i did some LN2 testing a couple of days ago
have a look at the amount of water on the board and the mofo worked perfect pushing an E6850 to 5350MHz 2D/3D stable
http://www.xtremesystems.org/forums/...d.php?t=170796
Printable View
i did some LN2 testing a couple of days ago
have a look at the amount of water on the board and the mofo worked perfect pushing an E6850 to 5350MHz 2D/3D stable
http://www.xtremesystems.org/forums/...d.php?t=170796
Well, I've finally got my water cooled rig up and running and so this board is now my main PC.
Currently busy installing everything I need before trying to push the OC further. However, I have noticed a couple of small problems.
I'm currently running F5h bios with Q6600 @ 3.6GHz (400*9) with 4x1GB Ballistix PC2-8500 @ 1066 24hr Prime stable.
I've just installed WinXP on 2 * 74GB raptors in RAID0 and here comes the first annoyance. In order to set up the RAID, I had to switch the SATA RAID/AHCI Mode to 'Raid' (obviously) and since doing so I do not get the logo screen during POST. The system goes straight to the RAID drive detection routine and ends with the option to press <CTRL> + I to configure. No option to get into BIOS. I've tried pressing <DEL> constantly, but no luck. I can occasionally get into BIOS using a combination of <DEL> and <CTRL> + I, although I cannot determine the pattern required to get in consistently. If anyone has any ideas on this, please let me know. Oh - and yes, I do have the option in BIOS set to show the logo screen :D.
The other annoyance is a cold boot bug (I think). As I'm not exactly sure what the 'cold boot bug' is, I'll describe the symptoms. If the PSU is disconnected, then the 1st boot after reconnecting fails. Lights come on briefly and then go off. The system then automatically reboots and this time it is successful. Now, seeing as I don't normally disconnect my PSU, and also considering that this problem requires no intervention on my part and only adds a fews seconds to the boot process, I'm not all that bothered about it. However, if someone knows of a fix to this, then I'm willing to give it a try.
Thanks in advance for any advice.
~aoe~: The bios does this... on first boot it checks everything (PSU Voltages, sets Memory timings (If is on Auto), CPU, and the other devices). If something is wrong the Bios automaticaly tryes to set it, and if it can't than it resets to it's defaults.
What You have there is not a cold boot problem. A cold boot problem is if your system doesn't start at all, or something fails to detect or start at boot (like it doesn't detect the hard drives for example), after a night (with the PSU conected), when all computer component are "cold". :)
The problem with entering the bios, I didn't had it when I tried raid or ahci. Try the latest Bios F7H (look in predvious posts there is a link).
Thx for the response ChrissTi.
I wasn't sure about the boot thing. It's new behaviour since I enabled the RAID function so I wasn't sure what was happening.
I'm sure to try other BIOS revisions once I've got everything installed and try to push the system a bit more. Maybe a newer revision will solve the problem of no logo screen. It is very frustrating not being able to get into the BIOS :ROTF:
Hmm, I see. I tried removing the overclock (again) so it was just at the stock 266 x 9, but that made no difference. So I gradually clocked it up to 400 x 9 (which is perfectly stable with my system, I just prefer to hold back a little), and that made no difference. During my attempts at solving this, I've also tried setting the PCI-e frequency (mhz) manually to both 100 and 105 (as was suggested by another user having similar problems to me) but neither helped.
Should that just be at 100, or something else? By PCI / PCI-e multipliers, do you actually mean the PCI-e frequency? I can't see anything else to change.
*edit* I also notice this odd form of display corruption in some games (TF2, Lost Planet, for instance). It's a little like extremely brief flashes of model corruption that starts at the dead center of my screen and arcs out in a small triangle. Extremely hard to describe, and actually quite easy to miss if I'm not watching out for it in particular. Remember, I have tested both of these cards individually in both my system and others, and this kind of corruption NEVER occurs outside of Crossfire mode... Whatever is happening is going awry when these cards try and merge their rendering together. Whether this is some bizarre one-in-a-million fault in both of my cards, or something this motherboard is causing (which I suspect because crossfire worked fine a mere day before in my CFX3200-DR, and my friends older P35-DS4). Man this is so lame.
Thanks CC and RD for the memory suggestions.
Looks like 2x2Gb of G.Skill's PC8000 PQ is the consistent suggestion.
Anyone else having success on the X38-DQ6 with 2x2GB?
yeah - I tried mashing the DEL from startup through to the <CTRL> + I message, but no luck. It doesn't go into BIOS and after the <CTRL> + I option, goes straight into windows. I've occasionally managed to get into BIOS after mashing the DEL and then hitting the <CTRL> + I at the prompt, but it's random and sometimes goes into the RAID configuration.
I'll try and clear CMOS and see if that fixes the problem. If not, I'll have to try a different BIOS revision.
Strange problem indeed :confused:
Clear CMOS did the trick. I've got my logo screen back and can now access BIOS :up:
I get the same thing - and I have to cut the PSU current, because it consumes 34 Watts doing NOTHING :mad:, PSU is on but PC is off (I contacted Thermaltake about this, asking them if they ever heard about things like energy efficiency or global warming or stuff like that). Never in my life I'd let something consume 34 Watts doing nothing, so I pull the plug. When rebooting, I get the same symptoms. Bios is F4b
Thanks for that. It's good to know it's not a problem.
I know what you mean about the PSU consuming power when the PC is off. However, I wouldn't blame the PSU manufacturer. It's the motherboard that is consuming the power, and keeping things like:
- USB ports active for charging stuff
- sensors to allow power on by keyboard/mouse/etc..
