Building an Affordable "Skulltrail" System

[Unregistered]

Mike, just another question since I noticed from your pics earlier in the thread that you appear to be using the same revision of the board as I am!

When the NB overheats, does the system just reset itself? If so, you're problem is identical to what mine was, see earlier post.

 

[Mike]

Northbridge

If you get hold of two Zalman 9700s (either the Nvidia edition or standard) and have the front one with the fan facing downard and the rear one with the fan facing forwards and also a 40x20mm <6.25 CFM fan blowing onto the northbridge, this should keep it well under 70 deg, at least it does seem to work for TJ09/TJ10 cases with all 4 120mm chassis fans installed. You will find that the fins on the left side of the front Zalman 9700 will just kiss the northbridge with the 40x20mm fan, but not enough to prevent installation.

What is your case and how many chassis fans do you have hooked up?

Hi, thanks for your reply.

I have a Thermaltake Armor Series VA8003BWS Black Full Tower Case w/ 25CM Fan. I have outlined my cooling in post #107 above, including 120mm front intake, 250mm side intake, 120mm rear exhaust, 90mm rear exhaust, and 90mm top exhaust, plus the PSU exhaust. The top of the case has 62% opening ratio (i.e. lots of honeycomb grate).

I'm using Thermalright Ultra 120 Extreme coolers (with attached fans), both oriented in the same direction pushing air from front to back (east to west when considering northbridge/southbridge) creating a continuous corridor to the main rear exhaust fan. Additionally, I have a 40mm x 10mm sitting on the northbridge.

At FSB 400 there are no problems at all, even without the northbridge fan. At 425 the northbridge hits 70 then blue screen. I've read reviews from others saying that they managed FSB 425 with 1.35v on the MCH and a small fan to keep it cool.

I can try turning my east cooler to point south, as you describe, and see if that helps. It might help my west CPU anyway, since it runs 5 C hotter than the east. If that doesn't work, I'll try lapping the northbridge heatsink and reattching with better thermal grease. If that still doesn't work, I'll look into buying a custom fan/heatsink combination and get rid of the stock one.

I'm going to have to hurry though. My wife's getting tired of the this beast sitting in the middle of the dining room floor!

 

[Mike]

Blue Screen

Mike, just another question since I noticed from your pics earlier in the thread that you appear to be using the same revision of the board as I am!

When the NB overheats, does the system just reset itself? If so, you're problem is identical to what mine was, see earlier post.

Actually, those pics were posted by someone else. But, as it turns out, I too (now) have the board with the Skulltrail logo on the southbridge. Originally I did not, but my replacement board does.

I'm using Xeon E5420's, so cannot play with multipliers as you can. As such, I can only play with the FSB. At FSB 400 with all stock voltages, no problems. Trying FSB 425 at stock voltages leads to freezing (i.e. lose keyboard and mouse control). I messed around with some other settings and would get system reboot (have not kept a proper journal of cause and effect, sorry). Finally, bumping the voltages to those mentioned in an article I read (FSB 1.3v, MCH 1.35v, CPU's 1.3v or something) would cure the freezing, but then I'd get a blue screen saying hardware failure. This would always happen as soon as the northbridge hit 70-71 C and I would have to manually reset the machine.

 

[Unregistered]

Actually, those pics were posted by someone else. But, as it turns out, I too (now) have the board with the Skulltrail logo on the southbridge. Originally I did not, but my replacement board does.

I'm using Xeon E5420's, so cannot play with multipliers as you can. As such, I can only play with the FSB. At FSB 400 with all stock voltages, no problems. Trying FSB 425 at stock voltages leads to freezing (i.e. lose keyboard and mouse control). I messed around with some other settings and would get system reboot (have not kept a proper journal of cause and effect, sorry). Finally, bumping the voltages to those mentioned in an article I read (FSB 1.3v, MCH 1.35v, CPU's 1.3v or something) would cure the freezing, but then I'd get a blue screen saying hardware failure. This would always happen as soon as the northbridge hit 70-71 C and I would have to manually reset the machine.

