Help with P5Q-E ...

Hello,

My problem is:

I have a decent Q6600 with a 1.2250VID.
He runs 3.7Ghz around 1.384V stable the previous owner said and i can confirm this.

Either, my mobo is strange acting (or bios is some strange thing...).

When i boot 400 x 9 and all voltages on minimum (except vcore) and i use the 333 strap with 1:1 divider, it runs stable!

When i set like 415 it boots but at windows it gives me BSOD. This keeps happening even if i up the voltages off NB/PLL/VTT etc etc...
BUT, when i use the 266 strap with 4:5 divider (lowest on this strap) it boots fine till 435 FSB... The northbridge needs some more voltage becuase off the tighter strap but it's so stable as what...

None off you guys with P5Q-E got this ? I tried every bios and all off them got it...

I just want to use the 333 strap with 1:1 divider and clock to 475 x 9 for some benching... If i got to use the 266 strap i'm on a ram clock off 593Mhz that my current ram can't handle... Tuesday i get PC9600 ram and then i can get this but i want to get this even without that ram. Why wont the 333 strap with 1:1 divider work above 400 FSB ... ??? Also, 400 strap, never ever posted since i get the board. You got this to ?

Sorry for the grammatical mistakes, but you understand me anyway ;) .

Any help is welcome !

EDIT:

To make it some understandable...

FSB : 400
FSB Strap : 333
RAM : 1:1
CPU Voltage : 1.40000 v
CPU GTL : AUTO
CPU PLL : 1.52 v
FSB/VTT : 1.22 v
DRAM : 2.10 v
NB : 1.12 v
NB GTL : AUTO
SB : 1.10 v
PCIe : 1.50 v

This is 1000% stable with IntelBurnTest and Prime95...

But when i use the same settings but then 401 FSB, it boots but is far from stable.

Even this:

FSB : 401
FSB Strap : 333
RAM : 1:1
CPU Voltage : 1.425000 v
CPU GTL : AUTO
CPU PLL : 1.54 v
FSB/VTT : 1.30 v
DRAM : 2.10 v
NB : 1.20 v
NB GTL : AUTO
SB : 1.20 v
PCIe : 1.60 v

But the good news is, if i use this :


FSB : 435
FSB Strap : 266
RAM : 4:5
CPU Voltage : 1.45000 v
CPU GTL : AUTO
CPU PLL : 1.60 v
FSB/VTT : 1.34 v
DRAM : 2.10 v
NB : 1.38 v
NB GTL : AUTO
SB : 1.20 v
PCIe : 1.60 v

It's perfectly stable... And the 400 strap never ever posted and gives me also 15 min work to get the board working again...

If you run 1:1, try a PL (Transaction Booster) of 10 or 11.

Also, your PLL might be too high at 1.54, so try reducing to 1.52 or 1.50. Also I have no idea what RAM you have, it could be the SPD table in your RAM uses totally different timings at 400 than at 401+, so you may have to set some of the timings by hand...

Try kets modded bios. That should at least eliminate memory incompat issues.

Oke, i will try a tRD of 10/11.

I saw jVidia test about the PLL voltages etc... That's why i set it at 2 notches higher then AUTO.

I you set PLL on AUTO or 1.50 the board puts ~1.80 through it !
That's why. This also happens for VTT etc...

For the ram, i have OCZ Platinum PC6400 CL4, they run stable 550Mhz CL5 on 2.1V and that's it... Will try Ket's bios to. About the timings, these are all set manual, with all i mean ALL :) .

I have also one problematic E board... all is fine up to 450FSB but once above it posts but is hardly stable (sometimes 30secs, other times for hours)... I've tried several biosses and I think I got a flakey one... the other E runs daily with 500FSB... this board has been so sensitive to any setting I applied... and it prefers to run AUTO on most of the ram settings

LLC doesn't work at all on my board and just seems to enhance instability with 45nm (can't vouch for ya 65nm)... I leave the strap at auto and to loosen the ram timings to gain stability (clock twister) at high FSB... leaving me to believe that the NB caps out on me... I reseated the heatsink, reapplied the TIM... tried every possible voltage setting, different ram, PSU, CPU's... all the same result... To me a borked board...

