Nanoblock.. :)

[UnaClocker]

Hey guys.. I'm pretty new to these forums.. I write articles over at ProCooling. I've written an article that was posted here on Overclockers.com awhile back.
Thought I'd introduce myself with a shot of my current waterblock..
http://una.procooling.com/nanoblock/glued_on_in_socket_above2.jpg
hehe.. It's Overclock-Watercool's mini v-tuned block. It's essentially just a rectangular block of copper, with a 3/8" passage through it. It's dirt cheap but doesn't come with a clip, so I glued it onto my Duron with plumbers goop. hehe.. I had expected it to not cool the chip worth a damn, figured I'd be yanking it off within an hour. Turns out, it performs better than my 1/2" Maze1C block. Granted, I only used my in socket thermistor, but according to that, it dropped 3c at full load compared to the Maze1C.. :) I've been running the block for about a week now, maybe a little longer.
Oh, if you look closely at that picture, you can see that I goofed when I glued the block on, it's hanging out over the side of the socket where the arm swings, and since it was glued on while the chip was out of the board, I had no way to get it in. So I had to take the socket apart and move the arm to the other side.. hehehe.. ;) Bet you've never seen that done before.. :)

[Intraveinous]

Welcome to the forums... Interesting... to say the least. :D How easy is it to remove if you ever needed to??? Were I to do something like that, I'd probably throw in a temp probe from a compunurse or something along with it...
Nice set up and interesting idea...
Peace
John

[UnaClocker]

I'm not sure how hard or easy it will be to remove, it'll probably be very difficult. :)

[wolfsid]

Originally posted by UnaClocker
I'm not sure how hard or easy it will be to remove, it'll probably be very difficult. :)

not a bad setup but how exacly did U put the right amount of pressure to the waterblock against the cpu & waiting for the goop to dry i think its crazy to glue anykind of heatsink to a processor but if U are not going to upgrade that one i say good JOB.. .. Also U can make a pretty easy hold down devise for that block... i was going to get that same one from overclocks-watercooled but i decieded to get the Z4 block from overclockerhideout.com which does 8 degrees better then the maze 2-1...

[UnaClocker]

MmmHmm.. 8 degrees huh? That with a pelt? Very odd, I'd written off the Z4 as a pile of dung.. :)

[wolfsid]

Originally posted by UnaClocker
MmmHmm.. 8 degrees huh? That with a pelt? Very odd, I'd written off the Z4 as a pile of dung.. :)


So far i had the maze 2-1 not bad block becooling blocks had two of there also not bad but need lapping alot.. then i got a Z4 nice block good cooling only problem its not like the maze2-1 where U can mount the pelter on it. with the Z4 i had to epoxy it and use a vise but after all the trouble it outpeforms the maze2-1... And Yes I use pelters why else get a waterblock :eh?: I have a 156 watt on cpu want to get the new 220 watt, then i have a 72 watt on both gpu and chipset computer at load runs 17c with prob next to core idle is 14c

then gpu runs at 16c its a gf3 ti overclocked to 250/495

and the chipset runs at 17.2 but i think thats a little high because i didn't use a vise on the copper plate and the pelter on the motherboard i think if i do i should get a colder temp. I might get board and do it one day but for now it works fine...


hack the planet..:D

[Frodo Baggins]

Very nice direct die cooling, I've seen better though. lol, can someone say Sunny D Bottle direct die cooling?

[Colin]

Hey UnaClocker - It's good to see you over here. Welcome to Overclockers from Mr. Thompson! :D

[Dissolved]

thats so cool, i love that mini block. i wonder how it performs..

[UnaClocker]

Hey thanks for the welcome. The guys over at hardocp kind of gave me a helping nudge, but I've been meaning to move over here anyways. :)
This isn't direct die watercooling, it's a solid block. It just doesn't come with a clip, that's why there's glue everywhere. heh..
You wonder how it cools? Well, 3c better than a Maze1C in my current setup, according to my in socket thermistor. About 13c over ambient at full load. Oh, and a little background on my cooling setup. I have 1/2" tube everywhere, scaled down to 3/8" for the block, which uses 1/8" NPT barbs (Chip@OCWC says velocity is very important, hence the narrowness). I'm running a Danner Magdrive model 2 (250gph) and 2 heatercores in serial. :)

[RoadWarrior]

I've seen brass brake parts about that size, right angle blocks. Wonder how they'd work out......