I see, so your issues are pretty similar to mine, despite the different CPUs. You might find if you twist one of those thermalrights that it will collide with the NB FAN since their wingspan is even greater than the Zalman 9700, still worth a try though.

The slight advantage of this Silverstone case is that it's front intake fan is directly in front of the video card area, meaning that cool can reach the fan on the front CPU cooler straight away, though I'd have though your huge side fan would more than make up for it!

I'll let you know the outcome of my dealings with Intel...

 

[TGG Productions]

Here is updated pic of my rig. Now with northbridge and southbridge on water cooling.

I don't have any crashing, even after using my rig for 5+ hours at peak!




The cooling on NB and SB really did the trick for everything.

 

[Mike]

Liquid Cooling

Here is updated pic of my rig. Now with northbridge and southbridge on water cooling.

I don't have any crashing, even after using my rig for 5+ hours at peak!

The cooling on NB and SB really did the trick for everything.

Cool (pardon the pun!).
So, why not use water cooling for the CPU's too? I would have thought that once you've gone through the trouble of introducing liquid cooling, that you would want to take advantage of the superior cooling throughout the whole build. Or was it just a cost control decision?

Mike

 

[TGG Productions]

Hey Mike,

To answer the question:
I originally bought the Zalmans, and they are working fine, no need to spend more $ unless you just need to silence the PC. The processers are currently overclocked at 2.97ghz (E5410 stock @ 2.33) default voltage in BIOS, 1.3v MCH, 1.35v FSB and the temps rarely go above 55C...

 

[Mike]

Cooler Orientation

I see, so your issues are pretty similar to mine, despite the different CPUs. You might find if you twist one of those thermalrights that it will collide with the NB FAN since their wingspan is even greater than the Zalman 9700, still worth a try though.

The slight advantage of this Silverstone case is that it's front intake fan is directly in front of the video card area, meaning that cool can reach the fan on the front CPU cooler straight away, though I'd have though your huge side fan would more than make up for it!

I'll let you know the outcome of my dealings with Intel...

Yeah, the Ultra 120's are pretty wide. But, I'll give it a try.

The front fan in the Thermaltake case is relocatable. It comes in a mounting cage that takes up 3 drive bays. You can move the mount up or down into any 3 consecutive bays of the front panel.

 

[Mike]

Liquid Cooling

Hey Mike,

To answer the question:
I originally bought the Zalmans, and they are working fine, no need to spend more $ unless you just need to silence the PC. The processers are currently overclocked at 2.97ghz (E5410 stock @ 2.33) default voltage in BIOS, 1.3v MCH, 1.35v FSB and the temps rarely go above 55C...

Fair enough. What kit did you go with for the liquid cooling? I have a feeling that my fate lays along the same path

 

[TGG Productions]

Thermaltake Bigwater 780i

Its been working great so far, and easy to install. The waterblocks were the hardest to install because I had to take everything out of my case!

One other thing to keep in mind, if you want to run SLI GTX 280's or any other video card which takes up 2 slots, you will need to move the cards to the bottom two slots and get a flexible SLI bridge. The water block tubes take up the second part of the first pci slot for a 2 slot card, if you understand that . A 8800GT would fit perfectly though, just not a 2 slot card on the first PCI Express slot.

I got my blocks at www.dangerden.com
A bit pricy, but really worth it because they are perfect .

 

[Unregistered]

Ok, if anybody has the time, I wonder if they could try running this:

http://www.mersenne.org/ftp_root/gimps/p95v257.zip

When first started up, you'll get a message inviting you to join GIMPS, select the "just stress testing" option, then goto advanced and click "rount off checking", a tick should appear beside it. Then goto "options" and select "torture test" the blend test is selected by default, so just click ok to commence it.

I reccommend running this:

http://www.cpuid.com/download/HWMonitor_112.zip

at the same time as Prime so that you can carefully monitor all temps and voltages and in the event of a lockup, the problem might be more easy to isolate.

