"ASUS P5Q Deluxe discussion".

[Martin.v.r]

no cooling at,NB/SB ..... fan on mem

[CryptiK]

Ive benched on 1.48vNB it doesn't exceed 47*C. Stock cooling on these boards is pretty good.

[carepolice]

Ive benched on 1.48vNB it doesn't exceed 47*C. Stock cooling on these boards is pretty good.

Am i right, does -E and Deluxe have same NB cooling ? Atleast it looks like perfectly same

[pangingIII]

475 FSB, 333 strap, 5:6, 8GB RAM at DDR2-1140, tRD 8 / vMCH 1.40, vDimm 2.0v









448 FSB, 266 strap, 4:5, 8GB RAM at DDR2-1120, tRD 7 / vMCH 1.44, vDimm 2.0v








from past experience with this board, in order for me to set tRD 7 when system overcloked above 400fsb, this used to be possible only to use 400 strap (e.g. 400fsb / DDR1066, 450fsb / DDR1200).
also, my gskill 8GB kits (2 kits of 2x2GB) have never overclocked well on this board, the ram kits are not a very good overclocking ram by themself, however, the ram kits can now be overclocked to DDR2-1140 (and not on a low fsb), this used to be max stable at around DDR2-1128 for 8GB.

[mac_69]

[QUOTE=CryptiK;3513244]Not too bad. Have you tried settings CPU GTL Refs to 0.67x or higher? There's a possibility that you could go a little higher or lower your cvore at the current speed with some GTL tweaking.


I'm just happy to be stable; the system is far more powerful than my needs anyway. Fast, quiet and able to game with insane FPS when needed. I enjoyed the learning process but I'd be quite happy never to sit doing prime95 runs again Can see this system lasting a few years for me.

[CryptiK]

Am i right, does -E and Deluxe have same NB cooling ? Atleast it looks like perfectly same

From pics it looks to be the same, except the P5Q-E has no heatpipe from NB to SB, and I've read that it has no backplates on the Mosfet sinks as it has 8 phase PWM and no fets on the backside of the board.

475 FSB, 333 strap, 5:6, 8GB RAM at DDR2-1140, tRD 8 / vMCH 1.40, vDimm 2.0v


448 FSB, 266 strap, 4:5, 8GB RAM at DDR2-1120, tRD 7 / vMCH 1.44, vDimm 2.0v


from past experience with this board, in order for me to set tRD 7 when system overcloked above 400fsb, this used to be possible only to use 400 strap (e.g. 400fsb / DDR1066, 450fsb / DDR1200).
also, my gskill 8GB kits (2 kits of 2x2GB) have never overclocked well on this board, the ram kits are not a very good overclocking ram by themself, however, the ram kits can now be overclocked to DDR2-1140 (and not on a low fsb), this used to be max stable at around DDR2-1128 for 8GB.

That's really nice to see that FSB with a quad and 8GB of ram at that speed stable. What is Ai Clock Twister set to and are you running static read control and dram read training enabled? They gave me a nice boost in performance, as did running clock twister on stronger. Not sure if you could manage that with a quad + 8GB at those speeds though.

I'm just happy to be stable; the system is far more powerful than my needs anyway. Fast, quiet and able to game with insane FPS when needed. I enjoyed the learning process but I'd be quite happy never to sit doing prime95 runs again Can see this system lasting a few years for me.

Good to see a wholly satisfied person, your setup is nice.

[pangingIII]

That's really nice to see that FSB with a quad and 8GB of ram at that speed stable. What is Ai Clock Twister set to and are you running static read control and dram read training enabled? They gave me a nice boost in performance, as did running clock twister on stronger. Not sure if you could manage that with a quad + 8GB at those speeds though.

opp sorry i haven't answered any of the questions...

static read control - enabled (in some of the previous bios versions i set this option on auto. but as far as i remember i never disabled this option completely)
dram read training - enabled
Ai clock twister - for my hardwares even before i use this ram kits, this setting is always set on moderated. however, i believe depending on hardwares quite a bit, i can't set Stronger when using this 8GB ram and before, and maybe include the voltage setting that also have effect on the setting. i also used to find some cool booting problems with Strong mode

the CPU batch that i got is a very hot CPU on full load, it's the very first retail q9300's that were available in retail store before q9x50, it was some time back in around Mar-April.
my max daily use for this particular CPU have been 467 at its default multi 7.5 and DDR1120, this is for me to keep a low voltage use in system, include RAM,
but 480FSB still can be stable on this p5q-d board. i might rebuild the cooling system if i'm not too lazy and start to use 475FSB for daily setting.