[IFMU]

Just a lil curious here... -Very nice btw!- But is it a solid block over the core? Or does the water actually touch the core? And beyond that can you actually run the water right on top of the core? Or would that cause a short of some sort? dunno... random thoughts... Ill be quiet now!~!

[Silversinksam]

Hey Welcome to the forums bro, Thats an interesting setup to say the least.

Actually I was pondering some thoughts regarding direct die cooling.

What if you took the block and actually made a cutout for the die, then sealed it all up and actually had the water making contact with the die itself since we are talking about semi permenent mods here.

[UnaClocker]

Mmm, I think that the copper is helping to spread the heat out, at least a little bit. I don't think direct die cooling is the way to go, but I suppose there's only one way to find out. I don't have the kind of equipment I'd need to do it right tho.
As for brass parts, brass conducts heat really poorly. I think it's worse than aluminum..
Here's another picture of the block, so you can see the solid bottom:
http://una.procooling.com/nanoblock/on_chip_low_pro2.jpg
That's the only pic I got of the bottom of the block.. That was me sizing it up for a really slim (like 1U type deal) application. :)

[Silversinksam]

Thermal Properties of Materials
Thermal Conductivity, W/cm-K
Metals
Aluminum 2.165
Beryllium 1.772
Beryllium-copper 1.063
Brass 70% copper, 30% zinc 1.220
Copper 3.937
Gold 2.913
Iron .669
Lead .343
Magnesium 1.575
Molybdenum 1.299
Monel .197
Nickel .906
Platinum .734
Silver 4.173
Stainless Steel-321 .146
Stainless Steel-410 .240
Steel, low carbon .669
Tin .630
Titanium .157
Tungsten 1.969
Zinc 1.024

[RoadWarrior]

wow, thought brass was pretty close behind aluminum and worth a try maybe, but those figures don't look good. Got any figures for silver solder btw??

[Frodo Baggins]

Yeha I kinda realized this morning it wasn't direct die cooling. I saw the block over at ocwatercool .com and noticed that heh. kinda jumped to conclusions when I saw that goop of glue

[UnaClocker]

Yeah, to make the price lower, he sells the darn thing without a clip. Figured it'd only be fair to test it as shipped.. hehe.. I put on some arctic silver, then tightened on a c-clamp and gooped all around it and threw it in the oven at 150f for like 3 hours.. ;) Can't bear to wait 24hours for that plumbers goop to cure. :)

Thanks for the numbers on those metals. I had thought the same thing about brass once too. My father is (was, he retired recently) a machinist, and made a drilled passage block for me out of brass, ended up performing horribly even though it was a maze-like design. That's when I learned how horrible brass was. I figured, hey, it's mostly copper isn't it? heh.. Might as well have run a chunk of steel. ;)

[The Overclocker]

do you think it was they way the block was glued down or the block it's self. is their any way to check?

[ButcherUK]

Originally posted by RoadWarrior
wow, thought brass was pretty close behind aluminum and worth a try maybe, but those figures don't look good. Got any figures for silver solder btw??
Solder comes in around 0.5, I wouldn't imagine silver solder would be much better as the amount of silver is actually quite low.

[UnaClocker]

Originally posted by the overclocker
do you think it was they way the block was glued down or the block it's self. is their any way to check?
I'm not sure exactly what your asking here.. Are you asking if the block is performing so well because of the way I glued it on, or just because it's a great design? I'd have to say it's the design, if anything, my attatchment method is hindering it. :)

[The Overclocker]

Originally posted by UnaClocker

I'm not sure exactly what your asking here.. Are you asking if the block is performing so well because of the way I glued it on, or just because it's a great design? I'd have to say it's the design, if anything, my attatchment method is hindering it. :)

i have seen some people glue down water blocks with silicon sealent and get better performance because of pressure between then was higher. but when you think of the ammount of copper to water it is probably the waterblock it's self. i want one - only prob is that i live in the uk, i will try and make one

[UnaClocker]

Good deal.. The trick I'm told is to keep the internal passage 1/8" just like the inside of a 1/8NPT hose barb. That way the water "shoots" through there like a jet. High velocity. :)

[Frodo Baggins]

Cjip said if I sent him money for shipping he'd send me a free waterblock (that was a couple months ago) still waiting....

