Sorted now :D
Printable View
Sorted now :D
Where is the place to do it ?Quote:
Originally Posted by zbogorgon
I don't know what I am doing right now anyways. I used intel 805 and played with resistance and didn't have much effect. Finally I decided to go with zero to see what happens. Zero gets vcore as close as possible to the one set in bios but that's for 805Quote:
Originally Posted by zbogorgon
i dont understand why coretemp should be the right temp and not asus probe...it is the same probe they are measuring....just getting different results..and coretemp always insanely hihg....
No p5wdh is reading a thermal probe under the cpu itself, on the motherboardQuote:
Originally Posted by [G.N.U.]Fragman
System has been running fine at 3.3ghz but now when i am playing a game the usb keyboard disconnects and reconnects and it takes like 30sec to reinitialize. any ideas?
Prove it, please.Quote:
Originally Posted by OBR
Thank you. You couldn't.
Please don't crap this thread k thx.
my E6600 will not be stable over 376-380FSB... but I have SPI 1m at 402FSB... What can I do to improve... Enabled SPD timings doesnt help and 1,6vmch does not... Maybe better cooling on NB? I have a 120mm 1200RPM on it
Set ICH chipset to 1.20V :)Quote:
Originally Posted by yike99
I want to be stable on 400FSB.. please help:_D
I have just upgraded my bios from 701 to 1101 and now PC probe talks even more rubbish. It says my cpu is at somthing like 256c then goes back down to 30c. It says my cpu rpm is dodgy and my case temp is high.
I say this program is not very good at monitoring the P5W DH.
Interestingly I had no problems with it apart from system temp on bios version 701.
hmm that 256 is not bios related i think... But is there any improvment.. stability or overclock on 1101.. try everest ultimate edition and core temp to see temps
with 1101 I can now boot with DDR 6400@1050 mhz 4-4-3-12 for benchesQuote:
Originally Posted by hot_fifty
I'm still stable under prime with less Vcore 1.51 v@3.5 ghz ( 1.47v with pc probe 2 ( 3 hours blend test now and temp cpu is 56° :D ) but I replace NB heatsink yesterday maybe this could explain it too
I got 256° temps spikes too , but I noticed it comes when I run Cpu-z , strange but I don't mind really :)
Finally fine with this 1101 , and it works with ES...:toast:
I did some characterization of FSB vs Vcore with all other factors being equal.
I started with FSB = 330 and went to 380 and gradually brought up Vcore until the point that my system went from unstable to stable. I plotted this and determined the slope of the curve.
I found that for every 4 mV Vcore increase you could get about 1 Mhz increase in stable FSB speed. In other words 330 FSB was the best I could do at 1.3 Vcore. However at 1.5 Vcore I could get to 380 FSB and remain stable. The Vcore vs FSB curve was almost a straight line. Although I didn't try it, it looks like approx. 1.6 Vcore would be required to get to 400 FSB.
I suspect that the slope is the same for most E6600's but obviously the starting points could be shifted depending on the specific device.
For my system I settled on a conservative 366 FSB = 3.3 Ghz at 1.475. I'm affraid to go >1.5 Vcore.
E6600
P5W DH
PCI=33
PCIE = 101
MCH, ICH, Mem = Auto
HP3 disabled
Did you try uping mch?Quote:
Originally Posted by davefr
Reading through the thread is evidence enough, most of these are just silly non-MB related issues to do with user inexperience/ineptitude, and most of those that aren't have been answered multiple times.Quote:
Originally Posted by Falkentyne
If anything, this thread shows that there are only a couple of _real_ issues with the board and has made me more confident to get one :)
This was pretty much my experience with my retail E6600 on the DS3 as well. Unfortunately, it looks like when I get my P5W DH Monday, I will be no better off with it than I was the DS3. :( So much for 975X being better for the 9x+ chips.
Quote:
Originally Posted by davefr
btw for me, my asus p5w undervolts by .05v so I'm fine running 1.55v in the bios for around 3.35 or something..
if you don't have a dmm then buy one..
I found that I needed the following voltages to get the following clocks:
1.40 = 333FSB Stable = 3000mhz
1.50 = 350FSB Stable = 3150mhz
1.55 =370FSB Stable = 3333mhz
1.65 = 400FSB stable for dual prime 32m, but crashes dual SP2004 and BF2
This is with everything else on Auto (Vmch, Vich, Vfsb, RAM, etc). Hyperpath disabled, etc.