See how long you can run the test for, I wouldn't bother going over 10 hours, 95% of the failures I've encountered occur within 9.

I'm just unable to establish why the test is freezing my machine in XP (even after a clean install), yet so far has worked in Vista, to find any kind of a pattern would be extremely useful to me.

Ideally, keep your voltages and speeds at their defaults.

 

[Kougar]

The Blend tests both the CPU and the northbridge/memory. Run the Small-FFT test to specifically test the CPU only... and if it still locks up in XP then it's probably the CPU. If Small-FFT tests fine but Blend does not, then it is likely NOT the CPU, and probably the chipset or memory.

 

[Unregistered]

The Blend tests both the CPU and the northbridge/memory. Run the Small-FFT test to specifically test the CPU only... and if it still locks up in XP then it's probably the CPU. If Small-FFT tests fine but Blend does not, then it is likely NOT the CPU, and probably the chipset or memory.

Yes, I've been monitoring very closely when the lock ups occur and there is no pattern, ie. no correlation between the freezes and the FFT size, sometimes it's 24k, sometimes, 640k, sometimes 320k, sometimes 48k etc....and I still can't think of a reason why Vista doesn't have the problem.

And I've already tried tweaking the voltage on the northbridge and tested two seperate sets of memory on two seperate boards, so it ain't the memory or the chipset, unless it's some kind of inherent fault.

Currently looking into the matrix drivers/write caching and swap files, all of which appear to have more issues with XP than Vista.

 

[Mike]

Northbridge heat problem

Unfortunately, the northbridge fan (40mm x 10mm) was not enough to control the heat at FSB 425. I tried with fan pointing up and pointing down. Temps rise to 70 C then blue screen. I think my next attempt will have to be to replace the heatsink, else remove it, sand it down and put it back using Arctic Silver. Hopefullly that will take it down the 5 degrees or so I need to stay operational.

Well, if anyone is keeping score; I took off the northbridge heatsink and lapped it. It had an obvious arch to the contact surface that took a good 2+ hours to sand down to flush. Once I reinstalled it, I was seeing temps a full 6 C cooler!

This gave me the incentive to try and push the FSB higher (up to 450) to see if I could match what I'd seen on other bulletin boards. As a result, I wonder if I've harmed my MCH in some way. My termperature improvements have completely gone away and now the northbridge quickly and steadily rises to the low 70's again before blue screen hardware failure, even at a lowly 400 FSB with all stock voltages that I had succesfully submitted to over 12 hours of torture a few days ago.

Seems like liquid cooling is looking more and more like a reality. Or, maybe I'll throw all caution to the wind and build an aquarium PC

 

[TGG Productions]

Check your FBDIMM temperatures too. I managed to get them up to 94C

 

[Rob Williams]

Thing is, I ran the test in XP twice and both times it reached 9 hours with the Vcore raised whereas before, the system always froze within 4.

That's strange to me, because usually I find that Vista is far more likely to experience issues with overclocking. As for personal issues, I didn't have any instabilities at all that were related to the the Skulltrail platform itself. The worst issue I had was with two NVIDIA GPUs in SLI, which overheated to the point of games being unstable. Once I removed one of the cards, it was fine.

Wow! That RAM fan unit made a HUGE difference. From mid 90's C down to mid 60's C for the Kingston's with the built in heat pipes (HyperX).

That's intense... I wouldn't expect a drop quite like that. Good to know, though.

What, if anything, has anyone done to control the heat of the northbridge? Would a 40mm x 20mm (as opposed to 10mm) make much difference?

It almost seems like little will help the temperature here. I can't see a slightly-larger fan doing much good, and to get a faster one would just become an annoyance. I'm wondering if a large 80mm fan pointed straight at it (not sure how it would be mounted) would help anything.

Since Core i7 is supposed to have its own Skulltrail, I'm curious if Intel is going to take further steps to prevent this kind of overheating. I'm doubting it would be based on a variant of the desktop X58, but still, their chipsets seem to be getting hotter than cooler, for some reason.