[wolfsid]

Originally posted by UnaClocker
Good deal.. The trick I'm told is to keep the internal passage 1/8" just like the inside of a 1/8NPT hose barb. That way the water "shoots" through there like a jet. High velocity. :)


OR YOU CAN JUST MAKE THE HOSE BARB BIGGER BY DRILLING IN THE CENTER WITH A DRILL PRESS... ALSO MAKE THE WATER FLOW QUICKER...

[The Overclocker]

how about buying lots and using them as ram blocks...

[Morphoius]

the only way i see that dirrect die cooling would be better than using copper as a heat spreader would be dependant on the temperature differance between your block at full load and your water at full load. if your block gets warm enough and the water stays cold then maybe running water over the block may yeild lower temperatures. I doubt it though because copper conducts heat a lot better than water. it would be interesting if someone were to try it though.

[Frodo Baggins]

Someone did alrdy. And he got better temps thena maze I think.

http://www.spodesabode.com/view.php?pageid=directdie1

[Diggrr]

Yeah, I was reading Spode's site tonight.

I've about convinced myself that I'm gonna try a direct die cooler on my T-bird 1G@1.2G. I've tried 3 different copper blocks without any change in temps...all different designs. (pool, maze, jet)

I've got lots of epoxy, goop, and time, so I'll probably pop out the proc. tonight and start the epoxy job on the bridges/resistors.
I'll have to stop at the hardware in the morning though, cause I'm out of brass barbs.

I'll post back here to let ya know how it goes.


BTW, I was also at Leufkens site tonight, and on his front page he said he's advising people paint the extra bits on the processor with nail polish (2 coats) to waterproof them (not on the core, I think) before attaching his new block. Hmmmm., not too bad an idea.

[wolfsid]

Originally posted by Frodo Baggins
Someone did alrdy. And he got better temps thena maze I think.

http://www.spodesabode.com/view.php?pageid=directdie1


someone goes that extreme to epoxy a processor and don't get nothing out of it... but a broken druon.

system I am running off of a z4water block and a 156 watt pelter...

amd xp1600 running at 2000+xp sorry pr# here 13x133fsb or 1755mhz LOL

ram 512 crucial pc2100 cas 2

gf3 ti running at 250/525 aircooled going to put a waterblock with a 72 watt pelter on her l8ter this week..

ibm 7200 rpm just finished doing a LLF on it dam thing crashed on me earlier this week luckily I have everything backup on my server...

temps are
cpu idle at 11c load at 16c :cool:


not bad for a 156watt pelter and a waterblock..new project coming up for me in like a week is mineral oil and a freezer.. going for below zero temps and going to try to push the xp 2000+ or 1755mhz up to 1900mhz+ I will post specs and temps maybe some picture l8ter when I am done until then....



Hack the planet..."USA USA"STAND TOGETHER..

[Diggrr]

Well, as I said I would, I jumped into direct die cooling today.
I modified a sewer cap block from Tim Whittaker's design. Note the fancy dremel work to clear the L1 bridges and the pads.

While I had the mobo out, I also added a northbridge waterblock, and the KT7A voltage mod.

The nail polish idea is working fine, but I'd suggest covering your pencil marks with a small piece of clear tape before painting it. I lost my oc-ability so I'm gonna have to take it back apart one day this weekend, remove the paint and reseal it again (marine goop is 'da bomb, and is removable)

Here's a pic....

[The Overclocker]

Originally posted by Diggrr
Well, as I said I would, I jumped into direct die cooling today.
I modified a sewer cap block from Tim Whittaker's design. Note the fancy dremel work to clear the L1 bridges and the pads.

While I had the mobo out, I also added a northbridge waterblock, and the KT7A voltage mod.

The nail polish idea is working fine, but I'd suggest covering your pencil marks with a small piece of clear tape before painting it. I lost my oc-ability so I'm gonna have to take it back apart one day this weekend, remove the paint and reseal it again (marine goop is 'da bomb, and is removable)

Here's a pic....

wow, that is a nice block... i wish i could get thoses end cap things in the uk but all of them are flat. ah well, i will find something i am sure

[Frodo Baggins]

post ur results

[wolfsid]

Originally posted by Diggrr
Well, as I said I would, I jumped into direct die cooling today.
I modified a sewer cap block from Tim Whittaker's design. Note the fancy dremel work to clear the L1 bridges and the pads.

