An orb guillotine
Oh snap, what a pun
An orb guillotine
Oh snap, what a pun
Team Rufio
Splash a little on your northbridge while your at it.
If you have a cooling question or concern feel free to contact me.
If you use that skull on the top of the rod you'll be my overclocking hero
The arc looks VERY nice, I can't wait to see if/how the direct pressure over the core effects your overclock!
Phenom II 940 BE / ASUS M4A79 / HD5770 Crossfire
3770mhz CPU 2600mhz NB | DDR1040 5-5-5-15 | 900/1250
Guess who's your new hero..Originally Posted by afireinside
The added pressure right over the core is great.
For me, I think the main advantage of the press is that there is no need for holddowns. This is key, as when working with ln2, you will get tons of condensation and ice where the inuslation is weak. The holdown area is always my worst spot. With one straight tube of 3/4 insualtion, I have eliminated all cond problems in that area. Of course your contact will be perfect everytime as well and there is less chance of damging a spreader-less chip with a crooked mounted container. For Dothan, the obvious advantage is no more need for the p4c mounting bracket and all the insulating issues that came with that.
@ k|ngp|n: Can you tell me, when do you know you have to stop turning? So you don't crack the core. Is there a spring in the rod holddown to prevent crushing?
well the core has some apllied neopreen or a pad next to it so it wont crack, but to many pressure will crack it indeed. but i think you will know when it is tight!
*EDITed by IFMU*
Signature Removed by XS Staff for not following XS Policies
Forum Policies
I don't think neoprene is needed here.Originally Posted by Waus-mod
Cos you put the pressure directly on the core. The container will always be horizontal.
I want to make one myself. I'm just not sure what kind of metal I'm gonna use for the arch.
Seem to me that the rod might shrink a little @ LN2 temps causing a little less pressure.
No spring, but thats a good idea. I was thinking the same thing, maybe the rod slides through the block smooth no thread and a nice thick spring could load up a little as you crank it down.Originally Posted by vegeta
The stainless is good tho, and gives just the right ammount of flex for a little spring action. You really have to overturn the rod alot to crack a core.
I knew it would make insulating a lot easier but I guess what I'm asking is will the direct over core pressure give better thermal transfer and thus higher clock speeds?
I think the motherboard would shatter into pieces before the die on the CPU cracked, especially with direct pressure. Last DI run I ditched my plywood back plate and tightened my tube to the point where the board was about to snap in two and the die was fine. Then again it was a clawhammer which is quite a bit larger than a dothan die...
Phenom II 940 BE / ASUS M4A79 / HD5770 Crossfire
3770mhz CPU 2600mhz NB | DDR1040 5-5-5-15 | 900/1250
Hmm how tight do you wanne get it then :P tight = tight.. no screw less or more!
*EDITed by IFMU*
Signature Removed by XS Staff for not following XS Policies
Forum Policies
Originally Posted by afireinside
I think with the rod, you will probably get your best contact that is possible...which in theory could lead to the best thermal transfer resulting in higher clocks.
With a container that weighs nearly 5lbs, imo the press/arc is probably the best way to ensure 100% proper contact every time just about.
For me the advantages were to be more functional than for better clocks, but I'll take the clocks too if I end up getting more out of it.
What's the best container design for an arc-bed?
One thing is for sure, the bottom must be flat in oder to work with the rod. So the one with nails, heatsink look-a-likes, holes in the bottom won't work or be efficient enough.
Have you tried an other container than the solid one of MickeyMouse?
I assume the Chilly1 container will be too tall for the arc. But i think that container will have the better result than the sollid ones. Cos it has surface increasement on top of the core.
And the top-platform of the Chilly1 block is flat so the rod can be used w/o interfering the spot above the core. The full cooling capacity of the Chilly1 block can be exploited .
Maybe you can shorten your Chilly1 container and try it , just an idea
its proven solid container is better with ln² then the chilly1 tube
Originally Posted by vegeta
Back in the phase change world
Now that I see how this works I really like the idea. I think I may have to attempt my own. It looks awesome anyway.
So far testing points to the arc/press getting better BIOS temps vs. container temps then traditional style holdown. I have switched back and forth from the arc to the regular holdowns now for 2 days of testing to try and see what differences if any there are.Originally Posted by speed bump
With arc I can get a few degrees colder on bios temps than I can get with the holdown/springs AT A SET CONTAINER TEMP.
