ga-7n400 Pro - no boot to bios

I have been testing my new ga-7n499 Pro, with AMD 2500+ and 512MB
ddr3200

today looks like i made a booboo

History:

Since the 2500+ was rated at 333, I bought 2700ddr. When I ramped up
to spec speed, 166/333 11x for 1834Mhz, I got lots of errors with
memtest86 v3.0. I opted to exchange my memory for 3200ddr and it
passed an 8 hr memtest86 assault.

For information purposes, had I kept the 2700ddr, what tweaks would I
have had to make to make the memory run stable at spec speeds? up
some voltages? which?

bios reported vcore = 1.646, ddr25v = 2.608, 3.3v = 3.312, 5v = 4.960
and 12v = 11.541
easytune4 reported slightly different readings once i booted to
windows: vcore = 1.640, 3.3v = 3.310, 5v= 4.960 and 12v = 11.710. I
didn't see a ddrV reading in ET4.

Are these typical, and which could I, should I tweak for reducing mem
errors like i was experiencing?

Now to the current 'crisis'...

Since my 3200ddr was running fine at 333, I decided to test it at 400.
But the way I did it I think was not so good. Maybe it exposed a flaw
in the bios design, I am not sure..

The bios was set to 166mhz and 'mem freq' was 100% which gave result
of 333 for fsb. I changed 'mem freq' to 'by spd' which changed
resulting mem clock? to 400, then clicked 'save and exit.

On exit, the screen blanked and the puter emitted a long tone,
repeatedly. It seemed to be hanging, so i pressed the reset button.
The lights on the HD and CD blinked, but nothing else happened. I
pressed the power off button. Nothing happened, not even blinking
lights. I pressed and held the power switch for about 5 sec and the
system finally powered down. I pressed it again to boot up. Nothing
appeared on the screen, and the same long recurring beep started. p
102 of the manual indicates a continuous long beep as being a dram
error.

So i figured i'd go into bios and undo the change from mem freq '100%
to 'by SPD' but now I can't even get into bios. All I get on boot is
a blank screen, with maybe a blinking cursor.. not sure even about
that detail atm.

What now?

Do I need to clear cmos? How? I looked on the MB and do not see a
jumper to clear bios cmos. Do I just take out the battery for a
minute?

Thanks in advance for any hints and facts.

On Thu, 14 Aug 2003 00:01:55 GMT, NewB wrote:

I have been testing my new ga-7n499 Pro, with AMD 2500+ and 512MB
ddr3200

today looks like i made a booboo

History:

Since the 2500+ was rated at 333, I bought 2700ddr. When I ramped up
to spec speed, 166/333 11x for 1834Mhz, I got lots of errors with
memtest86 v3.0. I opted to exchange my memory for 3200ddr and it
passed an 8 hr memtest86 assault.

For information purposes, had I kept the 2700ddr, what tweaks would I
have had to make to make the memory run stable at spec speeds? up
some voltages? which?

bios reported vcore = 1.646, ddr25v = 2.608, 3.3v = 3.312, 5v = 4.960
and 12v = 11.541
easytune4 reported slightly different readings once i booted to
windows: vcore = 1.640, 3.3v = 3.310, 5v= 4.960 and 12v = 11.710. I
didn't see a ddrV reading in ET4.

Are these typical, and which could I, should I tweak for reducing mem
errors like i was experiencing?

Now to the current 'crisis'...

Since my 3200ddr was running fine at 333, I decided to test it at 400.
But the way I did it I think was not so good. Maybe it exposed a flaw
in the bios design, I am not sure..

The bios was set to 166mhz and 'mem freq' was 100% which gave result
of 333 for fsb. I changed 'mem freq' to 'by spd' which changed
resulting mem clock? to 400, then clicked 'save and exit.

On exit, the
screen blanked and the puter emitted a long tone,
repeatedly. It seemed to be hanging, so i pressed the reset button.
The lights on the HD and CD blinked, but nothing else happened. I
pressed the power off button. Nothing happened, not even blinking
lights. I pressed and held the power switch for about 5 sec and the
system finally powered down. I pressed it again to boot up. Nothing
appeared on the screen, and the same long recurring beep started. p
102 of the manual indicates a continuous long beep as being a dram
error.

