Finally!

[MercutioUK]

Now have all my bits (NF7-s, XP 2500 AQZFA wk 0334) finished the basic build last night and will be tweaking and benching tonight, screenshots and scores to follow.
Only problem so far was forgetting to set my west digi's to master for the serial ATA raid (thanks to DOD for the second adapter!) other than that very straightforward, im impressed

[robbie]

Looking forward to seeing the pix!!!
Rob

[MercutioUK]

Originally Posted by robbie

Looking forward to seeing the pix!!!
Rob

Hmmmm.... having a few problems at the moment...
Running 3dmark03 and at some point the bench WILL fail with a BSOD, cant remember the exact message, at work at the moment.
Temps are a little high too.. the bottom hdd cage only leaves about a 1" gap for airflow between the drives and my GF4800... need to work on that methinks.
Currently 36 amb 49-51 cpu. Definitely got some work to do..
Water cooling or a new case might be on the cards.
Anyone got a good source for a watercooling setup? Looking at the waterchill kit but a little pricey so might try a self built.

[bad4u]

Originally Posted by MercutioUK

Hmmmm.... having a few problems at the moment...
Running 3dmark03 and at some point the bench WILL fail with a BSOD, cant remember the exact message, at work at the moment.
Temps are a little high too.. the bottom hdd cage only leaves about a 1" gap for airflow between the drives and my GF4800... need to work on that methinks.
Currently 36 amb 49-51 cpu. Definitely got some work to do..
Water cooling or a new case might be on the cards.
Anyone got a good source for a watercooling setup? Looking at the waterchill kit but a little pricey so might try a self built.

Don't spend a lot of money until you're sure there is a real problem.

Switching from 8RDA+v1.1 to NF7Sv2, I found that not only were CPU and System/board temperatures reported higher at both idle and load, but behavior pattern was different also.

The sensor for system/board temperature is closer to the CPU socket than my previous board, so system temps showed a greater rise and followed the rise and fall of CPU temps much more than the previous board, since the downflow from the CPU heatsink/fan flows over the system/board temp sensor.

I suggest that stability/instability is a better symptom of 'heat problem' than the higher temps reported with the NF7S board. My 'reported' load temps sometimes reach nearly 60C - with no instability whatsoever.

Hope this helps!

[prat]

What RAID level are you using?

[MercutioUK]

Originally Posted by bad4u

Don't spend a lot of money until you're sure there is a real problem.

Switching from 8RDA+v1.1 to NF7Sv2, I found that not only were CPU and System/board temperatures reported higher at both idle and load, but behavior pattern was different also.

The sensor for system/board temperature is closer to the CPU socket than my previous board, so system temps showed a greater rise and followed the rise and fall of CPU temps much more than the previous board, since the downflow from the CPU heatsink/fan flows over the system/board temp sensor.

I suggest that stability/instability is a better symptom of 'heat problem' than the higher temps reported with the NF7S board. My 'reported' load temps sometimes reach nearly 60C - with no instability whatsoever.

Hope this helps!

I did notice the heat sensor is in the group to the right of the sink so away from the exhaust and in the most likely place for higher temps. There is a LOT more heat coming from the system than as was even unclocked... ill try the free stuff first like removing the lower drive cage so I can get some more air flow through into the top of the case... water cooling was always on the cards though even before I bought the chip and board

Originally Posted by prat

What RAID level are you using?

Using RAID level 0 on a pair of west digi cav se`s.... they are contributing an awful lot to the heat in the case... have a spare case fan so I might try blowing a little air over them see if it helps much. Whats the max length for a serial ATA cable? I have 2 drive mounts in the top of the case with nothing above them so a blow hole might be a plan but the cable type included with the board is no-where near long enough.

[timbob2469]

Originally Posted by MercutioUK

Whats the max length for a serial ATA cable? I have 2 drive mounts in the top of the case with nothing above them so a blow hole might be a plan but the cable type included with the board is no-where near long enough.

According to this
The maximum cable length is 1 meter, or 39.37 inches.
Bob

[MercutioUK]

Originally Posted by timbob2469

According to this
The maximum cable length is 1 meter, or 39.37 inches.
Bob

Thanks bob, picked up a pair of 1m cables so should be good.

