3700+ SD clocking [vapo LS] and air cooling [massive pics]

[Razor_cut]

Recived my Diego (CABGE 0522TPMW) some days ago, and are now benching with air, but later on it will be LS all the way :cool:

System:

DFI nF4 Ultra-D
3700+ San Diego
1Gb OCZ VX 3200
ATI 7000 PCI 64Mb
Seagate 120Gb ATA
OCZ 520W Power Stream

Cooling:

Vapo LS
SLK-948U + Delta 80mm


http://img86.exs.cx/img86/3042/mrktestning0012be.jpg
Some cleaning befor the actuall test..

http://img86.exs.cx/img86/5654/mrktestning0034qf.jpg

http://img86.exs.cx/img86/7328/mrktestning0043ds.jpg

http://img178.exs.cx/img178/6576/mrktestning0136om.jpg
In the dark the camera needed such a long time to focus so the picture arent that great..

http://img178.exs.cx/img178/1692/mrktestning0072kn.jpg

http://img178.exs.cx/img178/8749/mrktestning0095gh.jpg

http://img228.imageshack.us/img228/4787/rooling0083bj.jpg

http://img156.imageshack.us/img156/5969/rooling0346yi.jpg

__________________________________________________ _

Test [air]

Click for bigger (1600x1200)

http://upl.silentwhisper.net/uplfolders/upload5/2717_-_1.56V_mini.JPG (http://upl.silentwhisper.net/uplfolders/upload5/2717_-_1.56V.JPG)

Memmory is at 200MHz so the points are far from briliant, with this bios 6.23-2 (http://home.no/trac3r/Bios/DFI.LANPARTY.NF4/N4D623-2.zip) I reched at most 2830MHz with a deacent high voltage.. Im now using the 5.10 XCEL2-LL (http://home.no/trac3r/Bios/DFI.LANPARTY.NF4/N4D510-XCEL2-LL.rar) bios :up:

http://upl.silentwhisper.net/uplfolders/upload8/2800_-_1.7V_-8mb-_mini.JPG (http://upl.silentwhisper.net/uplfolders/upload8/2800_-_1.7V_-8mb-.JPG)

http://upl.silentwhisper.net/uplfolders/upload7/2830_-_1.71V_mini.JPG (http://upl.silentwhisper.net/uplfolders/upload7/2830_-_1.71V.JPG)

The problem now is that I'v started to clock the memmory with a 180 divider. 280x10 so the FSB is 254. But when I tried SuperPI with the clocked memmory I get exactly same score. Whats wrong here? Its like the memmory clocking aint working as its so possed to..

http://img171.imageshack.us/img171/3690/254fsb1ou.jpg


http://img247.imageshack.us/img247/5158/255fsbmini3cw.jpg (http://img247.imageshack.us/img247/1455/255fsb1fs.jpg)

Whont go under 29s witch is weird, ,my winchester would do the same, with 290FSB 2.5-4-3-6. But now with 255FSB 1.5-2-2-2 the same points... seems weird to me.

Some of the voltage are probably to high on this screen, but the memmory would probably do the same with 3.5/3.6V.

[Rabidmoose171]

dont run the mobo on that bag... its conductive.

[Razor_cut]

u shure? Iv had no problem with it before. But If you say so I better remove it.

No clues about the memmory question?

thanks!

[Vincentvega18]

Its an antistatic bag, arent they supposed to not conduct..........

[Ashe]

Its an antistatic bag, arent they supposed to not conduct..........
They are supposed to be antistatic but i think they are conductive :stick:

[Razor_cut]

Some pics from the evening oc event that whent *ladidadida* !!

http://img291.imageshack.us/img291/2704/komponenter0016nj.jpg

http://img291.imageshack.us/img291/8936/komponenter0029nf.jpg
Deltan looks so much better then the Alu fan oh yeah!

http://img291.imageshack.us/img291/6282/komponenter0092rw.jpg

http://img103.imageshack.us/img103/8582/komponenter0100jp.jpg
Shaky shaky..

http://img209.imageshack.us/img209/8770/komponenter0139vv.jpg

[Der_KHAN]

They are supposed to be antistatic but i think they are conductive :stick:yeah, conductive yet antistatic, ... makes perfect sense ...