So i figured i'd go into bios and undo the change from mem freq '100%
to 'by SPD' but now I can't even get into bios. All I get on boot is
a blank screen, with maybe a blinking cursor.. not sure even about
that detail atm.

What now?

Do I need to clear cmos? How? I looked on the MB and do not see a
jumper to clear bios cmos. Do I just take out the battery for a
minute?

Thanks in advance for any hints and facts.

Read your mobo manual on how to clear the CMOS, usually found near the
end of the manual, it's an easy thing to do but some boards require
different steps taken.

But I think they all tell you to shut off the PSU, unplug the PC from
the wall socket if your not sure, wait a minute then proceed to clearing
it according to your mobo makers instructions.

Ed

Hi
I have the same mobo, you need to take the battery out for a few mins to
clear the BIOS. Stupind way to do it, much prefered the jumper on my old
ABIT board.

Paul

"Ed" wrote in message
news
On Thu, 14 Aug 2003 00:01:55 GMT, NewB wrote:

I have been testing my new ga-7n499 Pro, with AMD 2500+ and 512MB
ddr3200

today looks like i made a booboo

History:

Since the 2500+ was rated at 333, I bought 2700ddr. When I ramped up
to spec speed, 166/333 11x for 1834Mhz, I got lots of errors with
memtest86 v3.0. I opted to exchange my memory for 3200ddr and it
passed an 8 hr memtest86 assault.

For information purposes, had I kept the 2700ddr, what tweaks would I
have had to make to make the memory run stable at spec speeds? up
some voltages? which?

bios reported vcore = 1.646, ddr25v = 2.608, 3.3v = 3.312, 5v = 4.960
and 12v = 11.541
easytune4 reported slightly different readings once i booted to
windows: vcore = 1.640, 3.3v = 3.310, 5v= 4.960 and 12v = 11.710. I
didn't see a ddrV reading in ET4.

Are these typical, and which could I, should I tweak for reducing mem
errors like i was experiencing?

Now to the current 'crisis'...

Since my 3200ddr was running fine at 333, I decided to test it at 400.
But the way I did it I think was not so good. Maybe it exposed a flaw
in the bios design, I am not sure..

The bios was set to 166mhz and 'mem freq' was 100% which gave result
of 333 for fsb. I changed 'mem freq' to 'by spd' which changed
resulting mem clock? to 400, then clicked 'save and exit.

On exit, the
screen blanked and the puter emitted a long tone,
repeatedly. It seemed to be hanging, so i pressed the reset button.
The lights on the HD and CD blinked, but nothing else happened. I
pressed the power off button. Nothing happened, not even blinking
lights. I pressed and held the power switch for about 5 sec and the
system finally powered down. I pressed it again to boot up. Nothing
appeared on the screen, and the same long recurring beep started. p
102 of the manual indicates a continuous long beep as being a dram
error.

So i figured i'd go into bios and undo the change from mem freq '100%
to 'by SPD' but now I can't even get into bios. All I get on boot is
a blank screen, with maybe a blinking cursor.. not sure even about
that detail atm.

What now?

Do I need to clear cmos? How? I looked on the MB and do not see a
jumper to clear bios cmos. Do I just take out the battery for a
minute?

Thanks in advance for any hints and facts.

Read your mobo manual on how to clear the CMOS, usually found near the
end of the manual, it's an easy thing to do but some boards require
different steps taken.

But I think they all tell you to shut off the PSU, unplug the PC from
the wall socket if your not sure, wait a minute then proceed to clearing
it according to your mobo makers instructions.

Ed

I had a feeling I shouldn't have tried to ask about multiple things,
so I am separating the issues.

Additional information: dimms were in slots 1 and 3 and tested fine
with 400 mem at 333fsb. I assume it was in dual channel mode to
because I had a dimm in each bank.

Is the dual-channel ddr feature automatic, controlled solely by
placement of memory in separate banks per the manual charts? Is there
a way to confirm if it is in dual channel mode or do you just assume
it is if memory is installed in both banks per chart?

I would still like input on the prior questions below too, as no one
has of yet replied regarding them.

Thank you.


