ASUS Maximus Formula SE

[Nuckin_Futs]

I was able until the other day to run 450x8 ram at 1199mhz
now I cant seem to post beyond 425

Would like to get back to 450x8 for the bandwidth

How much tweaking to rest of the mobo and how much vdim to keep stable?
Can it sync 1:1 to 475~500 for DDR950~1000?

Mine did 500x8 but needed PLL of 1.742v and FSBT of 1.504v and 63x/67x on CPU/NB VTT and of corse 1.58v vcor

[Grnfinger]

What utility did you use to get those voltages? It seems like that I cant choose the Dram voltage in OCCT :S

You need Everest.

enable shared memory in preferences.

Then in OCCT load the vdim as a sensor, make sure to load Everest in the software monitoring drop down

[jonny_ftm]

what gave it away was in 0901 and before there was only things like [9] or [6]

now there is [09.0] and [06.0] i can't pick the 0.5 ones yet, so it has to be under testing.

It will never come. Multiplier is now a built in CPU function for Intel CPUs, unlike AMD. So, you'll need a Penryn to enable 1/2 multiplier steps. Do remember, penryn have high FSB and low multiplier, so unless you get a highly expensive QX unlocked edition, you'll be stuck to very low overclock performance because of the motherboard limiting any FSB increase

Owki I´ll give it a try.
I also downloaded afudos 2.11, 2.17, 2.19, 2.21 to check if any of those work.

AFUDOS v2.07 works fine on my P5B deluxe, but some new revisions of ASUS bioses are non downgradable whatever program you use. You could get CMOS checksum error and be obliged to upgrade again. Just give it a try

hey Safan, Do you know what Error Code 4 means? or does anybody know what it means

And Can someone get a Link to download OCCT Pro if anybody has it.

I saw many theories on error codes, so please, don't throw anything you didn't verify:

Code 1 : thread stopped because of another thread stopped
Code 2 : stop by user
Code 3 : memory error
Code 4 : calculation error (could be CPU, RAM, NB...)

The error code display is still wrong in the beta. If you get a code 1, you have to check the files thread0x.txt in OCCTPT folder to see what code it hides (usually a 3 or 4)

Safan, I got a question, My OCCT (Beta) doesn't break 75% on the test. No matter what. Whats up with that?

PS. No matter what Clock speed

Just a display bug in OCCTPT beta, a known bug that will be corrected. Just ignore it, it is a cosmetic error
Last beta will be the final with some display corrections, enhancements, and maybe a more detailed graph (xls?) if the author has the time

Prime small FFT most always pegs Task Manager @ 100%. I'd rather run OCCT than use Prime Blend--faster results and imho, a better indicator overall of stability.

Just read through this thread or the Gigabyte X38 thread to see you're fully wrong. OCCT in RAM mode is the ultimate test for the NB - RAM stability at high FSB, especially on a quadcore

I left the MB in the frig for about 3 hours and tried again. Just as you said, the NB and SB HS popped right off with no real force.

My bad, please, don't throw it to beginners, condensation could kill the board

I do believe that code 4 is a OCCT error

codes 1,2,3,6: calculation error, bad for you
codes 4,5: OCCT memory management error (bad for OCCT)
codes 7 and 7+: OCCT bugs

Wrong, don't belive, be sure...
read me few lines above in this same post

[Gunslinger]

Here is the response I have from ASUS for my problem:
Dear Sir/Madam


Thank you for your support of our products!

Your problem seems to be caused by an electrical stoppage of some kind. Check that your MB is not picking up EMI (Electromagnetic Interference), or improperly grounding against the case. Please remove the MB from the case and set it up DIRECTLY on the cardboard box it came in or some other non-static, non-conductive surface, like a telephone book. Install the power supply, processor (with heat sink and fan), memory and video card. If the motherboard has the vocal POST feature activated, plug speakers into the onboard audio jacks. If the board doesn't have this feature, or if it's deactivated, make sure to install a case speaker on the board.

If this eliminates the problem, then your problem has been caused by grounding. You will need to electrically isolate the MB from the case in order to correct it. To do this, you will need to use electrical tape to cover the brass stand-offs that support the MB, and insert paper washers between the MB and the heads of the mounting screws. The red paper washers should be included with your case hardware, or they can be purchased at Radio Shack or a computer store.
Make sure that no stand-offs are placed in a spot where the motherboard isn't designed to be grounded, as some cases will have different standoff locations to accommodate different styles of motherboards. Also check that there are no other metal objects that could contact the motherboard or any other electrical device attached to the system, such as a metal burr, loose screw, metal rod, or any other object that could cause a short.

Don't know what to do now, the hardware is not in a case, but set up on my wooden printer stand.

[Grnfinger]

How much tweaking to rest of the mobo and how much vdim to keep stable?
Can it sync 1:1 to 475~500 for DDR950~1000?