[Jasonhk]

yeah, conductive yet antistatic, ... makes perfect sense ...

lol

[s7e9h3n]

yeah, conductive yet antistatic, ... makes perfect sense ...
I've said this before, but maybe nobody noticed, but the outside of an anti-static bag offers no type of protection what-so-ever. Think about it: If you put something inside an anti-static bag, where does all the static end up? :p:

[JohnyKnoxville]

^^^^^^^Good point man.

[gundamit]

Think about it: If you put something inside an anti-static bag, where does all the static end up? So turn it inside out? :) Sorry OT.

Lets see the main event with the sub zero cooling or at least some more intense/longer SuperPi tests on air.

[s7e9h3n]

[QUOTE=gundamit]So turn it inside out? :) Sorry OT.
[QUOTE]
That definately would work better than outside-out :p:

@ OP: Download spi 1.4 mod to see if you're really running exactly the same times. It goes to 3 decimal places so you'll be able to tell any kind of difference in times.

[fAlCoNNiAn]

the reason its antistatic is because the actual bag conducts, therefore absorbing electricity. basic physics.

[craig588]

The bag is conductive to eliminate voltage differences between points in the board. If a charge happened to hit one point while another point was grounded (In fact it can be any difference in voltage, if one point gets hit with 30,000 volts and another gets hit with the same voltage nothing will happen, but if one gets hit with 30,000 and another only 29,000 then the board gets to experiance 1,000 volts, and differences in high voltages like that can be very common for static) it would make current flow. If both points are shorted (Though the conductive bag) then there will be no voltage difference, preventing current from flowing.

How else would you propose a anti static bag work? All it does is remove voltage differences, 30,000 volts spread everywhere is identical to 0V to the mobo because it can't know anything else exists.

I hope my explanation is good enough.

[ozzimark]

I hope my explanation is good enough.
it's perfect.


razor: nice pics, i always love seeing hardware :D

[Eldonko]

razor: nice pics, i always love seeing hardware :DYah nice pics, keep em comin :D

[Razor_cut]

Updated with some tests :up:

[Razor_cut]

Anyone know where I can download SSE3 patch for my san diego?

Thanks!

[Andrewv]

Nice oc

[Razor_cut]

Im at this point now = http://www.xtremesystems.org/forums/showthread.php?t=68345

The system hangs when coming into windows..
No cold issue either, lunches with -55C.

Appreciate help!

[craig588]

Anyone know where I can download SSE3 patch for my san diego?

Thanks!

Don't use it, it's not legit.

[jrw]

http://www.xtremesystems.org/forums/attachment.php?attachmentid=33905&stc=1

This is my preferred way to mount a board on the table :)

[Razor_cut]

The exact way im doing now :up:

[Rabidmoose171]

i just put mine on the padding that comes in the box... it may cause some heat but not enough to matter, i think nexst time i take it out for benching i will put standoffs in it and screw it to a piece of wood.


just take a multimeter and test the outside and the inside of the antistatic bag, youll see...

[kromosto]

static bag or table go for a nice oc :toast:

[Razor_cut]

Lets put that static behind us ok ;)
Im using standoffs now and its good.

Made me a new NB cooler. Its a Coolermaster Aero 7 something..

http://img275.imageshack.us/img275/562/workingpeaceofjunk0156wh.jpg

http://img25.imageshack.us/img25/9160/workingpeaceofjunk0166cr.jpg

http://img25.imageshack.us/img25/2053/workingpeaceofjunk0194ul.jpg

http://img25.imageshack.us/img25/1584/workingpeaceofjunk0183cu.jpg


http://img25.imageshack.us/img25/6746/workingpeaceofjunk0208wj.jpg
I used this antistatic paste around and under the cpu..

http://img219.imageshack.us/img219/6357/workingpeaceofjunk0230ex.jpg

http://img219.imageshack.us/img219/7343/workingpeaceofjunk0266sh.jpg
Lets see what happens :stick:

http://img219.imageshack.us/img219/6352/workingpeaceofjunk0271kb.jpg

http://img219.imageshack.us/img219/167/workingpeaceofjunk0294dj.jpg

http://img219.imageshack.us/img219/9041/workingpeaceofjunk0322qk.jpg

Happy and al having the comp working, started to clock some, for a while it seemd great. Al I can say is my stepping is totaly crap, if you'r going to use compressor cooling :nono:

http://img167.imageshack.us/img167/2158/vilketjklaskmtmini8yg.jpg (http://img202.imageshack.us/img202/7026/vilketjklaskmt7sx.jpg)

[NiCKE^]

You better sell it and buy a new one ;)
3ghz should be no problem imo with phase but you proved me wrong :)

[Razor_cut]

Well Nick you got that right :up:

Im not sure about my motives, Im probaly just gona stick with watercooling from now on..