Leeghoofd, i would say i agree with you that i maybe have a 'not so okay' board, but 400 FSB on a quad with only 1.12 vNB ? That's in mine opinion normal/good ... maybe very good... ?

update your bios ?

What settings are you guys running to get 4.1ghz or better

I have a P5Q with 4 gigs of PC8500 G skills and a E8500 with a custom water cooling kits on the way, want to know what bios and settings to get it stable over 4ghz for 24/7 :)

Quote:

---

Originally Posted by Prodigys

Leeghoofd, i would say i agree with you that i maybe have a 'not so okay' board, but 400 FSB on a quad with only 1.12 vNB ? That's in mine opinion normal/good ... maybe very good... ?

---

I fully agree but remember that these boards are foreseen to do 400FSB out of the box... so I would only expect that to be normal. Thing that bothers me is : I own 2 E's and one Dlx and 2 out of 3 do 600FSB/1200 RAM for benching while the other one doesn't even breach 450FSB, the latter is a bit out of bounce don't you think... this chipset is build for high FSB hence why we loose some ram performance over X48... and we are not here to run 400FSB... ,that was ages ago :ROTF:

Quote:

---

Originally Posted by Leeghoofd

I fully agree but remember that these boards are foreseen to do 400FSB out of the box... so I would only expect that to be normal. Thing that bothers me is : I own 2 E's and one Dlx and 2 out of 3 do 600FSB/1200 RAM for benching while the other one doesn't even breach 450FSB, the latter is a bit out of bounce don't you think... this chipset is build for high FSB hence why we loose some ram performance over X48... and we are not here to run 400FSB... ,that was ages ago :ROTF:

---

I know, but before i had the Q6600 i could do this:

http://valid.x86-secret.com/show_oc.php?id=402608

http://valid.x86-secret.com/show_oc.php?id=405584

And i see i got with a other Q6600 a good 490FSB in 5 min testing...

Didn't try Ket's biosses yet, but i'm thinking my ram is not good in QVL.
If i had Corsair Dominator PC8500 when i did that 490FSB on the Q6600.

Look : http://valid.x86-secret.com/show_oc.php?id=395591

With other words, my ram is maybe the down factor.


Trying tomorrow:

- Other ram
- Ket's bios
- tRD of 10/11

BTW, LLC works just normal here, so my mobo is not the down factor i think.
But, more news tomorrow !

Quote:

---

Originally Posted by Prodigys

I know, but before i had the Q6600 i could do this:

http://valid.x86-secret.com/show_oc.php?id=402608

http://valid.x86-secret.com/show_oc.php?id=405584

And i see i got with a other Q6600 a good 490FSB in 5 min testing...

Didn't try Ket's biosses yet, but i'm thinking my ram is not good in QVL.
If i had Corsair Dominator PC8500 when i did that 490FSB on the Q6600.

Look : http://valid.x86-secret.com/show_oc.php?id=395591

With other words, my ram is maybe the down factor.


Trying tomorrow:

- Other ram
- Ket's bios
- tRD of 10/11

BTW, LLC works just normal here, so my mobo is not the down factor i think.
But, more news tomorrow !

---

Disable LLC because you don't need it...

Quote:

---

Originally Posted by ahmad

Disable LLC because you don't need it...

---

Can you explain that ?

I notice now that the idle voltage from 1.376v jumps to 1.400V load with LLC enabled... This on 400 x 9 setting...

Quote:

---

Originally Posted by Prodigys

Can you explain that ?

I notice now that the idle voltage from 1.376v jumps to 1.400V load with LLC enabled... This on 400 x 9 setting...

---

I don't know exactly how it works it just tries to eliminate vdrop and vdroop, but there is no need for it because:

if you want for example 1.3v, then just set to 1.35v in the bios. If you want 1.376, just set it to 1.400v in the bios, and so on. On this board, I haven't seen any benefits to using it.

FWIW: I have done some research and will no longer use LLC myself. Anandtech and some other sources have done some testing that made me wary of it :eek:

Seems fine for some folks, but I'll keep it disabled for now.

I disabled mine but the vdroop is horrible. It's enabled again now, but I seem to have no issues with it enabled or disabled, apart from the horrible vdroop, of course.

I wont have LLC enabled, 1.4V in bios, about 1.32V load... worst Vdroop of any mobo I have, glad its got LLC.