On Thu, 14 Aug 2003 00:01:55 GMT, NewB opined in
alt.comp.hardware.overclocking.amd:

I have been testing my new ga-7n499 Pro, with AMD 2500+ and 512MB
ddr3200

today looks like i made a booboo

History:

Since the 2500+ was rated at 333, I bought 2700ddr. When I ramped up
to spec speed, 166/333 11x for 1834Mhz, I got lots of errors with
memtest86 v3.0. I opted to exchange my memory for 3200ddr and it
passed an 8 hr memtest86 assault.

[new note: memory was in slots 1 and 3 for this test]

For information purposes, had I kept the 2700ddr, what tweaks would I
have had to make to make the memory run stable at spec speeds? up
some voltages? which?

bios reported vcore = 1.646, ddr25v = 2.608, 3.3v = 3.312, 5v = 4.960
and 12v = 11.541
easytune4 reported slightly different readings once i booted to
windows: vcore = 1.640, 3.3v = 3.310, 5v= 4.960 and 12v = 11.710. I
didn't see a ddrV reading in ET4.

Are these typical, and which could I, should I tweak for reducing mem
errors like i was experiencing?

when i get my system working again, I would like to have the above
information as to proceed safely.

On Thu, 14 Aug 2003 08:34:31 GMT, NewB wrote:

I had a feeling I shouldn't have tried to ask about multiple things,
so I am separating the issues.

Additional information: dimms were in slots 1 and 3 and tested fine
with 400 mem at 333fsb. I assume it was in dual channel mode to
because I had a dimm in each bank.

Is the dual-channel ddr feature automatic, controlled solely by
placement of memory in separate banks per the manual charts? Is there
a way to confirm if it is in dual channel mode or do you just assume
it is if memory is installed in both banks per chart?

PAGE 21....
Dual Channel DDR:
GA-7NNXPV / GA-7NNXP / GA-7N400V Pro / GA-7N400 Pro / GA-7N400-L1
support Dual Channel
Technology. When Dual Channel Technology is activated, the bandwidth of
memory bus will be double
the original one, with the fastest speed at 6.4GB/s(DDR400) or
5.3GB/s(DDR333).
GA-7NNXPV / GA-7NNXP / GA-7N400V Pro / GA-7N400 Pro / GA-7N400-L1
include 4 DIMM slots, and
each Channel has 2 DIMMs as following:
Channel A : DIMM 1, 2
Channel B : DIMM 3, 4

Below are the explanations:
If you want to operate the Dual Channel Technology, please note the
following explanations
due to the limitation of Intel chipset specifications.
1. Only one DDR memory module is installed: The Dual Channel Technology
can't
operate when only one DDR memory module is installed.
2. Two DDR memory modules are installed: The Dual Channel Technology
will operate
when two memory modules are inserted individually into Channel A and B.
If you install
two memory modules in the same channel, the Dual Channel Technology will
not
operate.
3. Three or Four DDR memory modules are installed: Please follow figure
1 to achieve the
Dual Technology.


I would still like input on the prior questions below too, as no one
has of yet replied regarding them.

Thank you.


On Thu, 14 Aug 2003 00:01:55 GMT, NewB opined in
alt.comp.hardware.overclocking.amd:

I have been testing my new ga-7n499 Pro, with AMD 2500+ and 512MB
ddr3200

today looks like i made a booboo

History:

Since the 2500+ was rated at 333, I bought 2700ddr. When I ramped up
to spec speed, 166/333 11x for 1834Mhz, I got lots of errors with
memtest86 v3.0. I opted to exchange my memory for 3200ddr and it
passed an 8 hr memtest86 assault.

[new note: memory was in slots 1 and 3 for this test]

For information purposes, had I kept the 2700ddr, what tweaks would I
have had to make to make the memory run stable at spec speeds? up
some voltages? which?

bios reported vcore = 1.646, ddr25v = 2.608, 3.3v = 3.312, 5v = 4.960
and 12v = 11.541
easytune4 reported slightly different readings once i booted to
windows: vcore = 1.640, 3.3v = 3.310, 5v= 4.960 and 12v = 11.710. I
didn't see a ddrV reading in ET4.

Are these typical, and which could I, should I tweak for reducing mem
errors like i was experiencing?

when i get my system working again, I would like to have the above
information as to proceed safely.

I have two lots of 512mb in my board but didn't get the duall channel
working until I updated with an unofficial BIOS update. It now tells me when
I boot up that the Dual Channel is enabled. Try www.nforcershq.com and look
under the Gigabyte forum.