Even trying to run 375fsb with 1.6Vcore and MAX MCH, ICH, FSB, makes to different at all. The only setting that helps my OC stabalize is Cvore.
Maybe my CPU has reached its limit. It runs about 30c idle/42c load even at 1.55vcore though. But I think it's the board.
I would recommend no one else buys this board until Asus releases a proper BIOS (or maybe another revision). You can find higher OCs and more stability with other boards for much less.
This will also put pressure on ASUS to release a better BIOS and help the rest of us out.
wtf, no way lol :stick:Quote:
Originally Posted by John600rr
Yeah that's going to happen:poke:Quote:
Originally Posted by John600rr
I don't need anything like those voltages and I'm quite happy with mine thanks.
Quote:
Originally Posted by mion
Yes, I chose a point where I was on the threshold of unstable to stable and played around with mch. It didn't make any meaningful difference.
mine is fully stable at 400 x 9, 6600ES Step5
vcore in bios 1,55
vfsb 1,5
vmch 1,5
bios 801
2xsp2004
2xtoast
1xsuperpi 32m
1xprime95
1x3dmark2001 loop
during 1 night ( 8h ), no crash, temp is 51 for cpu, 60°c chipset in full
On air, water or SSQuote:
Originally Posted by noobzed
FSB WALL TESTING RESULTS AND A CONCLUSION
I did some work last night regarding FSB wall on this board....
P5W DH deluxe with Bios 1101
E6700 OEM Step 6 B2
2GB OCZ 6400 Plat 4,5,4,15
Results:
I have been having real issues at 380FSB and Above. I tried out an E6600 and that worked up to 400FSB, but no higher.
Another post of the foum put me onto crystal http://crystalmark.info/download/index-e.html which lets you change the multi in windows on this board.
So I tried the E6700 at 9x and 8x to see what happened.
The only Bios setting that helped at all was to set the ram to SPD - no voltage changes let me go higher than 379FSB. With SPD on i could do 410FSB before it would go no further. I should point out that SPD sets my ram at 5,6,6,20 - miles out from what it should be.... this got me thinking...
Next I booted with SPD settings and used memset http://rv.page.cegetel.net.perso.cegetel.net/MemSet.zip to see all the ram settings. I took a screen shot and saved the memset settings.
Then I rebooted changing back to manual RAM settings in the bios 4,4,5,15
When in windows I opened memset and changed all the settings apart from 4,4,5,15 to be the same as the screen shot at SPD.
I was then able to go right up to 440FSB before i suffered any issues.
Summary:
1. FSB Wall at 380FSB with manual Ram settings in Bios
2. FSB Wall at 400FSB with SPD ram settings in Bios
3. FSB Wall at 440FSB with manual Ram settings in Bios but changed to match SPD in windows with memset.
Conclusion:
Asus need to provide access to far more RAM settings in the Bios. This might well be the cause of the FSB wall and would explain why different people get the wall at different points... down to when the RAM fails due to its settings.
Nice summary Simon :)
Might be helpful
Nice findings.. I'll look into that when I get some time..
Interesting, something else to try :)
What BIOS is the best overclocking one so far?!?! I'm on 0701, as it came with my board, I'm using an E6400 with 2x1GB Mushkin EM6400 on it!!
I'm at 8x438 right now, vMCH at 1,65 Volts, I think I've hit a wall on the board, so I'm thinking of modding it, but I want to test some other BIOSes first... :)
I haven't seen any wall as such, it goes as high as 420x9 and the chip craps out. And all with manual memory settings from the get go. So maybe different memory handles the standard settings better than others. If you were to try your experiment again you may even be able to narrow it down to a couple of timings.Quote:
Originally Posted by simonmaltby
I've ordered some of the Geil Ultra 4,4,4,12 that is reported to do well... I will see if the settings the bios picks up change with these when they arrive tomorrow.Quote:
Originally Posted by fornowagain
I agree, it could well be the RAM that makes the difference.
thank you so much seems your findings seem to apply to my Intel D975XBX Badaxe too! i loosened tRFC and tRD from 35/6 to 40/7 and went from 433fsb max fsb to now benching at 440fsb on 1067fsb strap (will try higher later )Quote:
Originally Posted by simonmaltby
http://fileshosts.com/intel/Intel975...2s547ms_tn.png
Maybe all motherboard manufacturers need to do is open their bioses to the other memory timing options that memset has available so we can further fine tune how high the fsb can go ? :)
:) interesting...