/Have a good day

[Jamo]

shame to hear about your bad luck there, so it does seem the later steppings arent clocking so well :(

[kromosto]

must be something wrong it cant be that bad can it be ????

[Rabidmoose171]

Umm what are you volts at, cpu z and sg are reading different, maybe thats the problem?

also how are you going to get a vid card in with that nb cooler?

[NiCKE^]

Umm what are you volts at, cpu z and sg are reading different, maybe thats the problem?

also how are you going to get a vid card in with that nb cooler?
Do you see the PCI card in the bottom? :D

[Gautam]

Bad contact maybe? Shouldn't have the cold bug with an LS.

[Razor_cut]

DFI bios say -16/-14 so maby its a bad contact?

But I checked if there were good squized paste after seeing the evap on place and then unmounting it.

Its a no go probably with later 3700 san diegos..

[bachus_anonym]

DFI bios say -16/-14 so maby its a bad contact?

But I checked if there were good squized paste after seeing the evap on place and then unmounting it.

Its a no go probably with later 3700 san diegos..
Looking in BIOS, SD 3700+ @ 3300Mhz 1.78v, I see -16C... MBM shows -24C IDLE. But @ 3000Mhz 1.45v I get -23C in BIOS and -26C IDLE in MBM ...
If what you show as -15C was @ 1.45v then you might have a bad contact...

EDIT: Well, i take the "bad contact" possibility back ;) I don't know what's up with that last screenie (1.40v in CPU-Z but 1.76v in SG) but If SM is right, then your' temps are pretty much right. So, no bad contact there... But I don't believe that your SD can't do 1M @ 3000Mhz... There's something else that causes SPi fail on you so very quick. Play with it ;)

[brandinb]

not convergent in sqr5 error is one that only happens after spi has failed once before. reload that program and try again. i see lots of this on the web people dont know that after superpi has failed it cannot pass anymore untill you reload it!

like maybe you try it and it fails then you up the voltage and it now imediatly fails well its not your hardware its superpi bug
________
FREE XXXMOVIES (http://www.:banana::banana::banana::banana:tube.com/)

[bachus_anonym]

not convergent in sqr5 error is one that only happens after spi has failed once before. reload that program and try again. i see lots of this on the web people dont know that after superpi has failed it cannot pass anymore untill you reload it!
THAT is very true :toast:

[gundamit]

Way too early to give up on that CPU, but looking back at the shots with it back on air we never saw a stabilty shot higher than the 8M SuperPi at 2.8ghz with 1.71v. Seems like a lot of volts and based on that I don't believe many would have predicted great things for your SD.

Clear your SuperPi directory, reload the program and have another go. Crank the speed (and voltage) back down to stock and creep up slowly while making 32M runs. Get memory timings a bit looser, and use whatever dividers you need to keep the memory speed low, so you can be sure that isn't holding up your OC party. Hopefully you can find a memory/bios combination that'll allow you to use the full range of dividers. Some super flexible TCCD might be the way to go right now. You can always go back to the super tight stuff (Winbond UTT?) once you have your CPU dialed in. :)

[Razor_cut]

Way too early to give up on that CPU, but looking back at the shots with it back on air we never saw a stabilty shot higher than the 8M SuperPi at 2.8ghz with 1.71v. Seems like a lot of volts and based on that I don't believe many would have predicted great things for your SD.

Clear your SuperPi directory, reload the program and have another go. Crank the speed (and voltage) back down to stock and creep up slowly while making 32M runs. Get memory timings a bit looser, and use whatever dividers you need to keep the memory speed low, so you can be sure that isn't holding up your OC party. Hopefully you can find a memory/bios combination that'll allow you to use the full range of dividers. Some super flexible TCCD might be the way to go right now. You can always go back to the super tight stuff (Winbond UTT?) once you have your CPU dialed in. :)

Thanks, it kinda help with thous words. Il be doing such when I start oc again :up:

[Rabidmoose171]

Do you see the PCI card in the bottom? :D


Yes, but i figured he might do 3d benching sooner or later.

[Dynasty]

You can squeeze more out of that 3700 SD. Don't give up so soon.
Take step back and start trouble shooting every possiblity.
That 3700 should do over 3.2 on LS without too much trouble. :cool: