VDDQ are dynamic,
as a reflection of the various options for signal levels
on the AGP slot signal pins. VDDQ1.5 powering of signal pads,
handles both 1.5V and 0.8V signalling cases.

http://www.motherboards.org/files/techspecs/agp20.pdf

I have heard of cases, where the VDDQ voltage was out of spec,
but it was on a universal AGP slot. In that case, the video
card failed to pull down the TYPEDET# signal, according to
the method described in the AGP spec. The video card designer
placed a series resistor on the ground connection. That
causes TYPEDET# on the video card, to rest at the wrong
level, and caused the AGP regulator to put out a bit
more than 2 volts. For a 1.5V card. Again, in your case,
the regulator should *not* be looking at TYPEDET#, because
it is only supposed to run at one voltage - and that is 1.5V.

Paul

Paul wrote:
Xam wrote:
Hi Gang.

I was recently given a foxconn 600A01-6LRS (DDR400 Ram, 1.66 gig AMD
CPU, VIA chipset X4/X8 AGP slot) with the backup bios. Apparently it
keeps cooking graphics cards. I put another old PCI video card in it
and away it went. I can't see any bad blue caps, resisters or diodes
on the mobo (but there may be some I didn't find), so I'm wondering if
maybe someone here would be able to point me in the right direction.

-Adrian


Specs
http://www.foxconnchannel.com/en-us/...D=en-gb0000169


Picture
http://www.foxconnchannel.com/EN-US/...7V600A-FRS.jpg


It is a KT600/VT8237 motherboard for S462 processors.

The AGP slot is keyed for 1.5V, implying it always runs at 1.5V. That
means, as far as the I/O voltage regulator is concerned, it only
has to support one voltage, which is 1.5V. And that means the interface
won't be paying attention to the TYPEDET# pin. (On universal AGP,
TYPEDET# would be grounded by the video card, as a way for the
video card to say "I prefer 1.5V". And that can be tied into the
linear regulator. But I think the slot on this motherboard,
should be 1.5V all the time.)

http://www.playtool.com/pages/agpcompat/agpslots.gif (yours is second
one down)

You'd have to get a copy of the AGP spec, and examine the power
pins, to see a reason why it is frying cards. That would be, on
the assumption it is a power problem. An alternate explanation,
is the Northbridge AGP interface has failed, and perhaps
isn't getting power, and all the I/O pins are sitting at
ground potential. And perhaps shorting something out on the
video card, when it is present.

In my limited experience probing around the AGP slot, I
cannot say I really understood what the devices near the
slot were doing. In other words, the voltages I measured,
didn't add up, on MOSFETs and things.

There appears to be a linear regulator near the DIMM slots.
(8 pin DIP opamp plus two MOSFETs.)

Then, a linear regulator and a wide copper plane, feeding
the Northbridge. Perhaps that provides a Northbridge
core voltage or something. (8 pin DIP opamp plus one MOSFET.)

And that leaves the MOSFET near the faceplate end of the
AGP slot. But I don't see anything adjacent that appears
to be controlling it.

For AGP pinout, you can look here. PDF page 50. VDDQ1.5 would
be I/O power.

http://web.archive.org/web/200303140...0_final_10.pdf


For comparison, AGP20 spec is here. PDF page 231 shows
that VDDQ voltage level, reflects the kind of card plugged
in. A 3.3V I/O only card, has the I/O power signals labeled as
VDDQ3.3. But on your slot, the pins will be VDDQ1.5. The VCC3.3
power pins, are available to power core logic. The VDDQ are dynamic,
as a reflection of the various options for signal levels
on the AGP slot signal pins. VDDQ1.5 powering of signal pads,
handles both 1.5V and 0.8V signalling cases.

http://www.motherboards.org/files/techspecs/agp20.pdf

I have heard of cases, where the VDDQ voltage was out of spec,
but it was on a universal AGP slot. In that case, the video
card failed to pull down the TYPEDET# signal, according to
the method described in the AGP spec. The video card designer
placed a series resistor on the ground connection. That
causes TYPEDET# on the video card, to rest at the wrong
level, and caused the AGP regulator to put out a bit
more than 2 volts. For a 1.5V card. Again, in your case,
the regulator should *not* be looking at TYPEDET#, because
it is only supposed to run at one voltage - and that is 1.5V.

Paul

Hi Paul, thanks for your help, it's greatly appreciated.

Yes it's a KT600/VT8237 motherboard for S462 processors. I had to pull
the heat sink off to be 100% sure, as there are now three different
socket types for an AMD Sempron running at 1.66 GHz. Maybe the amount
of L1 cache displayed in the bios would have told me something. But
taking out the heat sink was easer than hooking the box up again to look
at the bios. Now, where did I put that tube of heat sink past? (-:

I'm sorry I didn't mention this at first, but the dead card I pulled out
was a PCI card (64mb DDR Redeon 7000). And I put an old (almost fried
it's self) trident 4mb card back in the same (upper most) PCI slot.

The owner said it keeps "shorting out video cards", but isn't computer
savy enough to know if any of them were AGP slot cards or not. It's a
shame, because it would make tracking down the fault a whole lot easier.

I was hoping it would be some easy to replace component, you know, three
leads or less!. (-: As I am yet to own a de-solder gun. )-:

There seems to be a lot of older video games installed, and the system
only had 256mb of ram, and no case fan. Is it possible that the cards
simply just overheated? They certainly wouldn't have been able to
access much system ram to play games. The bios was set to allow 64mb
for video use. But I'm guessing they would have been lucky to get
anywhere near that much.

And thanks for all the useful links. If/when I do get down to measuring
voltages, they will come in very handy.

-Adrian



--
Oxymoron: Microsoft Works.
--
NetscapeMozilla SuiteSeamonkey
A blatant plug for the latest offering in a fine tradition.
http://www.mozilla.org/projects/seamonkey/releases
Browser, Mail & Usenet, HTML Editor and IRC Client.
All in one internet application suite.

Xam wrote:


Hi Paul, thanks for your help, it's greatly appreciated.

Yes it's a KT600/VT8237 motherboard for S462 processors. I had to pull
the heat sink off to be 100% sure, as there are now three different
socket types for an AMD Sempron running at 1.66 GHz. Maybe the amount
of L1 cache displayed in the bios would have told me something. But
taking out the heat sink was easer than hooking the box up again to look
at the bios. Now, where did I put that tube of heat sink past? (-:

I'm sorry I didn't mention this at first, but the dead card I pulled out
was a PCI card (64mb DDR Redeon 7000). And I put an old (almost fried
it's self) trident 4mb card back in the same (upper most) PCI slot.

The owner said it keeps "shorting out video cards", but isn't computer
savy enough to know if any of them were AGP slot cards or not. It's a
shame, because it would make tracking down the fault a whole lot easier.

I was hoping it would be some easy to replace component, you know, three
leads or less!. (-: As I am yet to own a de-solder gun. )-:

There seems to be a lot of older video games installed, and the system
only had 256mb of ram, and no case fan. Is it possible that the cards
simply just overheated? They certainly wouldn't have been able to
access much system ram to play games. The bios was set to allow 64mb
for video use. But I'm guessing they would have been lucky to get
anywhere near that much.

And thanks for all the useful links. If/when I do get down to measuring
voltages, they will come in very handy.

-Adrian


If the owner of the system wasn't very careful with their usage of drivers,
perhaps you can tell what kind of cards were present, by looking at the
driver packages that are still installed. You know, install a Matrox card,
install Matrox drivers, pull card, slap in Nvidia, install Nvidia driver,
leaving Matrox ones sitting in Add/Remove. Maybe you can guess at the machine
history, from the accumulated junk ?

For the cards that were "shorted out", did he get to keep them, or
were the cards kept by whoever was "helping" him ?

Paul

Paul wrote:
Xam wrote:


Hi Paul, thanks for your help, it's greatly appreciated.

Yes it's a KT600/VT8237 motherboard for S462 processors. I had to
pull the heat sink off to be 100% sure, as there are now three
different socket types for an AMD Sempron running at 1.66 GHz. Maybe
the amount of L1 cache displayed in the bios would have told me
something. But taking out the heat sink was easer than hooking the
box up again to look at the bios. Now, where did I put that tube of
heat sink past? (-:

I'm sorry I didn't mention this at first, but the dead card I pulled
out was a PCI card (64mb DDR Redeon 7000). And I put an old (almost
fried it's self) trident 4mb card back in the same (upper most) PCI slot.

The owner said it keeps "shorting out video cards", but isn't computer
savy enough to know if any of them were AGP slot cards or not. It's a
shame, because it would make tracking down the fault a whole lot easier.

I was hoping it would be some easy to replace component, you know,
three leads or less!. (-: As I am yet to own a de-solder gun. )-:

There seems to be a lot of older video games installed, and the system
only had 256mb of ram, and no case fan. Is it possible that the cards
simply just overheated? They certainly wouldn't have been able to
access much system ram to play games. The bios was set to allow 64mb
for video use. But I'm guessing they would have been lucky to get
anywhere near that much.

And thanks for all the useful links. If/when I do get down to
measuring voltages, they will come in very handy.

-Adrian


If the owner of the system wasn't very careful with their usage of drivers,
perhaps you can tell what kind of cards were present, by looking at the
driver packages that are still installed. You know, install a Matrox card,
install Matrox drivers, pull card, slap in Nvidia, install Nvidia driver,
leaving Matrox ones sitting in Add/Remove. Maybe you can guess at the
machine
history, from the accumulated junk ?

For the cards that were "shorted out", did he get to keep them, or
were the cards kept by whoever was "helping" him ?

Paul


I seem to remember something about an ATI card in add/remove. I will
have a look and get back to you.

I think he would have taken it to a shop, so probably doesn't have the
old cards.

-Adrian

--
'You've got a magic carpet for three people to fly to the King of the
Potato People to plead your case and you're trying to tell me you're
sane?' -Rimmer, Red Dwarf.

--
NetscapeMozilla SuiteSeamonkey
A blatant plug for the latest offering in a fine tradition.
http://www.mozilla.org/projects/seamonkey/releases
Browser, Mail & Usenet, HTML Editor and IRC Client.
All in one internet application suite.

On Sun, 01 Jun 2008 11:36:49 +1000, Xam wrote:

Hi Gang.

I was recently given a foxconn 600A01-6LRS (DDR400 Ram, 1.66 gig AMD
CPU, VIA chipset X4/X8 AGP slot) with the backup bios. Apparently it
keeps cooking graphics cards. I put another old PCI video card in it and
away it went. I can't see any bad blue caps, resisters or diodes on the
mobo (but there may be some I didn't find), so I'm wondering if maybe
someone here would be able to point me in the right direction.


I suspect a capacitor(s) on the 5V rail inside the power
supply has failed. Unplug power supply from AC, remove it,
wait a few minutes then open the lid and look at the
capacitors near the exiting wiring harness.

If it is the PSU, make sure you get one rated for 200W
combined 3V+5V rating, not one of the newer models biased
mostly for 12V current as it may not regulate as well and
may cost quite a bit more to get the necessary capacity on
the 5V rail.

The system should have a case fan added. This may have been
a contributory factor and if the case is set up that poorly
it seems more likely it would also have poor front air
intake area. Given these two possiblities, also examine the
motherboard for failed capacitors - vents tops or bottoms
possibly with a leaky, crusy residue.

Xam wrote:
Paul wrote:
Xam wrote:


Hi Paul, thanks for your help, it's greatly appreciated.

Yes it's a KT600/VT8237 motherboard for S462 processors. I had to
pull the heat sink off to be 100% sure, as there are now three
different socket types for an AMD Sempron running at 1.66 GHz. Maybe
the amount of L1 cache displayed in the bios would have told me
something. But taking out the heat sink was easer than hooking the
box up again to look at the bios. Now, where did I put that tube of
heat sink past? (-:

I'm sorry I didn't mention this at first, but the dead card I pulled
out was a PCI card (64mb DDR Redeon 7000). And I put an old (almost
fried it's self) trident 4mb card back in the same (upper most) PCI
slot.

The owner said it keeps "shorting out video cards", but isn't
computer savy enough to know if any of them were AGP slot cards or
not. It's a shame, because it would make tracking down the fault a
whole lot easier.

I was hoping it would be some easy to replace component, you know,
three leads or less!. (-: As I am yet to own a de-solder gun. )-:

There seems to be a lot of older video games installed, and the
system only had 256mb of ram, and no case fan. Is it possible that
the cards simply just overheated? They certainly wouldn't have been
able to access much system ram to play games. The bios was set to
allow 64mb for video use. But I'm guessing they would have been
lucky to get anywhere near that much.