A shame however that my cpu craps out before I hit over 400fsb :(
Good summary. i would be very interested to know if the couple of asus "Engineering Sample" boards that are floating around that allow higher Vmch up to 1.85v would exhibit this same pattern or would the vigher chipset voltage counteract this? Maybe someone who has don a hardwrae vMCH pencil mod could repeat this test.Quote:
Originally Posted by simonmaltby
Quote:
Originally Posted by davefr
the problem with vcore increases is that temps scale considerably higher as well with even slight vcore increase steps.
Yes, I think most of this is "platform" related limitations and not bugs at all.Quote:
Originally Posted by unclean
so far so good
1.5v real
:woot: Nice one :)
I'm SP2004'ing at 3.7Ghz (370x10) at 1.4750v at the moment. Previously done an 18 hour run at 3.6Ghz (360x10) at 1.4250v, and a 3.4Ghz run at 1.3250v before that...
Getting a bit worried now because it looks like I'm gonna run out of sensible 24/7 voltage before I get to 4Ghz. Anyone know if sticking it under a single stage modded Vapo LS give me more stability at the same volts? I haven't changed any of the other settings yet, only vCore (and the obvious stuff like disabling Hyperpath 3, etc). At what point will I need to change vMCH, vFSB, ICH etc voltages from "Auto" ?
CPU showing ~60 degrees C loaded at the moment in PC Probe, I daren't look to see what it's showing in Coretemp.
ofcourse it will, lower temps are (almost) always better :)Quote:
Originally Posted by Durzel
nice clocks.. wish mine did 3.6+
BINGO, Simon !
I had the same problem with G SKILL HZ DDR 2G for my 939 setup-
I ve printed some of the settings from other ppl, wham ! :eek: , no boot , so I
did basically the same -SPD settings and A64tweaker in the XP .
It s the proof that DFI MB do not have all those mem settings in BIOS in vain
:D
Well, it crapped out a few moments ago in SP2004 - only did 30 mins so I guess 3.7Ghz/1.4750v isn't stable. Hopefully 1.4875v or whatever the next highest one up will do the job.Quote:
Originally Posted by lawrywild
I'm sure you'll get further with your chip lawry, fingers crossed for you :)
Mines an ES so I'd be disappointed if I didn't get 3.6Ghz to be honest!
Great post Simon :)Quote:
Originally Posted by simonmaltby
The question is now I guess - will Asus (or someone who is able to mod it) provide a BIOS with these memory options available, or is there some way to set them automatically when Windows boots? (I've never used memset before so there might even be a way to do that now)
dam..
I just passed 2 hours.. didn't get any errors but evrything just shutdown :(
They have to be in the Bios otherwise if you suffer you cant even get the board to post. memset can be set to make the changes at logon... but will not change the CAS from 5 to 4 without crashing first.Quote:
Originally Posted by Durzel
Also my method of changing the settings in windows proves that the SPD settings are all over the place and therefore its hit and miss if you get a clean post at all.
maybe a bad PSUQuote:
Originally Posted by lawrywild
psu is fine, it's not new, it's been running my x2 rig which drew more power than thisQuote:
Originally Posted by hot_fifty
Nicely done.Quote:
Originally Posted by simonmaltby
Yeah I had the Ultra-D, it really can be one settings that makes all the difference. Anyone got the DFI Infinity 975X? What that got for mem settings?Quote:
Originally Posted by famich
Anyway, here's my sub timings untouched on the HZ, these go well over 400 without SPD. Any major difference?
http://img291.imageshack.us/img291/5651/timequ3.jpg
Does changing the normal-turbo bios setting change any of those mem timings?Quote:
Originally Posted by fornowagain
Last time I'd tried turbo it locked out my vcore to auto. The only way I could get it off once on, was to load defaults, 903 bios btw. 1101 crashes on mine.Quote:
Originally Posted by Plaicd
Hmm voltage scaling on conroes is funny :D
3ghz stable - 1.3v real
+(0.25ghz)
3.25ghz stable - 1.4v real
+(0.13ghz)
3.38ghz not stable - 1.5v real
Is there anyway to enable manual timings after setting the SPD to automatic and getting the higher OC? Memtest just locks my system up if i try to lower the timings, and I can't get past 357ish without setting it to Auto :slapass:
I cant beleive all the trouble people are having with this obviously shonky m/b. I might change my order to something less shonky, if there is such a thing. What else would be recommended for a retail 6600?
