![]() | ||
|
|
General Liquid/Water Cooling Discussion For discussion about Full Cooling System kits, or general cooling topics. Keep specific cooling items like pumps, radiators, etc... in their specific forums. |
![]() |
Thread Tools |
![]() |
#1 |
Cooling Neophyte
Join Date: Mar 2002
Posts: 79
|
![]()
Here's a sneak preview of a concept I have been tossing about in my head for some time, and that I've finally had the time to try out. Believe it or not, finding a 4mm drill bit here in singapore is not easy. Most hardware stores seem to only carry bits in imperial sizes. I needed a bit that was exactly 4mm, had to head to a specialty DIY store to get one (and pay DIY store prices for it!).
In any case, here's what I'm talking about: ![]() It's a clip-on RAMsink. That's right, only mechanical retention. No goddamned thermal epoxies or frag tape. ![]() As you can see this is just a proof-of-concept prototype. It's missing its fins, and it could also use being thicker. I'm currently thinking of how I can put fins into it with nothing but a drill press. I'm currently pondering two possible methods: 1) Use a jigsaw and try to carefully cut fins into it. 2) Use a drill press and try to carefull drill pins into it. (a la #Rotor) Input? ![]() The sink is fairly easy to clip on. Just put in position and tighten the nut. Wing-nuts may make this job even easier, at the risk of overtightening however. Since the clipping is only at one end there is more force on the outer as compared to the inner part of the ram chip. I am worried this may damage it if over-tightened. Also may lead to gaps at the inner (WTR to the card itself) end between the sink and the chip. I looked at it though, visual inspection seems to indicate good contact all the way down. ![]() Here you can see clearly how the ramsink is built. I got the basic idea from F-clamps. ![]() Here you can see the components. The two copper pieces obviously have not been finished yet. Like I mentioned earlier some kind of finning is required. I also think a thicker block of copper with a longer screw may be a good idea, to prevent the copper from bending due to the uneven pressure. I obviously don't have any performance results as this is prototype was only meant to assess the feasibility of the concept. In this respect it has been a success. What do you guys think? Comments/suggestions? |
![]() |
![]() |
![]() |
#2 |
Cooling Savant
Join Date: Feb 2002
Location: SWEDEN(cold...)
Posts: 195
|
![]()
nice idea, but wont they heat up quickly and have trouble to dissapate the heat?
__________________
www.webraced.com <- addicted as hell /me is! |
![]() |
![]() |
![]() |
#3 |
Cooling Neophyte
Join Date: Mar 2002
Posts: 79
|
![]()
Like I mentioned these still lack finning. I'm also pondering making them thicker. With finning they would pretty much be like any copper ramsink out there...
|
![]() |
![]() |
![]() |
#4 |
Big PlayerMaking Big Money
Join Date: Aug 2001
Location: irc.lostgeek.com #procooling.com
Posts: 4,782
|
![]()
Could you incorporate a spring on the bolt in between the two pieces? I would think you could even out the pressure somewhat like that.
|
![]() |
![]() |
![]() |
#5 |
Cooling Savant
Join Date: Jan 2002
Location: Waukesha, Wi
Posts: 698
|
![]()
problem i see with the clamp is that all of the pressure is on one side. i could see the blocks actually lifting off the farther side of the clamp.
|
![]() |
![]() |
![]() |
#6 |
Cooling Savant
Join Date: Aug 2001
Location: KS
Posts: 374
|
![]()
Yes, but pHaestus has a good point, a spring on the bolt would help some by taking up the slack. Also, moving the screw as close as possible to the card. I would have the screw actually touching the card. A little AS2 or 3 and you would be set. You dont have to crank the hell out of the screws to get a good connection.
__________________
MeltMan Lurker Supreme! |
![]() |
![]() |
![]() |
#7 |
Cooling Savant
Join Date: Jan 2002
Location: Waukesha, Wi
Posts: 698
|
![]()
perhaps a spring in there on the other side of the bolt (opposite the card) would be good, instead of around the bolt. if it was around the bolt i would think that the symmetrical pressure would cancel out any possible counter pressure that you are needing.
|
![]() |
![]() |
![]() |
#8 |
Cooling Savant
Join Date: Feb 2002
Location: Dark Side of the Moon
Posts: 365
|
![]()
The combination of the bolt very close to the card and the spring on the far end of the clamp would be about perfect if the spring had a high enough tension. After all, you aren't wanting enough pressure on the RAM to crack it
![]() Curious though, what's your beef with using an Arctic Alumina mix? Holds well and can be removed if done with the epoxy based version used only on a small section of each end and the non-epoxy version used in the middle. |
![]() |
![]() |
![]() |
#9 |
Cooling Savant
Join Date: Aug 2001
Location: KS
Posts: 374
|
![]()
I think someone also suggested (if you dont want to shell out the dough for AS epoxy, a 2/8 mix of superglue and AS2 <- I think it was that way.. not the other way around... but heck if i know..