I also noticed that if I run at -160 and below for extend periods, I lose my contact and need to re-tighten the rod a little bit to get it back. There is definitely some thermal contraction/expansion going on there.
Some action pics from last few days.....
On p4gd1
This is about as bad as the condensation/frost gets...this is after hours...
DFI nf4 with a 3678mhz fx-57..ing fx could not run below -69c no matter what. At -65c, it's stable for 2d and 3d at 3660mhz or so. Not really impressed. Maybe I am asking/looking for too much out of these chips..LOL. I was hitting almost 3600mhz for 3d with a fx-55@ -80c almost a year ago
I'm thinking intel for sli now. Intel and Pentium M have spoiled me. Cold bugged chips are so booooring. Maybe some really sick clocks on ln2 can make up for the performance differences in SLI between AMD and INTEL.
Last edited by k|ngp|n; 07-30-2005 at 08:49 PM.
lolOriginally Posted by k|ngp|n
Has anybody ever been able to run an AMD below -100º?
Intel Core 2 Quad Q6600
GSkill F2-8000CL DDR2-1000 2x2GB
Abit IP35 Pro
XFX NVidia GeForce 6800GT @ 435/1210Mhz
2x WD Raptor 36GB RAID 0
Antec TruePower Quattro 1000w
Windows Vista Ultimate x64
Sure top jap's...TOM HOLCK..OPB....and others I'm sure. I have benched with a fx @ around -90c.Originally Posted by PoL
-95c seemed to be the the sweetspot with the fx-55 sledge's.....especially the early ones(0432-0438) Memesama and other pro jap overl:banana::banana::banana::banana:ers were running those chips on ln2 at 3700mhz.
I am not looking for below -100c anyways...-90c would be just fine
Last edited by k|ngp|n; 07-30-2005 at 06:15 AM.
Nice results with the arc kingpin... But As you said the expansion of the rod is a little bit bad, but who cares if you control it once in a while. I assume youre happy with the arc?
*EDITed by IFMU*
Signature Removed by XS Staff for not following XS Policies
Forum Policies
Actually it still is only dependant on contact pressure. The major advantage is hang time. With a solid container you don't have to add Ln2 as much due to the mass. However the large mass also means that under load the base temperature will be higher. As we see with thick copper blocks there is still some thremal resistance and a temperature delta. But with a need to not run the processors at absolute lowest values this is not a problem. SO both containers have their advantages.Originally Posted by Jort
I never had much luck with the clamps you squeeze. I prefer the screw type clamp. There comes a point with the squeeze clamp where you can't tighten it any more.
I'm making my homemade arc ATM with a screw rod with a piece of teflon at the end of the screw to make contact with the container base. Hopefully this will eliminate shrinking of the screw rod.
I remember Fugger making a comment about the baker block as compared to the Chilly1 block. Fugger himself said the advantage to the Chilly1 block was the fact that it was solid and didn't have the braze joint which gave you a heat transfer lose. Shouldn't this hold true with these LN2 containers?
Jason and I have 20 of these on order. I will be testing mine next weekend with the screw type hold down. I'm gonna ask Vince for LN2 training.
Last edited by runmc; 07-30-2005 at 11:17 AM.
The arc is rockin'....getting better with adjustments and pressure at max temps.
Worked hard all day to make 19 sec club super pi 1m, and I finally got it on my last run at only 3900mhz/1.72v(1.74 actual). I believe this chip has 18 seconds in it. That's my next goal Once I get ovp dialed in right on my lucky p4c, I should have it.
780es
p4p800-e
buffalo tech bh-5 pc3200
ocz 600watt
19 sec cpuz validated
Picture of the rig after about 5 litres of ln2 and a few hours of hammering
This 780 has taken some xtreme punishment and it keeps on goin'....strong chip.
Last edited by k|ngp|n; 07-30-2005 at 08:41 PM.
Man that is so nice looking. I am major jealous Vince. You certainly have it all dialed in.
Do you have the screw rod directly against the bottom of your solid container?
NICE!....cant wait to start with LN2....but sadly im strill raising money for it
Bookmarks