Paul

"Ed" wrote in message
...
On Thu, 14 Aug 2003 08:34:31 GMT, NewB wrote:

I had a feeling I shouldn't have tried to ask about multiple things,
so I am separating the issues.

Additional information: dimms were in slots 1 and 3 and tested fine
with 400 mem at 333fsb. I assume it was in dual channel mode to
because I had a dimm in each bank.

Is the dual-channel ddr feature automatic, controlled solely by
placement of memory in separate banks per the manual charts? Is there
a way to confirm if it is in dual channel mode or do you just assume
it is if memory is installed in both banks per chart?

PAGE 21....
Dual Channel DDR:
GA-7NNXPV / GA-7NNXP / GA-7N400V Pro / GA-7N400 Pro / GA-7N400-L1
support Dual Channel
Technology. When Dual Channel Technology is activated, the bandwidth of
memory bus will be double
the original one, with the fastest speed at 6.4GB/s(DDR400) or
5.3GB/s(DDR333).
GA-7NNXPV / GA-7NNXP / GA-7N400V Pro / GA-7N400 Pro / GA-7N400-L1
include 4 DIMM slots, and
each Channel has 2 DIMMs as following:
Channel A : DIMM 1, 2
Channel B : DIMM 3, 4

Below are the explanations:
If you want to operate the Dual Channel Technology, please note the
following explanations
due to the limitation of Intel chipset specifications.
1. Only one DDR memory module is installed: The Dual Channel Technology
can't
operate when only one DDR memory module is installed.
2. Two DDR memory modules are installed: The Dual Channel Technology
will operate
when two memory modules are inserted individually into Channel A and B.
If you install
two memory modules in the same channel, the Dual Channel Technology will
not
operate.
3. Three or Four DDR memory modules are installed: Please follow figure
1 to achieve the
Dual Technology.


I would still like input on the prior questions below too, as no one
has of yet replied regarding them.

Thank you.


On Thu, 14 Aug 2003 00:01:55 GMT, NewB opined in
alt.comp.hardware.overclocking.amd:

I have been testing my new ga-7n499 Pro, with AMD 2500+ and 512MB
ddr3200

today looks like i made a booboo

History:

Since the 2500+ was rated at 333, I bought 2700ddr. When I ramped up
to spec speed, 166/333 11x for 1834Mhz, I got lots of errors with
memtest86 v3.0. I opted to exchange my memory for 3200ddr and it
passed an 8 hr memtest86 assault.

[new note: memory was in slots 1 and 3 for this test]

For information purposes, had I kept the 2700ddr, what tweaks would I
have had to make to make the memory run stable at spec speeds? up
some voltages? which?

bios reported vcore = 1.646, ddr25v = 2.608, 3.3v = 3.312, 5v = 4.960
and 12v = 11.541
easytune4 reported slightly different readings once i booted to
windows: vcore = 1.640, 3.3v = 3.310, 5v= 4.960 and 12v = 11.710. I
didn't see a ddrV reading in ET4.

Are these typical, and which could I, should I tweak for reducing mem
errors like i was experiencing?

when i get my system working again, I would like to have the above
information as to proceed safely.

On Wed, 13 Aug 2003 19:27:46 -0500, Ed opined in
alt.comp.hardware.overclocking.amd:

On Thu, 14 Aug 2003 00:01:55 GMT, NewB wrote:

I have been testing my new ga-7n400 Pro, with AMD 2500+ and 512MB
ddr3200

[installed in slots 1 and 3]

bios v. F9

today looks like i made a booboo

or uncovered a glitch in the bios coding?

I was running spec settings:
166 clock for 333fsb
11x multiplier for 1837mhz CPU speed
[afaik, mobo and mem is rated at 400fsb and amd 2500+ 333 so I think
my settings were conservative and not over clocked at all. then...

Since my 3200ddr was running fine at 333, I decided to test it at 400.
But the way I did it I think was not so good. Maybe it exposed a flaw
in the bios design, I am not sure..

The bios was set to 166mhz and 'mem freq' was 100% which gave result
of 333 for fsb. I changed 'mem freq' to 'by spd' which changed
resulting mem clock? to 400, then clicked 'save and exit.