P5W DH is a good mobo IMO, and with another one you'll get its own issues , be sure about that .Quote:
Originally Posted by Crankybugga
This mobo was one of the earliest to chip, so thread about issues is very consequent I think and lot of people right here blame it cos they can't reach clockings as we saw .
I think I'm a noob clocker but I learn every day from XS and I know too what are my limits beside guys more experienced
This mobo has to get more mature bios to be un "bon cru" as we said in France for good wine ;)
regards
I think all we need is the really basic issues to be fixed in a bios update and then more memory timings added. the spd thing is ridiculous, it's really the only real problem I have with this board..
Quote:
Originally Posted by lawrywild
Lawry ... One more issue to add to the list
BSOD when installing Windows XP SP2 . I got it installed when i disable the 1394 and the Jmicron controler in bios . Does anyone know what is going on ? :confused :
New version of DH Remote driver has been released:
ftp://ftp.asus.com.tw/pub/ASUS/misc/utils/DHRemote
BTW, does anybody know is it somehow possible to use MBM with P5W DH?
Strange. I installed Windows XP SP2 with no issues with nothing disabled and the WiFi card plugged in.Quote:
Originally Posted by toledo
It really does not make any sense . I was using bios 0401 at that time . What bios were you using ?Quote:
Originally Posted by Durzel
have you tried another PSU?Quote:
Originally Posted by toledo
there was the thread regarding Antec NEO, but someone stated they had the same trouble with a Powerstream...
just a thought.
EDIT: im dumb, that person was... YOU !!! Sorry dude.
I guess it only happens to some people. I disable those types of devices in the bios usually on first boot, I never use or will use them so I automatically disable them.
I have every onboard device disabled in bios and the wifi card removed. Saves me the trouble of troubleshooting issues like that..
hahahaha :D . No problemQuote:
Originally Posted by bullet2urbrain
I`ve posted the same in other thread , i need to disable firewire and Jmicron controler using bios 0401 ..... I really don`t understand this board .
I am having problems with my Samsung 16x DVD burner as the only conection on the PATA port. Ive used both the Asus supplied drivers and the new ones from Windows update. In each case when I access the optical drive my mouse becomes very slow and the system slows to a crawl. When the access is finished the computer returns to normal.Quote:
Originally Posted by MacClipper
My temporary solution is to remove the jmicron raid software from the system . This gives me back my performance but I dont see any jmicron drivers in the device manager.
I'm considering just getting a SATA burner and selling my Samsung :(
Jeff
perhaps double check your settings/drivers with these posts,
http://club.cdfreaks.com/showpost.ph...5&postcount=13
http://club.cdfreaks.com/showpost.ph...5&postcount=19
You can check the burner DMA status with Nero Infotool too or CDSpeed (test the burst rate).
PS: I have no problems with running the new 18x Samsung burner on the JMicron.
Quote:
Originally Posted by BulldogPO
I forget exactly which one but one of the Asus P4C800 choices in MBM5 works everything except vcore.
OK, I got my P5W last night and got everything installed fine and the thing's FAST. RAID0's running great. Cool.
PROBLEM:
Whenever I restart the computer from Windows, one of my HDDs clicks off and spins back up again. It's like the board kills power for a split second to during a reboot, causing it to spin back up. This then causes a slowdown during POST, because the BIOS has to wait to detect the HDD before it's at full spin speed. It does NOT do this with my 2 RAID0'd WD740ADFD Raptors, just my WD 250GB 16mb SATAII HDD. Or if it does do it to the Raptors, I can't tell because the BIOS detects them immediately. It's the weirdest thing.
The ONLY thing that's changed is going from a DS3 to a P5W. This never happened with my DS3. How can the mobo be killing power to a HDD during a reboot??? Any ideas?
My PSU is more than adequate and again, it never did this with my DS3. I'm on 1101 BIOS.
Thanks.