__________________
MeltMan Lurker Supreme! |
![]() |
![]() |
![]() |
#10 |
Put up or Shut Up
Join Date: Dec 2001
Location: Spokane WA
Posts: 6,506
|
![]()
I can think of one better way to do this!!! But it would require some work. Instead of having the bolt on the outside where the other end of the clamp will pull apart, have the nut on the inside to where when you turn it it actually created down pressure on the opposite end and will squeez together.
|
![]() |
![]() |
![]() |
#11 |
Put up or Shut Up
Join Date: Dec 2001
Location: Spokane WA
Posts: 6,506
|
![]()
Pic of what I am saying. The red are nuts. The inside ones will tighten and press the other end of the clamp downward to create a nice even pressure on the other side.
![]() |
![]() |
![]() |
![]() |
#12 | |
Cooling Neophyte
Join Date: Mar 2002
Posts: 79
|
![]() Quote:
![]() I had a similar idea. Instead of putting a spring on the opposite side, put a thin bolt with two nuts inside instead of outside it. Like that, by cranking the two nuts, you can get leverage that will force the entire setup to stay parallel. What do you think? |
|
![]() |
![]() |
![]() |
#13 |
Thermophile
Join Date: Oct 2001
Location: Nuu Zeeelin
Posts: 3,175
|
![]()
why not use a thin piece of copper to attach to the ram, then epoxy to this copper whatever heatsinks you want, it will cool better, look better, and be easier
__________________
2x P3 1100's at 1400, Abit VP6, 2x Corsair 256mb PC150 sticks, 20gb 'cuda ATA-III, 2x 40gb 'cuda ATA-IV in raid 0. 20" Trinitron. No fans 2x 2400+ at 2288mhz (16.0 x 143), Iwill MPX2, 2x Kingmax PC-3200 256mb sticks, 4x 20gb 60gxp in Raid 5 on a Promise SX6000. Asus Ti4200 320/630. Cooled by Water |
![]() |
![]() |
![]() |
#14 |
Cooling Savant
Join Date: Aug 2001
Location: KS
Posts: 374
|
![]()
Yeah, but I never listened to "keep it simple stupid". Hehe. If it were up to me, I would just glue the darn things.
__________________
MeltMan Lurker Supreme! |
![]() |
![]() |
![]() |
#15 | |
Cooling Neophyte
Join Date: Mar 2002
Posts: 79
|
![]() Quote:
|
|
![]() |
![]() |
![]() |
#16 |
Thermophile
Join Date: Oct 2001
Location: Nuu Zeeelin
Posts: 3,175
|
![]()
but with the heatsinks on, I doubt it would bend one little bit.
__________________
2x P3 1100's at 1400, Abit VP6, 2x Corsair 256mb PC150 sticks, 20gb 'cuda ATA-III, 2x 40gb 'cuda ATA-IV in raid 0. 20" Trinitron. No fans 2x 2400+ at 2288mhz (16.0 x 143), Iwill MPX2, 2x Kingmax PC-3200 256mb sticks, 4x 20gb 60gxp in Raid 5 on a Promise SX6000. Asus Ti4200 320/630. Cooled by Water |
![]() |
![]() |
![]() |
#17 |
Cooling Neophyte
Join Date: Mar 2002
Posts: 79
|
![]()
Average heatsink is not that long, you'd bend at the interface point probably.
|
![]() |
![]() |
![]() |
#18 |
Put up or Shut Up
Join Date: Dec 2001
Location: Spokane WA
Posts: 6,506
|
![]()
I have some 1/8" copper stock sitting right here that I use for the tops of my copper WB's. It will not bend in the short amount, but you do not have much area for fins unless you soldered some on.
|
![]() |
![]() |
![]() |
#19 | |
Cooling Neophyte
Join Date: Mar 2002
Posts: 79
|
![]() Quote:
![]() IMHO soldering fins on is not a good idea. Unreliable thermal interface at the solder joint. |
|
![]() |
![]() |
![]() |
#20 |
Thermophile
Join Date: Oct 2001
Location: Nuu Zeeelin
Posts: 3,175
|
![]()
soldering can't be that bad, remember how well the themalright sk6 performs? Some heatsinks with soldered fins have performed really poorly, but the sk6 was consistantly good
__________________
2x P3 1100's at 1400, Abit VP6, 2x Corsair 256mb PC150 sticks, 20gb 'cuda ATA-III, 2x 40gb 'cuda ATA-IV in raid 0. 20" Trinitron. No fans 2x 2400+ at 2288mhz (16.0 x 143), Iwill MPX2, 2x Kingmax PC-3200 256mb sticks, 4x 20gb 60gxp in Raid 5 on a Promise SX6000. Asus Ti4200 320/630. Cooled by Water |
![]() |
![]() |
![]() |
#21 |
Put up or Shut Up
Join Date: Dec 2001
Location: Spokane WA
Posts: 6,506
|
![]()
You need to use silver based solder for good transfer. Also note ram does not need much extra cooling to begine with. Copper is really overkill. If you put enough pressure on 1/8" copper to bend it you will damage the memory chip!!! They are very touchy.
|
![]() |
![]() |
![]() |
#22 | |
Cooling Neophyte
Join Date: Mar 2002
Posts: 79
|
![]() Quote:
As for the SK6, they are produced industrially in a factory, not by an amateur on a table in his bedroom, and the former tends to produce better quality products if only for the better equipment present. ![]() |
|
![]() |
![]() |
![]() |
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
|
|