Quote:

---

Originally Posted by Willhub

I wont have LLC enabled, 1.4V in bios, about 1.32V load... worst Vdroop of any mobo I have, glad its got LLC.

---

Yeah but what is the problem with setting it to 1.4 in the bios??

Nothing, but I'd need 1.50V in bios though for 1.42V.

Okay, just tested 415 x 9 with tRD 10 and the OCZ ram, started up IBT and after 30 sec it suddenly restarted (same problem i got before).

Next i'm going to try other ram (G.SKill PC6400), and if that's no go, i go for Ket's bios.

About the LLC, i have it enabled, got me 1.4000 in bios, gives me 1.376 idle and 1.4000 load in windows.

With other words, what i set in the bios, i get in windows under load. But what's the problem with LLC ?
I mean, if i disable it, and i set 1.4000 in bios, i maybe only get 1.376 idle but a lousy 1.328 load... Why not enable i then ? Gives it extra instability ?

Why is Vdroop so bad on this mobo without LLC? Its got 8 phases sure that must be enough?

I found this mobo more picky now, at anything under PL10 it wont boot when 400fsb strap is set and 1:1 ram, disabling the fsb strap, well setting it to Auto and running ram at 5:6 I now can boot at PL8.

Quote:

---

Originally Posted by Willhub

Why is Vdroop so bad on this mobo without LLC? Its got 8 phases sure that must be enough?

I found this mobo more picky now, at anything under PL10 it wont boot when 400fsb strap is set and 1:1 ram, disabling the fsb strap, well setting it to Auto and running ram at 5:6 I now can boot at PL8.

---

That's some intresting thing you say there.
I always had PL on auto setting, and since i had the board it never ever posted on 400 strap.
But now you say, it wont boot when you set 400 strap and PL lower then 10.
Maybe that's the reason my board never posted, becuase it sets a to low PL when auto with 400 strap.
going to try that to then !

Set 400 strap with PL11-12... For 3DMark it doesnt matter :) .

EDIT:

Strap 400 - PL12 = no boot...

Now, i did test this :

FSB : 415
FSB Strap : 333
RAM : 1:1
CPU Voltage : 1.43750 v
CPU GTL : AUTO
CPU PLL : 1.52 v
FSB/VTT : 1.38 v
DRAM : 2.10 v
NB : 1.22 v
NB GTL : AUTO
SB : 1.10 v
PCIe : 1.50 v

And it runs 3DMark fine.
So, i go to bios and try 425 x 9, it doesnt boot... BUT, when i boot 415 x 9 (like above) and use SetFSB, it is stable untill 432 x 9 and then it begins to act strange...
So, with the settings from the 415 x 9 boot i can SetFSB to 432 x 9, but with the same setting i can't boot with just 425 x9...
I tested this with 2 x OCZ and 1 G.Skill stick on the board. Even tried only the OCZ, same thing. With the G.Skill i got the same.

Now, going to try Ket's bios, and if that doesn't work, i dont know it anymore.
The only thing i can do then is use the 266 strap for 415 + FSB... Wich is no problem, but man, the NB needs allot voltage then ...

Got mBIOS 1306 P5Q-E on the board, noticed now after 3 min that only 3 cores work off my Q6600...

Trying to reach Ket on MSN, maybe he forgot something.

If i got some time, going to test the 333 strap with +415 FSB...

EDIT:

Reboot gives me back 4 cores... !

EDIT2:

Just played with some settings, and 333 strap is unstable as soon as i hit 3.8Ghz...
The 266 strap works good, but needs some crazy voltages :p: ...

So, if somebody know's something to try, just shoot!

I tried Keys mBIOS 1206 and I see no difference in the BIOS.

Also.

At the moment with my E8500 I am running prime blend, 4.23Ghz, 1.44V, hitting 75-76 degrees but now and then goes lower, this is going to be my 24/7 overclock I think, I might just wack it to 4050Mhz and run it at 1.35V though and say no more.

In bios my current settings:

FSB : 470
FSB Strap : 400
RAM : 1:1
CPU Voltage : 1.5000 v
CPU GTL : 0.64/0.64
CPU PLL : 1.52 v
FSB/VTT : 1.30 v
DRAM : 2.04 v
NB : 1.30 v
NB GTL : AUTO
SB : AUTO v
PCIe : AUTO v

Other settings:
5-5-5-5-15 on ram, the other timing that people say manually set is at 50.