Mine did 500x8 but needed PLL of 1.742v and FSBT of 1.504v and 63x/67x on CPU/NB VTT and of corse 1.58v vcor

I set ( currently testing ) 450x8
FSB Strap to 333
Dram 1081
CPU Volts 1.55
Dram 2.20
NB 1.55
SB 1.10

This has passed 2 hours prime95 when b4 it wouldntmake it past 7 mins
Going to try OCCT 30min test now, if it passes I'll let her run all night

[Grnfinger]

Quote:
Originally Posted by Grnfinger
I do believe that code 4 is a OCCT error

codes 1,2,3,6: calculation error, bad for you
codes 4,5: OCCT memory management error (bad for OCCT)
codes 7 and 7+: OCCT bugs

Wrong, don't belive, be sure...
read me few lines above in this same post
__________________


My info came from Toms Hardware
I figured it wold be correct

My bad

[Unbornchild]

Gunslinger: i'd take the Logitech Speaker control device off the PSU.

And if ASUS is right about 'improper grounding': of course you're using a grounded wallsocket, but have you checked if you get the same result from another socket (to check if the grounding is ok)?

[NysoO]

You need Everest.

enable shared memory in preferences.

Then in OCCT load the vdim as a sensor, make sure to load Everest in the software monitoring drop down

Thanks!
Another question: In bios you can set the LCD poster in a "User string" mode. Does this mean that you can make the poster show like your nick? but I can't seem to find anywhere in which you can change what the poster will show. Cause' I aint lookin' at the clock on the poster, and it would be a little bit more fun if you could make it post whatever you want

[Gunslinger]

The PSU was moved under the stand shortly after the picture, the Logitech system is has not been hooked up yet. Will look for extension cord to check another socket.

Update: Moved to another outlet, same result.

[Talonman]

Just install all modules it will work fine.

Thanks for answering my question...

Both sticks (2 x 2GB) will be installed in the mobo, before my Vista Ultimate x64 OS install.

I guess i'll take some of the users advice here, and use the white slots... ??

[Grnfinger]

Thanks!
Another question: In bios you can set the LCD poster in a "User string" mode. Does this mean that you can make the poster show like your nick? but I can't seem to find anywhere in which you can change what the poster will show. Cause' I aint lookin' at the clock on the poster, and it would be a little bit more fun if you could make it post whatever you want

I am still trying to find how to enter a "user string"
I have a few thoughts for the display

[Cumulonimbus]

anyone still has the temp problem?

Yes, since 0602 the CPU-Temperatur increase from 29 to 46°C. Now PC Probe report's 46°C CPU Temp, but the Cores 0 to 3 are only 33 to 36°C :-)

If i flash 0602, PC Probe reports the correct CPU-Temp.

[Grnfinger]

Stable again at 450x8!

The insane ram voltage ripple is gone, I can live with this





Thanks guys for the tips and advice

[Cumulonimbus]

Stable again at 450x8!

The insane ram voltage ripple is gone, I can live with this

Stable, after 30 minutes? :-) Run OCCT for 4h+, than you can say it's stable not earlyer. :-)

[kup]

Stable again at 450x8!

The insane ram voltage ripple is gone, I can live with this





Thanks guys for the tips and advice

How'd you sort it out?

[Grnfinger]

Stable, after 30 minutes? :-) Run OCCT for 4h+, than you can say it's stable not earlyer. :-)

Generally I run it 24 hours
this is a settings test to make sure I'm on the right path

I have been fighting with 450x8 and a nasty ram volt ripple ever since I tested bios 902

[NysoO]

I am still trying to find how to enter a "user string"
I have a few thoughts for the display

Cool, please make us notified if you find out how to do it.

[Grnfinger]

How'd you sort it out?

I did a test today

moved the ram from blue slots to white slots, and stopped over volting my ram from 2.2 to 2.0 ( rated for 2.1 ) Set transaction booster from disabled to Auto.
Changed
DDR2 Channel A REF Voltage : DDR2_REF
DDR2 Channel B REF Voltage : DDR2_REF
DDR2 Controller REF Voltage : DDR2_REF
to all AUTO

[kup]

And which of those changes made the difference or did you do them all at the same time?

[Grnfinger]

I did things in kinda of a logical order

moved ram to white slots, little difference
set transaction booster to auto and it would load windows without bsod
changed ddrII ref to auto , ripple was there but not as bad
drop the volts back and ripple was almost gone to what I posted

[CERO]

hmz so putting to much volts @ your ram would cause ripple?

[kup]

Well I've just gone and flashed 0903 again and taken your pointers into consideration when loading up my 24/7 Profile. Lets see how it fares now.

Last time I flashed 0903 I couldn't keep the board stable @ these clocks, but hopefully this time it's going to play a little nicer for me? :P

[simonmaltby]

Howdy guys... new to this thread and boy does it take some reading

Just changed by Bitz Formula for this one and finding my way.

All instabillity i have found so far when comparing the Bitz to the Maximus seem to relate to memory. Some dividers just painly dont work.

I have not tried beta bioses yet... just the 802 official one.

I was having terrible stability until i set all my ram settings manually... Its fairly pointless me listing them all unless you have the same ram as me.

One thing that strikes a cord with the previous post is the DDR2 REF setings. I have all of mine on DDR2 REF except the controller with i set on +10 (this infor i found re the Striker on the Mushkin forums). Anyway since that and the ram times i have better ram performance and stability than i has on the Blitz.

Still working through the other bits... but thanks to all for the pointers so far.

I'm stable at 400x9 with ram at 1066

[Talonman]

Are you running on the white or blue memory slots?

[pazza316]

I just flashed to the 0903 bios and found things to be more stable at 1:1 Ratio. My fsb can clock much much further now also.