And thanks for all the useful links. If/when I do get down to
measuring voltages, they will come in very handy.

-Adrian


If the owner of the system wasn't very careful with their usage of
drivers,
perhaps you can tell what kind of cards were present, by looking at the
driver packages that are still installed. You know, install a Matrox
card,
install Matrox drivers, pull card, slap in Nvidia, install Nvidia driver,
leaving Matrox ones sitting in Add/Remove. Maybe you can guess at the
machine
history, from the accumulated junk ?

For the cards that were "shorted out", did he get to keep them, or
were the cards kept by whoever was "helping" him ?

Paul


I seem to remember something about an ATI card in add/remove. I will
have a look and get back to you.

I think he would have taken it to a shop, so probably doesn't have the
old cards.

-Adrian



No, the only other driver was the ATI Radeon™ 7000 PCI card I pulled out.

But I will give the guy a call, and see if he remembers anything more
specific.

-Adrian

PS. I'm really going to have to replace the four fried caps in that old
trident of mine, it's the only spare PCI video card I have left.

kony wrote:
On Sun, 01 Jun 2008 11:36:49 +1000, Xam wrote:

Hi Gang.

I was recently given a foxconn 600A01-6LRS (DDR400 Ram, 1.66 gig AMD
CPU, VIA chipset X4/X8 AGP slot) with the backup bios. Apparently it
keeps cooking graphics cards. I put another old PCI video card in it and
away it went. I can't see any bad blue caps, resisters or diodes on the
mobo (but there may be some I didn't find), so I'm wondering if maybe
someone here would be able to point me in the right direction.


I suspect a capacitor(s) on the 5V rail inside the power
supply has failed. Unplug power supply from AC, remove it,
wait a few minutes then open the lid and look at the
capacitors near the exiting wiring harness.

If it is the PSU, make sure you get one rated for 200W
combined 3V+5V rating, not one of the newer models biased
mostly for 12V current as it may not regulate as well and
may cost quite a bit more to get the necessary capacity on
the 5V rail.

The system should have a case fan added. This may have been
a contributory factor and if the case is set up that poorly
it seems more likely it would also have poor front air
intake area. Given these two possiblities, also examine the
motherboard for failed capacitors - vents tops or bottoms
possibly with a leaky, crusy residue.

Okay Kony, will do.
Thanks for that.
One other thing the owner said was that he had taken it to the local
computer shop, and they said that the PSU was good. But they would have
only put a multi-meter on the harnesses. They definitely didn't take
the lid off, you should see the smeg floating around in there. (-:

I'll get back to you on the state of the PSU, but I had yet another look
at the motherboard while I had the heat sink out, and didn't see any
obviously fried components. Unlike four of the caps on the PCI video
card I put in!, which still just works bty. (-:

-Adrian

Xam wrote:
kony wrote:
On Sun, 01 Jun 2008 11:36:49 +1000, Xam wrote:

Hi Gang.

I was recently given a foxconn 600A01-6LRS (DDR400 Ram, 1.66 gig AMD
CPU, VIA chipset X4/X8 AGP slot) with the backup bios. Apparently it
keeps cooking graphics cards. I put another old PCI video card in it
and away it went. I can't see any bad blue caps, resisters or diodes
on the mobo (but there may be some I didn't find), so I'm wondering
if maybe someone here would be able to point me in the right direction.


I suspect a capacitor(s) on the 5V rail inside the power
supply has failed. Unplug power supply from AC, remove it,
wait a few minutes then open the lid and look at the
capacitors near the exiting wiring harness.

If it is the PSU, make sure you get one rated for 200W
combined 3V+5V rating, not one of the newer models biased
mostly for 12V current as it may not regulate as well and
may cost quite a bit more to get the necessary capacity on
the 5V rail.

The system should have a case fan added. This may have been
a contributory factor and if the case is set up that poorly
it seems more likely it would also have poor front air
intake area. Given these two possiblities, also examine the
motherboard for failed capacitors - vents tops or bottoms
possibly with a leaky, crusy residue.