DRAM Static Control:Disabled
DRAM Read Training:Disabled
MEM. OC Charger:Enabled
AI Clock:Lighter

PL setting to 10, dont want to try 9 with fsb unstrapped as anything but 1:1 was too high.

Does SB and PCI-E overvolted give more stability?

When you tried without FSB strap did you have 1:1 still? Apparently the P45 is very sensitive to 1:1 and when I tried it I noticed settings it to 5:6 or something made it boot so it might not be the fsb strap but the 1:1 that makes it not boot under PL10.

Mine does 400FSB semi-stable, but craps out completely @ 412FSB. That could be my chip though.

You cant run 1:1 with pl less than 10. 1:1 I use pl 11, divider 333 strap I use 8. 400 strap broken with the bios I am using.

This is a fun board to play with, but the pll/vtt voltage readouts hard to reach with a tru installed.

Quote:

---

Originally Posted by Prodigys

Why not enable i then ? Gives it extra instability ?

---

Yup thats what it does for me. Creates unstability and I don't need any more factors when I am ocing..

Odd, dont create instability here, well I dont want to know really if it does on my mobo if I enabled it that means 1.48V idle for 1.43 load..

Willhub if you enable LLC you get 1.48V idle and 1.43 load ? don't get that...

Quote:

---

Originally Posted by Leeghoofd

Willhub if you enable LLC you get 1.48V idle and 1.43 load ? don't get that...

---

1.5V set in bios, gives 1.48V idle, 1.43V load, Vdroop I think it called.

So LLC is at AUTO or disabled ? I didn't get ya first phrase at all : "Odd, dont create instability here, well I dont want to know really if it does on my mobo if I enabled it that means 1.48V idle for 1.43 load.. read it 5 times and still not sure if it was on or not... apparently it's not...

Vdrop:
When you select a Voltage setting from the BIOS, and the real voltage applied is less than the selected voltage

Vdroop:
If the voltage flunctuates from idle to load, then the difference between the idle & load voltage value is named after Vdroop

Quote:

---

Originally Posted by Willhub

When you tried without FSB strap did you have 1:1 still? Apparently the P45 is very sensitive to 1:1 and when I tried it I noticed settings it to 5:6 or something made it boot so it might not be the fsb strap but the 1:1 that makes it not boot under PL10.

---

That's intresting to say, becuase if you think, with the 266 strap i'm direct on the 4:5 divider, this it the lowest setting, and i'm back very stable.
Maybe the 333 strap with a other divider then 1:1 is good to boot above 420 FSB becuase this fails now with 333 strap and 1:1 divider.

Quote:

---

Originally Posted by ahmad

You cant run 1:1 with pl less than 10. 1:1 I use pl 11, divider 333 strap I use 8. 400 strap broken with the bios I am using.

This is a fun board to play with, but the pll/vtt voltage readouts hard to reach with a tru installed.

---

Same here, good to hear that 1:1 never boots under PL10.
These are things you only know from testing, with some ppl testing it goes much better :cool: .

Quote:

---

Originally Posted by ahmad

Yup thats what it does for me. Creates unstability and I don't need any more factors when I am ocing..

---

Well, LLC isn't any problem here, enabled or disabled, no difference :clap: .

EDIT: Gives no difference in stability. For Vcore it helps much ...

Quote:

---

Originally Posted by Leeghoofd

So LLC is at AUTO or disabled ? I didn't get ya first phrase at all : "Odd, dont create instability here, well I dont want to know really if it does on my mobo if I enabled it that means 1.48V idle for 1.43 load.. read it 5 times and still not sure if it was on or not... apparently it's not...

Vdrop:
When you select a Voltage setting from the BIOS, and the real voltage applied is less than the selected voltage

Vdroop:
If the voltage flunctuates from idle to load, then the difference between the idle & load voltage value is named after Vdroop

---

Hi.

I just tried LLC off now.

1.46V in bios is 1.42V idle and 1.4V load.

My E8500 must be the crappest ever, at 1.4 4.2Ghz the second core failed after 1 min, the first is still going now!!

Well guys, maybe i got tomorrow a P5E3 with some OCZ DD3 ram, but going to try some more before i switch.