This is my current plan for cooling... will this work?
______^^^^______
|***|******2x HDD
|PSU|******CDROM
|**********CDROM
=>***HS******FDD
=> ****^^^ ****|
|--Geforce--*****|
|**************|
|*PCICARDS*****<=
|*************<=



Both intake and the usual exhaust sucking air in (21cfm each) and find a 120mm fan blowing the same amount (42 or nearest match) for a blowhole in the top of the case. a fan of some sort (poss an spare 80mm) mounted on the back of the geforce (I have a huge vga heatsink to mount it on) blowing into the top of the case just to help direct air to the blowhole. I was thinking of trying to make some kind of deflector to make sure air doesnt get caught under the geforce as well?

Issues I can see:
The rear intake is likely to go straight to the psu, but this should hopefully take the hot air off the heatsink with it?

Was also considering one of these puppies:
http://www.overclockers.co.uk/acatal...oolers_57.html (top of the page)
Its basically a quieter version of the old "vacuum cleaner" HSF on the 5800FX. Will be a little pointless if I go water but as an option...

[MercutioUK]

Hmmm.... a rather unhappy me...
Got the setup all working... never even TOUCHED the vga settings, set all the bios settings to keep the hell away from it... still started getting artifacting etc followed by LARGE graphics corruption that slowly spread to the point where even the bios/bootup screens were corrupt. Removed the heatsink (for the first time) on the card and the shiney blue of the sink was a slightly lighter, less shiney blue around 1 of the memory chips RIP my geforce. Gonna try an RMA but managed to get the TINYEST scratch on the heatsink as I uncliped it so not likely to happen.

[tech8]

Looks like you still have a few tweaks to make before you get it stable. Anyway I like to suggest a few things when it comes to case cooling. Please see my post low cost and efficient "airflow" design

I strongly suggest that you use only a few auxillary fans for case cooling so as to minimize the Electro Magnetic Interference. Also, quantity doesn't really mean you get quality (think you have too many fans) and there's also the power consumed by the fans from your PSU.

Hope it helps. Good Luck

[danrok]

I think the top blow-hole is over kill. I would stick to the standard setup of rear exhaust, psu exhaust and front intake. You could always add the blow-hole later.

Also, the rear intake will draw in warm air from the psu exhaust.

It is difficult to predict air flow - the best thing is to use the trial and error method until you get it right. Ideally try and get the best air flow with as few fans as possible.

[MercutioUK]

Originally Posted by danrok

I think the top blow-hole is over kill. I would stick to the standard setup of rear exhaust, psu exhaust and front intake. You could always add the blow-hole later.

Also, the rear intake will draw in warm air from the psu exhaust.

It is difficult to predict air flow - the best thing is to use the trial and error method until you get it right. Ideally try and get the best air flow with as few fans as possible.

Ta for the suggestions - with all the system trouble I was having I havent got around to messing with extra fans just yet, just intake and exhaust. Ive managed to move the drives up into the top of the case and moved them apart. Ones currently on the hard drive plate built into the top of my case the other is in the spare 5 1/4" below, jury rigged to the side with some *coughs* handmade supports on the other side (wedged a little cardboard under there till I can get something more permanent). Removing the bottom cage and the death of my geforce along with re-routing of a lot of cables behind the motherboard back plate and into the top of the case out of the way seems to have improved matters quite a lot. Yet to give it a proper stress-test but now running stable @ 200 x 11 (yet to fine tune) with a Tyan radeon 9600pro. 3dmark03 gives me 3280 marks which for the time being will probably do me

[MercutioUK]

Done some tweaking with the graphics, now at 3674.

[MercutioUK]

I know this generally comes down to system stability but what sort of voltages should I need on the chip to get 210+ FSB? as soon as I push the fsb (anything over 209 really) I seem to be hitting a wall. How bad is the cooling on the NB with default sink, paste etc? What would folks recommend if I was to replace and would it then be worth cleaning up/lapping the NB coller and attaching that with some thermal adhesive to the SB? I quite fancy going for the 220 FSB. My CPU temps get a little too high (52c) to do much with the multi for the time being. I get the feeling I should be able to do more with my current cooler (volcano 7+) so might have to get my hands on some finer sandpaper and give it another go