On exit, the
screen blanked and the puter emitted a long tone,
repeatedly. It seemed to be hanging, so i pressed the reset button.
The lights on the HD and CD blinked, but nothing else happened. I
pressed the power off button. Nothing happened, not even blinking
lights. I pressed and held the power switch for about 5 sec and the
system finally powered down. I pressed it again to boot up. Nothing
appeared on the screen, and the same long recurring beep started. p
102 of the manual indicates a continuous long beep as being a dram
error.

So i figured i'd go into bios and undo the change from mem freq '100%
to 'by SPD' but now I can't even get into bios. All I get on boot is
a blank screen, with maybe a blinking cursor.. not sure even about
that detail atm.

What now?

Do I need to clear cmos? How? I looked on the MB and do not see a
jumper to clear bios cmos. Do I just take out the battery for a
minute?

Thanks in advance for any hints and facts.

Read your mobo manual on how to clear the CMOS, usually found near the
end of the manual, it's an easy thing to do but some boards require
different steps taken.

But I think they all tell you to shut off the PSU, unplug the PC from
the wall socket if your not sure, wait a minute then proceed to clearing
it according to your mobo makers instructions.

Ed


I thought I had read the entire manual a few times, before posting,
but do admit I missed this. The manual information is ambiguous, and
not specific for this board family, much unlike the nice board
specific details elsewhere through out the manual. I found no jumper
on my board, so followed plan B and removed the cmos battery for a
while, shorted the contacts and reinstalled the battery.

This had no noticeable effect. On boot the continuous long tone was
still there and no bios boot/info screen.

On a hunch, and for no logical reason I can understand, I decided to
pull one of the dimm modules [from slot 3] and reboot. voila! With a
dimm in just slot 1, it gave 1 short beep and booted to the bios
window. I reinstalled the dimm to slot 3 and tried again. the boot
to bios failure returned.

Now things turn more confusing. I moved the dimm from slot 3 to 2,
accepting that this would kill the dual-channel ddr feature. On
power-up, it booted to bios. I changed the bios settings back to what
I think they were, but even though I had tried to clear cmos by
removing the battery, it appeared that some [all?] of my previous
settings were still there. I saved and exited bios and reboot was
normal. I shut down, moved the dimm from slot 2 back to slot 3, to
enable dual channel, afaik. On reboot it worked.

It seems that dimm location/dual channel was the culprit in
combination with the bios setting i guess. I still don't understand
it but I did get it working again.

To start testing the mem, and cpu at more than 333, would I be advised
to reduce the multiplier and start increasing the fsb from 166?

On Thu, 14 Aug 2003 11:43:16 GMT, NewB wrote:


To start testing the mem, and cpu at more than 333, would I be advised
to reduce the multiplier and start increasing the fsb from 166?


That's what I do, under clock the CPU and then play with the bus speed
so your really only pushing the chipset and ram. I also keep my ram at
100% (sync) of the FSB.

Ed

On Fri, 15 Aug 2003 21:27:24 GMT, Paul Kilbourne
opined in
alt.comp.hardware.overclocking.amd:

On Thu, 14 Aug 2003 11:43:16 GMT, NewB wrote:

On Wed, 13 Aug 2003 19:27:46 -0500, Ed opined in
alt.comp.hardware.overclocking.amd:

On Thu, 14 Aug 2003 00:01:55 GMT, NewB wrote:

I have been testing my new ga-7n400 Pro, with AMD 2500+ and 512MB
ddr3200

[installed in slots 1 and 3]

bios v. F9

today looks like i made a booboo

or uncovered a glitch in the bios coding?

I was running spec settings:
166 clock for 333fsb
11x multiplier for 1837mhz CPU speed
[afaik, mobo and mem is rated at 400fsb and amd 2500+ 333 so I think
my settings were conservative and not over clocked at all. then...

Since my 3200ddr was running fine at 333, I decided to test it at 400.
But the way I did it I think was not so good. Maybe it exposed a flaw
in the bios design, I am not sure..

The bios was set to 166mhz and 'mem freq' was 100% which gave result
of 333 for fsb. I changed 'mem freq' to 'by spd' which changed
resulting mem clock? to 400, then clicked 'save and exit.