At FSB 400 there are no problems at all, even without the northbridge fan. At 425 the northbridge hits 70 then blue screen. I've read reviews from others saying that they managed FSB 425 with 1.35v on the MCH and a small fan to keep it cool.

I'm curious, what temperature does the Northbridge reach at 400MHz? 425MHz isn't that much of a boost, so I'm wondering if the instability is more the fact that you are hitting the overall limit of the Northbridge. Maybe temperature isn't the main issue here.

Here is updated pic of my rig. Now with northbridge and southbridge on water cooling.

Looking good... that thing is a beast. For future reference though, our forums can accept uploading of images. It's better to do it that way since it doesn't stretch the entire forum ;-)

My termperature improvements have completely gone away and now the northbridge quickly and steadily rises to the low 70's again before blue screen hardware failure, even at a lowly 400 FSB with all stock voltages that I had succesfully submitted to over 12 hours of torture a few days ago.

Ugh, geez :-/ That has got to be frustrating. If you damaged the MCH, you'd probably notice it with regular use of the PC. Even USB devices could be effected.

Since subjects keep getting intertwined here, is someone still having issues with the rig at stock speeds, or are those now gone?

 

[Mike]

FB-DIMM Temperatures

Check your FBDIMM temperatures too. I managed to get them up to 94C

Yeah, I had FB-DIMM temps in the mid 90's last week. But upon installing a triple fan unit over the RAM, they've been holding steady at mid 60's without exception. Best $30 I've spent on this thing so far

My issue appears to be coming from something else. I'm still assuming the northbridge.

 

[Mike]

Northbridge

Ugh, geez :-/ That has got to be frustrating. If you damaged the MCH, you'd probably notice it with regular use of the PC. Even USB devices could be effected.

Since subjects keep getting intertwined here, is someone still having issues with the rig at stock speeds, or are those now gone?

It's happening with regular use, simple web browsing. USB mouse and keyboard seem to be working fine. All stock voltages, 400 FSB (which is stock for the board and the memory, but not for my E5420's). Northbridge rises to 70 and blue screen. I'll have to try with it back down at FSB 333. If it's still a problem there, then I've definately broken something. If broken, I wonder if liquid cooling would be enough to keep it functional?

P.S. What's the answer to the random question for posting regarding the two colors of the site? I never seem to get it right.

 

[Unregistered]

That's strange to me, because usually I find that Vista is far more likely to experience issues with overclocking. As for personal issues, I didn't have any instabilities at all that were related to the the Skulltrail platform itself. The worst issue I had was with two NVIDIA GPUs in SLI, which overheated to the point of games being unstable. Once I removed one of the cards, it was fine.


Since subjects keep getting intertwined here, is someone still having issues with the rig at stock speeds, or are those now gone?

Hehe! Poor old Rob trying to tie up all the loose ends!!

As far as my case goes, I uninstalled the intel Matrix 8.6 drivers and reverted to the native XP Sp2 ones (also disabled write caching on the raid 0 array), this time the test made it to nine hours in xp without locking up and with no vcore increase (I manually stopped it this time)!

I doing one more test now with the drivers re-installed.

I accept what you say about Vista aswell, that's why I was thinking this may have been a compatability issue all along....oh well...hopefully I'll have resolved this by the end of next week...once I've had another go at Intel!

It's tricky to compare all our issues since we're each running a slightly different configuration - specifically on the CPU front - this is bound to yeild varying results.

 

[TGG Productions]

It's happening with regular use, simple web browsing. USB mouse and keyboard seem to be working fine. All stock voltages, 400 FSB (which is stock for the board and the memory, but not for my E5420's). Northbridge rises to 70 and blue screen. I'll have to try with it back down at FSB 333. If it's still a problem there, then I've definately broken something. If broken, I wonder if liquid cooling would be enough to keep it functional?

P.S. What's the answer to the random question for posting regarding the two colors of the site? I never seem to get it right.

Yeah I got 94C WITH a corsair memory fan on the FBDIMMS.

Didn't help much lol!!