Has anyone resolved the problem on ram divider over 340FSB?...cause I got Gskills with D9 chips and I have them under 400mhz because the board don`t let me have them on any divider more than 1:1 :slapass:
Regards,
Primoz
I'm pretty sure it's meant to do that, my Asus board does the same thing, cuts power for a second then reboots - got to admit it's odd as I'm used to my old DFI NF4.Quote:
Originally Posted by Brahmzy
Yep mine does this when it's overclocked, powers off & and then powers back on when rebooting. Very strange.Quote:
Originally Posted by matt1
Anyone Know what it means if the power led is flashing all the time?
have I done something wrong or is it worse? I have never seen the LED flashing all the time before.
hmmm no ideaQuote:
Originally Posted by SteveS
did you plug it at the right place ?
Yup I just finished putting everything together, and plugged in the power, turned it on at the back and instead of the LED coming on as it should it continuiously flashes.
Needless to say I pulled the power straight away. I have disconnected everything that has power attached and tried again. Same thing.
Mine does that too I think, I usually notice it during shutdown. I shutdown the rig, windows shutsdown and screen turns out then harddrive seems to go off and back on again, then 5 seconds or so later evrything will actually all turn off...Quote:
Originally Posted by Brahmzy
ok.. I'm really started to dislike this motherboard the more I try to push ocs
6 hours orthos stable at 3.24ghz.. I keep the same settings and run orthos, runs for about 10 mins without any errors then suddenly decides it wants to restart itself just like it did when I was testing 3.38ghz
WTF is up with this thing. It's not my psu.. not my harddrives.. not my cpu (errors come before restarts and already primed at that speed).. could be ram I guess.. it's stable in dual windows memtest so..
I sometimes get that too, testing stability is becoming a real pain at the moment because it's difficult to get consistent results. Originally primed 3.1G on 1.4v for 4 hours, so thought I'd see lowest volts for 3G, kept failing/crashing so now I'm on 1.38v and it seems to be happy. A64 + DFI NF4 was alot more straightforward. :(
true but that doesnt mean that he can call 80% of the ppl suckers in here....it might be platform related, but motherboard is still a part of the problem, even though it is not the real issue......but the mysterious way it acts when hitting walls really makes me think simon is on the right track......the real issue might be the chipset and how the board handles the RAM timings with high FSB....not something the users as such can do much about other than adjust to the situation.Quote:
Originally Posted by xgman
Problem : Major incompatibility issue with Antec NeoHE Power Supply Units.
Symptoms (not exclusive):
- Multiple (20 or more) un-installable Texas Instrument OCHI Compliant IEEE 1394 Host Controllers devices in Devices Manager;
- Windows XP SP1/SP2/whatever hangs on install at exactly 34 minutes remaining;
- User only able to install Windows XP after multiple (20+) tries;
- Windows freezes on bootup or takes several minutes to boot up to desktop.
Solution : Use another brand of PSU! The incompatibility is more likely when using ASUS powersupplies (not only P5W DH). Antec pretends a BIOS update fixes this (no bios up to 1101 on P5W DH works). They also offer to fix the PSU via RMA (yet to be tested).
Props to Carcamel for the solution! You da man!
See my thread for more info about this:
http://www.xtremesystems.org/forums/...d.php?t=111370
That was my most frustrating build ever so far. But the reward should be plenty considering I upgrade from an Athlon XP :D
Almost the same here , i shutdown the rig but seems that the hard drive shuts down first ....i mean the mobo is cutting power from the drive before windows is off :confused: . This mobo is really weirdQuote:
Originally Posted by lawrywild
BTW i am using the same drive as Brahmzy (WD 250GB 16mb SATAII )
I put my multimeter on both +5V and +12V leads on the HDD that's affected. Not even a blink of power loss. The HDD has FULL power when it flips off then on again.Quote:
Originally Posted by toledo
So....SOMETHING in the mobo is sending some kindof signal through the SATA cable to bounce power on the HDD during a restart. WTF?...annoying as hell.
Maybe the ICH7 controller does not like this HDD or it is crazy for some reason . I will install the latest chipset driver and see if it helpsQuote:
Originally Posted by Brahmzy
Thats funny - the opposite of my experiences. And I would suggest that what you might be seeing is a result of thinking that 4 hours priming = stable. That is nowhere near enough to tell.Quote:
Originally Posted by matt1
I have seen failures at 16 hours, and even a heartbreaker on 36 hours on my DFI. In my experince, ASUS boards tend to be 'true" - they tell you rather quickly where their limits are. DFI was tantalising - awesome overclocks, but the Vcore needed to make them truly stable (24+hrs Prime95) was always way above what you started with.