On exit, the
screen blanked and the puter emitted a long tone,
repeatedly. It seemed to be hanging, so i pressed the reset button.
The lights on the HD and CD blinked, but nothing else happened. I
pressed the power off button. Nothing happened, not even blinking
lights. I pressed and held the power switch for about 5 sec and the
system finally powered down. I pressed it again to boot up. Nothing
appeared on the screen, and the same long recurring beep started. p
102 of the manual indicates a continuous long beep as being a dram
error.

So i figured i'd go into bios and undo the change from mem freq '100%
to 'by SPD' but now I can't even get into bios. All I get on boot is
a blank screen, with maybe a blinking cursor.. not sure even about
that detail atm.

What now?

Do I need to clear cmos? How? I looked on the MB and do not see a
jumper to clear bios cmos. Do I just take out the battery for a
minute?

Thanks in advance for any hints and facts.

Read your mobo manual on how to clear the CMOS, usually found near the
end of the manual, it's an easy thing to do but some boards require
different steps taken.

But I think they all tell you to shut off the PSU, unplug the PC from
the wall socket if your not sure, wait a minute then proceed to clearing
it according to your mobo makers instructions.

Ed


I thought I had read the entire manual a few times, before posting,
but do admit I missed this. The manual information is ambiguous, and
not specific for this board family, much unlike the nice board
specific details elsewhere through out the manual. I found no jumper
on my board, so followed plan B and removed the cmos battery for a
while, shorted the contacts and reinstalled the battery.

This had no noticeable effect. On boot the continuous long tone was
still there and no bios boot/info screen.

On a hunch, and for no logical reason I can understand, I decided to
pull one of the dimm modules [from slot 3] and reboot. voila! With a
dimm in just slot 1, it gave 1 short beep and booted to the bios
window. I reinstalled the dimm to slot 3 and tried again. the boot
to bios failure returned.

Now things turn more confusing. I moved the dimm from slot 3 to 2,
accepting that this would kill the dual-channel ddr feature. On
power-up, it booted to bios. I changed the bios settings back to what
I think they were, but even though I had tried to clear cmos by
removing the battery, it appeared that some [all?] of my previous
settings were still there. I saved and exited bios and reboot was
normal. I shut down, moved the dimm from slot 2 back to slot 3, to
enable dual channel, afaik. On reboot it worked.

It seems that dimm location/dual channel was the culprit in
combination with the bios setting i guess. I still don't understand
it but I did get it working again.

To start testing the mem, and cpu at more than 333, would I be advised
to reduce the multiplier and start increasing the fsb from 166?

I had the same problem. Try putting the memory in slots one and two.
It is not documented, but it worked for me.


At this point, a short while after power has been restored after that
blackout, I haven't even looked to see if that system is still
working. I had left it under mem stress test when I left for work and
it was running when the blackout happened. :\

anyway.. Does it appear the breakdown might be in bios, failing if the
mem freq does not match fsb. Should a user be able to run mem at
other than fsb? or is this a common characteristic of all systems?

I can run mem in slots 1 and 3 to effect dual channel ram access, as
long as mem freq = fsb in bios. But as stated, I had to either remove
all but 1 dimms or install in only slots 1 and 2 in order to boot
after I changed mem freq != fsb.

On Sun, 17 Aug 2003 21:26:10 GMT, Paul Kilbourne
opined in
alt.comp.periphs.mainboard.gigabyte:

I also put two Sunon 80x80x38mm Tornado
Fans in the case, bringing my fan count to four. They dropped the
tempature to:
CPU - about 52C and the system to 32C under a moderate load.

The fans are load.


I think that I was having problems with heat and that may have
contributed to the problems I was having. I bought my fans at Best
Byte for a lot less than any other place I have found. No, I do not
own them nor am I related to anyone who does. I do not even no anyone
who works there. They do have great prices on Fans.

Do the fans support rpm feedback?

My ga-7N400 pro has 2 open fan ports besides the cpu fan and the mobo
supports sense feedback. I bought 2 'generic' 3 wire fans which
plugged direct to the mobo, but both bios and easytune4 do not show
any rpm sense feedback from them, only the CPU fan. I moved the CPU
fan wire to one of the alternate fan connections and the mobo did
report it running at ~3700rpm. So I know the mobo and software is
working.

Does the 3rd sense wire on a fan 'usually' indicate a feedback design?
Or is not the sense line standardized yet? How do you know if a 3
wire fan will work on my GA MoBo?