[CryptiK]

Nice work pangingIII, 480 FSB / DDR2-1152MHz PL8 is very nice What's the max load temp at those settings? What vNB did this take to stabilize?

I decided to re-do the TIM (thermal interface material) under the NB & SB heatsinks of my board. It was very easy to remove and replace; after undoing the 2 bolts on the rearmost mosfet heatsink and popping out the pushpins from the NB & SB sinks, I simply gently lifted the SB heatsink, the heatsink seperated from the SB chip and the whole heatpipe assembly subsequently lifted off easily. The stock TIM is a sticky gum substance, and is quite thick, and most likely would not have optimal thermal transfer properties. After removal, I initially cleaned the stock TIM off the NB, SB and associated heatsinks with my fingernail (it just peels off easily), and finished the cleaning process with isopropyl alcohol. I re-did the TIM using arctic ceramique. I also had to slightly twist the SB heatsink so it was perfectly flat on the SB when mounted, as the bottom side of it was inititally sitting slightly higher than the rest.

Pics:

NB chip as it looked after I removed the heatpipe assembly



SB as it looked when I had removed the stock TIM with my finger nail



NB Heatsink as it looked after I removed the heatpipe assembly 1



NB Heatsink as it looked after I removed the heatpipe assembly 2 - you can really see how thick and sticky the stock TIM looks



NB chip cleaned



SB chip cleaned



NB heatsink cleaned



SB heatsink cleaned



Mosfet heatsink & thermal pad - notice the excellent contact between all mosfets and the heatsink. This is due to the use of the backing plate under the board, which prevents the board from bowing. My M2F does not have a backing plate, and the centre mosfets don't have as good a contact as those on the ends.

[Gir92]

Can anyone give me advice for CPU and NB GTL settings?

I'm using 1.6v VTT on my Q6600 @4Ghz.

Thanks.

[lexus232]

Cryptik, thanks for photos and responses!

[CryptiK]

Can anyone give me advice for CPU and NB GTL settings?

I'm using 1.6v VTT on my Q6600 @4Ghz.

Thanks.

Q6600's seem to like ~0.65x, but GTL Ref settings are trial and error and must be worked out by each person Try between 0.60x and 0.70x (0.70x is almost the absolute max and gives you a Vref of ~1.12v due to your Vtt of 1.6v) for both GTL Refs. Set a clock that you know is unstable, run Orthos small FFT and note the fail time. Then adjust the CPU GTL Refs and see if you can increase the fail time. The point at which it takes the longest to fail, is teh sweet spot.

NB GTL Ref seems fine at auto up to 550 - 560 FSB for me, but for your board it may be different as each board's MCH has slightly different behaviour, as CPU's do. Drop the NB to a level that makes the chipset produce errors (I like to use HCI Memtest in windows for this but before you do you must be certain your ram is stable or it will produce errors too) and tweak the NB GTL Ref. Again, the setting the decreases the errors generated or decreases the time before the first error is generated is the sweet spot.

Cryptik, thanks for photos and responses!

No problem at all.

[Gir92]

Thanks Cryptic. Also, what's the max NB volts you recommend? I'm using 1.4v right now. North bridge is watercooled as well, so temps aren't an issue.

[CryptiK]

For 24/7 use I'd have no problem running 1.4v on air keeping max temps under 50*C, even on water I'd say around 1.45v is around max for 24/7, 1.5v would worry me a little. Remember the P45 chipset is 65nm, and 1.5v is pretty much max for 65nm cpu's 24/7, and intel rate the P45 at ~1.2v absolute max, so even at 1.4v it's already over that by a bit- this is why it's hard to advise anyone on a safe voltage as most of us are already over the rated absolute max. Just depends on how hard you want to push it really.

[lexus232]

For 24/7 use I'd have no problem running 1.4v on air keeping max temps under 50*C, even on water I'd say around 1.45v is around max for 24/7, 1.5v would worry me a little. Remember the P45 chipset is 65nm, and 1.5v is pretty much max for 65nm cpu's 24/7, and intel rate the P45 at ~1.2v absolute max, so even at 1.4v it's already over that by a bit- this is why it's hard to advise anyone on a safe voltage as most of us are already over the rated absolute max. Just depends on how hard you want to push it really.

what NB volts u are using with 4005mhz 1.336v? I now using 1.20 at 4005mhz 1.256v.

[RAdoxx]

Lately I've been getting bluescreens caused by the AHCI-driver, iaStor.sys, and I can't figure out why I'm getting them.

