Cold Boot? Memory Died? [email protected] = who you need to contact

[pershoot]

its very wierd. it doesnt happen for all of us. i only use high voltage ram on this board. board has been set to 3.4 since i recieved it in late feb. both orange and yellow slots. only twice have i had a cold boot situation(it was in orange). but in those 2 times, power off and on once, and it booted right up. this is with ac power being out and in for extended periods.

 

these days, i have noticed that once powering on, it takes about 2 seconds for signal to come to monitor, and then POST has already happened. i have also noticed these days that after cold booting, itll POST, but once it goes to load the OS, black screen will happen (and no OS loading (on the first time the rig is powered up after cold boot)). power off and on, and this is cured. but the board always POSTS (*knock on wood*).

 

i have UTT BH coming, so I will observe and let you all know if I too experience this issue. my friend uses UTT BH, and he never gets it.

 

however, this issue happens with classic BH (from what i have seen) as well as the newer UTT CH/BH modules.

 

there must be something up with these particular boards where it happens, within certain configurations.

[OzSnoal]

The point here is that the BIOS first looks for the RAM and then if validated adjusts Vdimm.

So if it cannot find the ram, it cannot adjust the voltage - you're in a catch22 situation.

 

So to avoid this, the BIOS should always set VDimmto what you applied in the CMOS and thereafter try to detect the RAM. Not the other way around.

 

Thus, back to the simulation, by removing the RAM you are proving that the BIOS will not adjust the voltage unless it finds the RAM. When using RAM like the OCZ VX which requires 3.5V and coupled with the DFI NF4 series which was designed for and supports such high voltages on Vdimm it's a shame that this is occuring.

 

DFI needs to change one of two things in the bios.

1) Either set the voltage on Vdimm before trying to detect the memory.

2) Fixup the RAM tables/FSB startup to get the VX ram to startup at 2.7V.

 

Again, to prove all the above - you must remove AC power.

[h_Us3r]

Great post Oz, I had a sneaky feeling about how power is distibuted in this board and your findings semi-support a hunch I have. I personally feel the problem lies in the BIOS chip.

BIOS chip goes==instant RMA;

 

h

[pershoot]

oz, i get no cold boot issues when ac plug is out for extended periods of time (12+ hours), with 3.4v set in bios. only those 2 times (like i stated above) while the ram was in orange, on my previous winchester.

 

its all very interesting how it happens for some and not for others...

[OzSnoal]

Seems like the 0.2V may be what makes the difference. I have to run my VX at 3.5V, and when measured with the multimeter is at 3.6V.

 

Could someone please recommend a free something or other where I can upload my pics and the AVI.

 

I have taken a couple of pics and video:

The pics are especially for ExRoadie - doubting Thomas (Just teasing, I understand all your questioning - I'm the same way).

 

I took pics of how Vdimm is set without RAM present :-)

1st I shutdown, did not unplug AC, pulled out the RAM:

Pic shows 0V on Vdimm

 

2nd I powered up WITHOUT RAM

Pic shows Vdimm at 3.6V (Proves that Vdimm is remembered even without the RAM)

 

3rd Pulled the AC Plug, and reinsert it, WITHOUT RAM, PC is OFF

Pic shows Vdimm at 2.7V

 

4th I powered up WITHOUT RAM

Pic shows Vdimm at 2.7V (Proves that Vdimm has been reset to standard and does not get adjusted up to 3.6V - because the BIOS has not detected the RAM and thus not adjusted the voltage)

 

5th Shutdown the PC, pulled AC, put in 1 stick of VX in slot 2

Made an AVI -

of how Vdimm shows 0V without AC power (duh)

of how Vdimm rises to 2.7V when plugging in AC power

of how I turn on the PC and how long it takes for the RAM to reach the voltage of 3.6V

[red930]

OzSnoal has done a fine job of working this problem. Give this thread a chance to work its way through the system to see if there is a resolution.

 

That's what we like to see. Facts, screen-shots and common sense are the only things that matter.

[Wingz]

Could someone please recommend a free something or other where I can upload my pics and the AVI.

 

Use http://www.imageshack.us for pics and maybe http://rapidshare.de/ for vids.

[OzSnoal]

EDIT testing image stuff

 

I took pics of how Vdimm is set without RAM present :-)

1st I shutdown, did not unplug AC, (EDIT, above I said I pulled the RAM, pic shows otherwise)

http://img122.echo.cx/img122/5976/0vshutdo...cremoval2bp.jpg

 

2nd I powered up WITHOUT RAM

Pic shows Vdimm at 3.6V (Proves that Vdimm is remembered even without the RAM)

http://img122.echo.cx/img122/3076/3v6nodim...cremoval6mm.jpg

 

 

3rd Pulled the AC Plug, and reinsert it, WITHOUT RAM, PC is OFF

http://img122.echo.cx/img122/7387/2v7pcoff...vereinsert9.jpg

 

4th I powered up WITHOUT RAM

(Proves that Vdimm has been reset to standard and does not get adjusted up to 3.6V - because the BIOS has not detected the RAM and thus not adjusted the voltage)

http://img122.echo.cx/img122/4844/2v6nodim...acremovedre.jpg

 

 

5th Movie showing various Vdimm stages with 1 stick, PC shutdown, AC off, AC on, Power ON

Movie Download

(Note the above is a free service, waited 60 seconds in a queue, thereafter, download maxed out my 2mbit connection)

The MOV file all in one smooth shot showing single stick bootup:

[Sharp]

Hello,

 

I tested the voltages and here are my results,

Set BIOS to 2.8v.

 

In windows = 2.8V

Shutdown PC = 0V

Power PC = instantly goes to 2.8V

 

Now the other method.

 

In windows = 2.8V

shutdown PC = 0V

Remove power plug = 0V

Put power plug back = 2.5V

Power PC = 2.5V

 

Now at this point only when the BIOS does that beep.

(you know when you turn the PC on it always does 1 beep)

1 beep = goes to 2.8V.

 

 

Things that I noticed.

The orange LEDs on the board.

When the PC is on they are lit.

If you shutdown the PC in windows the lights go out.

 

If you remove the power plug and put it back in the LEDs come on but the PC is still off.

And LEDs on = 2.5V

 

My board went to 2.5V instead of 2.7V.

My board rev = A02, BIOS = 125.

[h_Us3r]

This could be a dangerous thing for anybody runnin@ 3.2v+...

[fhertlein]

I just changed some hardware and encountered this issue. I have the OCZ 4000VX with vdimm set to 3.3, and I encountered only one other time doing the same thing.

 

I like and agree with the analysis.

[red930]

Is there a reason you haven't updated to a more recent BIOS? 1/25 has some know issues!

BIOS = 125.