These are gimmicks that some people like to have. Until we change peoples mindset, this isn't going to change sadly :shakes:
Guns81: I know it may not make sense, but try 1 stick of ram in each memory slot, then try the other stick in each memory slot and see if one any of those work, if they do not, you may have to RMA it.
Apparently any bios after F3i has the memory speed and timings set up for 1333, which is supposed to work on your memory, just like it is supposed to work on my 1600 memory, but something is set wrong. F3i and below have the speed and timings set for 1066, which works fine for me to boot into.
Hope this helps,
Isepick
i'd try..there is only one configuration with a stick on first slot in this case i can see the boot image and then if i press canc del ins canc f9 f12 or something else this is blocked..only tab sometime work...
thanks for help!
I agree that the board consumes power when the PC's off, but I had a 500W Seasonic PSU before and I measured 11 watts (PSU ON, PC OFF) and 6 watts (PSU OFF). So I do blame Thermaltake for the 34 Watts (it's 16 watts when PSU OFF). It's just that I've got a fileserver running 24/7 that consumes 45 watts @ 10% CPU load, so the Thermaltake ToughPower 700 uses 75% of that amount for nothing - seems pretty insane to me. BTW: never got an answer from Thermaltake customer care :down:
~aoe~
cool, you got it working i just read your post and thought maybe a ps2 keyboard would be better than usb ,if thats what you were using ? anyway,glad you sorted it..hey this makes me drool a bit for raid 0 on mine too....whats it like? did you get 2 x 10000 rpms? are they loud? did it literally speed your system up by x2 in certain areas.
im thinking 2 x raptor 36gig for xp ....strange how all these raptors are sata150? i thought they would be sata2(300)
anyone else using raid on a x38 dq6 too?
:)
on
http://www.jzelectronic.de/jz2/index.php
found
26.12.07
GA-X38-DQ6 - TestBetaBIOS Version F7j
More Loadline Calibration fixes
And for those with cold start problems when overclock a cold start fix
But the bios still is not available. It will be available any moment :)
Guess that would be doable with 2 HDs aswell (be innovative), did this once because of noise, used bike-tube, never heard that HD whirling again..;) Hang em' high.
Attachment 69552
haha good one 2diff . :) never seen that before. cool, stops the whole case vibrating like a speaker box
Just wanted to ask a question to be sure I mounted the Heatpipe system right when removing Crazy Cool.
Since I had to use a regular screw and nut for one pinhole on the X38 heatsink, and the spring loaded nut screw and screw combination for the other, I'm afraid I'm putting too much pressure on the chipset.
How much presssure can it take? For the screw I had to use from my own collection, I tightened it just enough for it to be noticeable. When I felt the change in grip as the heatsink contacted the IHS plate of the chipset, I went at most a quarter turn further, maybe less. But with the Spring loaded screw, I can't tell when the spring is doing it's job and when the screw itself is applying pressure. So now I'm left wondering as to how tight the heatsink is grappled with the chipset, and the possibility of damage cause by long term use like that.
Right now it's running fine, but I'm afraid something might "Pop" unexpectedly in the future.
All WD Raptors are 10k rpm. They are louder than other drives but how much louder largely depends on how you install them into the case. I like 2DIFF's solution :up:
Mine were previously installed in a TJ07 using the standard hdd modules supplied. In this configuration, they were rather noisy as there was no vibration dampening at all (and flimsy aluminium modules). However, in the Lian-Li modules, they are much quieter as they rest on rubber washers so the disks are not in contact with the case.
The advantage of Raptors is a greatly reduced seek time (4.6ms read seek compared to around 8.9ms for most latest generation 7200rpm disks). The advantage of RAID0 is faster transfer rates (both read and write). Put the 2 together and they really fly. I wouldn't say you get twice the performance, but it's definitely noticeable. Remember that the slowest component in any PC is the hdd so any improvement tends to be noticeable.
It's not necessary to go with Raptors though. Any RAID0 with recent generation hdds will provide a noticeable performance improvement.
If you do decide to go RAID0, then you will need to decide what data to store on it. Personally, I give priority to my OS and programs (excluding game installations). I like to keep my games on a separate disk as most games will also require OS files (drivers) so this helps improve load times.
I've just moved all but 1 of my Raptors into the new build, so I now have 2 x RAID0 arrays, 1 for my OS and programs and the other for game installations.
There will be some people who will tell you that RAID0 creates more risk of data loss. It is true that the risk is doubled as if either disk fails, any data stored on the disks cannot be retrieved. However, disks tend to be very reliable, so the risk of one failing is minimal. 2 x minimal is still very low risk (still minimal in my opinion). However, again personally, I don't store any data on a RAID0 array that I am not prepared to lose. All my data (docs, videos, pics, program installation files, music, etc..) is on a single disk and also backed up onto an external disk. If 1 of my RAID0 disks were to fail, I would need to replace the disk and do a complete reinstall. For the performance gain, it's a risk I'm prepared to take.
Can someone stick the new F7j beta up on Rapidshare when it's available for download?
Also...can anyone translate from German to English?...i'd say GB are close to an official F7 bios now, i'm guessing boards with F7 will ship soon.
26.12.07 - GA-X38-DQ6 BIOS F7j ... Test läuft...
Loadline Calibration jetzt Auto-Disabled-Enabled
OC bei Strap266 CPUen [2.50A]... Test fehlerhaft ...[2.00A] und [3.00A] OK
OC bei Strap333 CPUen... Test OK
Standby Bug bei Loadline Calibration [Enabled]... Test fehlerhaft ... [Disabled] OK
Info: Empfehle z.Z. noch weiterhin die F7f Version