It usually happens while I play Warcraft 3, I've never had it happen while just using Firefox etc. I don't know if it has something to do with the latest AHCI-drivers (8.6.0.1007) or if it's caused by something else, like a failing harddrive.

Although it seems like I'm getting them pretty random, sometimes I can play for hours, and sometimes I get them just within like an hour or so of playing.

I usually have uTorrent, Firefox, mIRC, Live Messenger and Nod32 running in the background.

I'm running the system in my sig at stock settings, but with set voltages because of the overvolting on this board.

[Melhisedek]

Anything I should think of when using SATA ports? I have 2 SATA HD and one IDE DVD burner. I don't use RAID. I just plugged SATA drives into 2 top slots. Now I'm experiencing some really strange freezes. They are really random, last exact 5 minutes and I mean EXACT. I can move the mouse but not much else. Sometimes I can minimize current window. Afterwards there is nothing strange in Event Viewer, but under Windows-Diskcheck-performance (or something similar can't remember now am at work) I can see that some service have been using a lot of HD and thus lowered performance. I've checked most of my freezes and it can be anything that causes disk trashing. From Firefox, Warcraft, Explorer to some ntfskrnl or something.

I've tried a lot of stuff... PC is rock stable no matter the stresstest. Prime, Orthos, Pi, Memtest. It can run them all for at least 24 hours. But I still get these weird freezes. After 5 minutes everything is just fine and dandy...

Any ideas?

[x3mgamer]

same problem with my seagate 640 gb, made the problem worse when switched to ahci native sata mode

[RAdoxx]

Did you install some IDE controller driver? I remember back when I had my P5B Deluxe, the system would go crazy if I had the IDE driver from the CD installed. Jmicron something IIRC.

[CryptiK]

what NB volts u are using with 4005mhz 1.336v? I now using 1.20 at 4005mhz 1.256v.

I run PL8, static read control and dram read training enabled, and Ai Clock twister = stronger, and it takes ~1.30v for stability using my Max 2 Formula. I think my deluxe doesn't need quite that much due to the fact it appears to have slightly looser chipset timings clock for clock at the same settings compared to the M2F, but I have just been benching on it recently not using it 24/7.

[x3mgamer]

Did you install some IDE controller driver? I remember back when I had my P5B Deluxe, the system would go crazy if I had the IDE driver from the CD installed. Jmicron something IIRC.

yeah i think it was installed when i made the switch, the hdd would just freeze multiple times each day, it still freezes but rarely

[Melhisedek]

Oh so I'm not alone... Do you use Vista to? My drive is Seagate as well... It is driving me crazy. I disabled SMART in BIOS closed unused SATA ports and set SATA to Compatible it is "acting" as IDE. Any harm in disableing SMART? Also is there some kind of seagate health check software?

[MarkJohnson]

I too get lockups randomly after passing all stress tests. I just readjusted my overclock (bumped vcore) and no issues since (only been a week or so)

What errors does it give to make you think it is AHCI?

-=Mark=-

[x3mgamer]

Oh so I'm not alone... Do you use Vista to? My drive is Seagate as well... It is driving me crazy. I disabled SMART in BIOS closed unused SATA ports and set SATA to Compatible it is "acting" as IDE. Any harm in disableing SMART? Also is there some kind of seagate health check software?

yeah i use vista ultimate 64, i installed vista in emulated ide mode but switched to ahci with this procedure:


1. Exit all Windows-based programs.
2. Click Start, type regedit in the Start Search box, and then press ENTER.
3. If you receive the User Account Control dialog box, click Continue.
4. Locate and then click the following registry subkey:
HKEY_LOCAL_MACHINE\System\CurrentControlSet\Servic es\Msahci
5. In the right pane, right-click Start in the Name column, and then click Modify.
6. In the Value data box, type 0, and then click OK.
7. On the File menu, click Exit to close Registry Editor.

go to bios, change to sata ahci mode and restart

im not sure is it better to leave smart disabled but there r seatools on seagate site for download

I too get lockups randomly after passing all stress tests. I just readjusted my overclock (bumped vcore) and no issues since (only been a week or so)

What errors does it give to make you think it is AHCI?

-=Mark=-

it doesnt give any errors, it just freezes the entire desktop, u can move the mouse but u cant click on anything, and the browser is kinda whitish, like its been hit with a glass of milk

[Melhisedek]

Is AHCI better than IDE mode? should I be using that instead?

[x3mgamer]

im not sure, i never noticed any difference