Okay Kony, will do.
Thanks for that.
One other thing the owner said was that he had taken it to the local
computer shop, and they said that the PSU was good. But they would have
only put a multi-meter on the harnesses. They definitely didn't take
the lid off, you should see the smeg floating around in there. (-:

I'll get back to you on the state of the PSU, but I had yet another look
at the motherboard while I had the heat sink out, and didn't see any
obviously fried components. Unlike four of the caps on the PCI video
card I put in!, which still just works bty. (-:

-Adrian



Well, the PSU looks to be in good nick. It's a Auriga 9806B with a hard
power 1/0 switch. Although it's only 300 watts total, (+12v 10A) (-12v
0.8A) (+5 VSB 2A), with 160 watt for the +5v(30A) -5v(0.5A) & +3.3v(14A)
rails combined.

I still haven't rung up the owner to ask more closely about it's
history. Although that might now have to weight till tomorrow.

Thanks for your help Kony.
As I said to Paul, I really do appreciate you guys giving up your time
to help people who you will probably never even meet.

-Adrian

Paul wrote:
Xam wrote:
Hi Gang.

I was recently given a foxconn 600A01-6LRS (DDR400 Ram, 1.66 gig AMD
CPU, VIA chipset X4/X8 AGP slot) with the backup bios. Apparently it
keeps cooking graphics cards. I put another old PCI video card in it
and away it went. I can't see any bad blue caps, resisters or diodes
on the mobo (but there may be some I didn't find), so I'm wondering
if maybe someone here would be able to point me in the right
direction. -Adrian


Specs
http://www.foxconnchannel.com/en-us/...D=en-gb0000169

Picture
http://www.foxconnchannel.com/EN-US/...7V600A-FRS.jpg

It is a KT600/VT8237 motherboard for S462 processors.

The AGP slot is keyed for 1.5V, implying it always runs at 1.5V. That
means, as far as the I/O voltage regulator is concerned, it only
has to support one voltage, which is 1.5V. And that means the
interface won't be paying attention to the TYPEDET# pin. (On universal AGP,
TYPEDET# would be grounded by the video card, as a way for the
video card to say "I prefer 1.5V". And that can be tied into the
linear regulator. But I think the slot on this motherboard,
should be 1.5V all the time.)

http://www.playtool.com/pages/agpcompat/agpslots.gif (yours is
second one down)
You'd have to get a copy of the AGP spec, and examine the power
pins, to see a reason why it is frying cards. That would be, on
the assumption it is a power problem. An alternate explanation,
is the Northbridge AGP interface has failed, and perhaps
isn't getting power, and all the I/O pins are sitting at
ground potential. And perhaps shorting something out on the
video card, when it is present.

In my limited experience probing around the AGP slot, I
cannot say I really understood what the devices near the
slot were doing. In other words, the voltages I measured,
didn't add up, on MOSFETs and things.

There appears to be a linear regulator near the DIMM slots.
(8 pin DIP opamp plus two MOSFETs.)

Then, a linear regulator and a wide copper plane, feeding
the Northbridge. Perhaps that provides a Northbridge
core voltage or something. (8 pin DIP opamp plus one MOSFET.)

And that leaves the MOSFET near the faceplate end of the
AGP slot. But I don't see anything adjacent that appears
to be controlling it.

For AGP pinout, you can look here. PDF page 50. VDDQ1.5 would
be I/O power.

http://web.archive.org/web/200303140...0_final_10.pdf

For comparison, AGP20 spec is here. PDF page 231 shows
that VDDQ voltage level, reflects the kind of card plugged
in. A 3.3V I/O only card, has the I/O power signals labeled as
VDDQ3.3. But on your slot, the pins will be VDDQ1.5. The VCC3.3
power pins, are available to power core logic. The VDDQ are dynamic,
as a reflection of the various options for signal levels
on the AGP slot signal pins. VDDQ1.5 powering of signal pads,
handles both 1.5V and 0.8V signalling cases.

http://www.motherboards.org/files/techspecs/agp20.pdf

I have heard of cases, where the VDDQ voltage was out of spec,
but it was on a universal AGP slot. In that case, the video
card failed to pull down the TYPEDET# signal, according to
the method described in the AGP spec. The video card designer
placed a series resistor on the ground connection. That
causes TYPEDET# on the video card, to rest at the wrong
level, and caused the AGP regulator to put out a bit
more than 2 volts. For a 1.5V card. Again, in your case,
the regulator should *not* be looking at TYPEDET#, because
it is only supposed to run at one voltage - and that is 1.5V.

Doesnt explain why it killed a PCI video card, if it did that.