While I had the mobo out, I also added a northbridge waterblock, and the KT7A voltage mod.

The nail polish idea is working fine, but I'd suggest covering your pencil marks with a small piece of clear tape before painting it. I lost my oc-ability so I'm gonna have to take it back apart one day this weekend, remove the paint and reseal it again (marine goop is 'da bomb, and is removable)

Here's a pic....


man that looks nice but how did U seal it up around the processor it self, also how did U attach the cooler let see some more picture...

[Diggrr]

The block is glued to the processor with the modder's favorite: Marine Goop.

I made a hold down out of plexiglass (perspex for the Brits) so that pressure will hold the seal just in case. I used hardware from work, and it'll provide too much pressure...enough to bow the motherboard if not carefull.

As for cooling results, I'm running at 34C. That's about a 5C improvement to a similar block with 9mm copper on the bottom.
I've also added the northbridge cooler, but it cuts way down on the waterflow, so I'm guessing it may get better when that's fixed.

[Warlord2]

awesome mod

I would really like to see the results when compared to a maze2

[Diggrr]

As soon as someone sends me a maze 2, I'd be happy to supply a direct stack-up. :D

Actually, I'm broke right now, so the direct die cooler was the cheapest option-however radiacal.

[IFMU]

Ok where is this mod for the direct die cooling? I would also like to know did you use the paint polish on the actual core? or just the jumpers? (is that the right word???) thanks for the link if anyone can supply it...

[Diggrr]

http://www.spodesabode.com/view.php?pageid=directdie1
As posted on the first page of this thread.

I just gathered information and did it my own way.

Leufken tech. suggested using nail polish (enamel, non-metalic) to paint over the bridges and up to the core for their new block. The pic at the top of page two of this thread shows this. I painted right up onto the sides of the core, without puting any on the top.

I can't say that I got more out of direct die cooling than with the copper. I fixed my flow problem, but temps didn't change any. I've worked up to 36C which is 4C better than I had with the copper, but the nail polish killed my pencil marks, so I'm not overclocking right now. I guess that would about even-out in the end.

But ya sure can't beat the looks on my friends faces. They think I'm on crack or something, but it's a monkey of a different color, the MOD MONKEY!
And yup, that's a whirlpool in my res, and no it doesn't effect temps at all.

[IFMU]

Ok sorry bout the repost of the link... didnt realize that was the link to that story... But thanks none the less... Anf is there a way to get a bigger pic of that res? looks pretty kewl there...

[Diggrr]

As you wish.

The res is made out of 3.25 inch cast acrylic. I made the top with a hole for filling (by adding a tube and syringe). The bottom is solid. There is a second bottom inside with a single hole in the middle of it to keep the vortex going. The outlet to the pump is between the two bottoms. The inlet pipe is off center (tangental) to provide the vortex's swirl.

One thing I found was that the epoxy I used absorbs the green die in the antifreeze that was in it. This shows on the (inner) bottom. I now run almost pure distilled water with 10cc of antifreeze so it looks clear, and shows off the blue neon.

[Al666]

Una it's good to see you here.
Also thx for all your help with my RAIP (Redundant Array of Independent Pumps) idea over @ HardOCP forums. Unfortunately I couldn't finish it yet, due to lack of space in the Enermax 989AL-00 case (secondary Eheim 1048 didn't fit;). But I'm waiting for the Dual MPX motherboards that will be released soon, then I'm going to do the RAIP setup in my Addtronics7896A case.
I also loved your idea of using lower rated secondary pump for the radiator.
AL

BLACK ICE II w/ 2 screamin' 130CFM 12CM DELTAs
SWIFTECH MCW462
OCHIDEOUT Aluminum Reservoir w/ Universal 145GPH pump (the pump will be replaced with a better one soon)
DigitalDoc5
PCMODS Rheostat Baybus 17W per knob
XP 1600+ @1680 MHZ
KG7-RAID
3DLabs GVX1-Pro AGP and VX1 PCI
IBM Ultrastar 18ES 9.1GB Ultra2 LVD SCSI as the boot-disk
2X IBM Deskstar 60GXP 60GB as RAID-0
IBM Deskstar 75GXP 45GB in Macpower IceCube Firewire Enclosure
Enermax 989AL-00 w/431W PSU