I found using systool the quickest way to get there - do things the same way every time, use it to establish freeze/crash limits, the back off a couple of percent and test P95 for a long time. Quickest way to learn the limits.
I know 4 hours isn't stable, it was just a quick test, but it was just the inconsistency of results that worried me.Quote:
Originally Posted by davexl
I may have to try this :)Quote:
Originally Posted by davexl
or use setFSB.. works like a charm... So nice
Does anybody know what Vich is supposed to improve in o/c ?
Just saw differents settings but always let them on auto for now and didn't retreive any interresting info about it
thanks :)
Well I consider 6 hours orthos stable to be stable enough to run orthos again for 10 mins a few hours later.. no errors just shutdown..Quote:
Originally Posted by davexl
I just reset mine to stock again... this shutdown issue is really starting to get on my nerves.. atleast the board is booting WAY faster at stock :)
PS: Anyone know if 0088 works with retail conroes?
Boot faster... You talking about the "cold boot OC protection feature" or just slower Windows XP boot?Quote:
Originally Posted by lawrywild
I noticed mine takes more bars to load XP when OCed. Not 100% sure though
Mainly the cold boot thing + POST time (it takes less time to detect drives and enable the USB controller when at stock it seems)
windows load slightly faster oced
My mobo is showing 53C on the same ambient temperature . Hot as hellQuote:
Originally Posted by xpsentity
I`m having problems to have a Presler at 4Ghz that use to run at 4.2ghz@1.3v on the P5WD2 ..... I can`t boot at 4Ghz no matter what :confused: , using bios 903 btw
Quick question, do all boards have these issues?
Also, is there a "best" bios for overclocking results?
any tips or advice (settings) greatly appreciated.
I have this mobo, mushkin xp2 pc2 6400 and an x6800. I am going to build it in 5 hours from now.
thx
:dammit: after doing the Illustrative prepare your ASUS P5W DH Deluxe i heard a pop sound and had smell smoke upon first boot after that. Keep in mind i ran the the motherboard flawlessly before the Illustrated Preperations. Now i get no post, no beep. Everything powers up and fans spin.
Bios is 0801
Is this motherboard officially defective or is there still hope
Ok, I may be a noob, but I think you're in trouble.Quote:
Originally Posted by 450tantrum
Have U used the AS5?...I think you made sth wrong and if there was some smoke your MOBO is gone :confused: ...put all the coolers apart again and check all the chips and resistors is there is sth strange or burned...Quote:
Originally Posted by 450tantrum
Reagrds,
Primoz
p.s. Hope it is just your MOBO and not the CPU or ram...cause in my case (DFI NF4+opteron) the MOBO has burned the CPU too...:slapass:
Quote:
Originally Posted by Primoz78
I cant find any physical damage and i didnt see smoke but i did smell it...Everything still powers on but no post....i also noticed it killed my RAM that were in at the time, but again no physical damage.....tommorrow my new RAM comes in i hope that helps and doesnt become damaged themselves from being used on this motherboard
Quote:
Originally Posted by Primoz78
what do you mean by "AS5" and "sth"
If I were you I won`t put the new ram in this mobo!...just send it in RMA if there are no physical damage! I`m pretty sure it will kill your new ram if you put it in!
Regards,
Primoz
p.s. AS5= Arctic silver 5 thermal paste...it is conductive so there is a possibility you have made a short-circuit (I don`t know it is the wright word for this...)...sth=something
I believe that is Exactly what happened...So is a short circuit fixable or is that automatic RMAQuote:
Originally Posted by Primoz78
Just RMA it. If it's nuked your RAM something is seriously wrong and I wouldn't even trust it at all. For the sake of a few days you're better off getting a new board.Quote:
Originally Posted by 450tantrum
I assume you didn't remove the covers from the heatsinks yet? If you did, find a way to glue them back on :)
AS5 is not conductive. It's very slightly capacitive and usually nothing to worry about.
I doubt you shorted something with as5. Smell around the board, usually your nose will take you to where the burning smell is coming from so you will know if it's the nb smds..
http://img127.imageshack.us/img127/8715/3ghzrg7.png
Stable enough? :D