I haven't and I don't intend too right now until I give it a few trys without. Back in the day IHS didn't exist and aftermarket and even factory provided coolers were direct contact.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Removed the IHS from an Ivy Bridge i5 3570k
- Thread starter Moogoos
- Start date
I haven't and I don't intend too right now until I give it a few trys without. Back in the day IHS didn't exist and aftermarket and even factory provided coolers were direct contact.
Agree. This is the goal.
Activate: AMD
[H]ard|Gawd
- Joined
- Nov 6, 2004
- Messages
- 1,994
yea that sounds like a brilliant idea.. i'm sure it would work fantastically!
Yea, but things change.. what worked well by then isn't guaranteed to work today. These chips are putting out more heat per unit area than the old chips, maybe the lid is really doing something worthwhile.
Possibly and I get that but there's only one way to find out and that is by testing various methods. I don't understand how die, to TIM, to IHS, to TIM, to HS can be better than direct but we'll see. Test. Test. Test. That's the plan.
Acer_Sheep
[H]ard|Gawd
- Joined
- May 18, 2007
- Messages
- 1,201
Can I do that with sandy bridge as well?
Not easily. Sandy Bridge is actually soldered down, which means you need razors+blowtorch to disassemble it, rather than just razors.
You wouldn't want to anyway. SB does not have the heat issues.
While you're at it you might as well remove the water from your loop too, because it won't be doing much if you don't use any TIM. And if you go that far, you should probably just remove the power cord from the back of the PSU to make sure that no electricity gets to the system. Your PC won't have much need for electricity if you don't use any TIM. And hell, why not just go all out and complete the process by removing the PC from your house entirely? You won't have much need for a PC that doesn't have any TIM between the CPU and the waterblock.
Last edited:
True, but good luck getting them that perfect. Even microscopic air bubbles will act as insulation. There's a reason people use TIM.
I think most people here know what TIM is used for, and if you think you're going to get the CPU die and the surface of the waterblock to be "perfectly mated" by lapping the block to a mirror finish...follow the directions in my post above.
Last edited:
Definitely go with the best TIM you can possibly get, otherwise you might be disappointed. The "bottleneck" is not the TIM+IHS itself, but the CPU trying to push all of it's power through a small area of bad TIM. The TIM you apply here will be far more important than the TIM you would otherwise apply on top of a larger area IHS.
Pixel Eater
Limp Gawd
- Joined
- Oct 25, 2006
- Messages
- 241
See, this was Intel's wisdom in switching to thermal paste.. yeah.....
Well I have had good luck with IC Diamond. It could be a good choice here.
Good luck with the mounting i hope it turns out well, For all of us.
Yea... I'm still waiting for someone to grab some KappZap fluxless sliver/tin solder and give that a try under the lid. Seems like solder would obviously have the best conductivity for heat transfer.
IC Diamond is pretty abrasive. I would not want to put it on a bare cpu.
Someone earlier mentioned that Intel preps the dies they use for soldered IHS's - and that the solder won't adhere to a non-prepped die (like these, presumably).
Hmmm... It was my understanding the dies were electroplated leaving a thin copper coating on the chip? If that's the case, it doesn't seem like it would be difficult to get the right kind of solder to hold. Of course my understanding of how the wafers are made could be completely off.
Just got home from work and The Home Depot. Scored some neoprene washers to help with adjusting the height of the mounting brackets relative to the placement of the heatsink on the die. This should help me get a better seat.
I removed the block and noticed that I looked quite liberal with the TIM so I am going to clean it down, again, and go with a very small dose. The Shin Etsu has a big outlet so I might need a toothpick to apply it on this little rectangle.
Last night I remember thinking I was spinning the screw that holds the heatsink down. These was from scrap parts from around the house so who knows what really fits. So when I took em apart I screwed them on manually and they go tight, no spinning so I may have been loose as well.
I can't remember where I got these screws but they're perfect for this hack job. Glad I toss all the screws and crap over the years into a box.
I am going to try and gauge the thickness of the IHS and then reduce that thickness from the overall height of the stock mount base. This should be calibrated with the correct level, I think. I didn't finish college.
Hope I don't f this up. Been thinkin' about it all day.
I removed the block and noticed that I looked quite liberal with the TIM so I am going to clean it down, again, and go with a very small dose. The Shin Etsu has a big outlet so I might need a toothpick to apply it on this little rectangle.
Last night I remember thinking I was spinning the screw that holds the heatsink down. These was from scrap parts from around the house so who knows what really fits. So when I took em apart I screwed them on manually and they go tight, no spinning so I may have been loose as well.
I can't remember where I got these screws but they're perfect for this hack job. Glad I toss all the screws and crap over the years into a box.
I am going to try and gauge the thickness of the IHS and then reduce that thickness from the overall height of the stock mount base. This should be calibrated with the correct level, I think. I didn't finish college.
Hope I don't f this up. Been thinkin' about it all day.
Last edited:
If results aren't favorable without the IHS I'll go ahead and try it with.
So here is the height of the heat spreader measured in washers.
3. I'll take one out to account for the gap that is under it, so we'll say the IHS is 2 washers thick.
The gap between the Intel mount and the AMD mount height is 1 neoprene washer and 4 metal washers. If I understand this part right then I should remove 2 washers from the stack and I should be back in alignment as if I had the IHS on.
3. I'll take one out to account for the gap that is under it, so we'll say the IHS is 2 washers thick.
The gap between the Intel mount and the AMD mount height is 1 neoprene washer and 4 metal washers. If I understand this part right then I should remove 2 washers from the stack and I should be back in alignment as if I had the IHS on.
Cerulean
[H]F Junkie
- Joined
- Jul 27, 2006
- Messages
- 9,476
Subscribed
Thermite Paste
Supreme [H]ardness
- Joined
- Dec 17, 2003
- Messages
- 5,650
Subscribed, looking forward to seeing your results. I plan on doing a full watercooling loop for my 3570k eventually, I'd love to do direct die if it helps temps.
I enjoy your applied mathematics as well... but maybe next time you should invest in a caliper
I enjoy your applied mathematics as well... but maybe next time you should invest in a caliper
As an Amazon Associate, HardForum may earn from qualifying purchases.
but maybe next time you should invest in a caliper
lol. You're totally right and I am so lazy. I have a caliper but I have no idea where it is. Use it to measure length of .45 reload casings.
As an Amazon Associate, HardForum may earn from qualifying purchases.
So I went with a little bit o TIM.
First time PC didn't post. I think 'cause I was too high so the block wasn't keeping the pressure on the chip to keep it seated so I took out one more washer.
Then I was able to seat it and the results are now inline with my original stock.
About 30º idle and 51º under load, stock 3570k. So far I am not convinced that Intel switching to TIM changes much in this CPU.
I need to take a break from all the tweaking, cleaning, reapplying and such and get my Diablo III on. I'll try it with the IHS in the next couple days.
First time PC didn't post. I think 'cause I was too high so the block wasn't keeping the pressure on the chip to keep it seated so I took out one more washer.
Then I was able to seat it and the results are now inline with my original stock.
About 30º idle and 51º under load, stock 3570k. So far I am not convinced that Intel switching to TIM changes much in this CPU.
I need to take a break from all the tweaking, cleaning, reapplying and such and get my Diablo III on. I'll try it with the IHS in the next couple days.
Last edited:
jojo69
[H]F Junkie
- Joined
- Sep 13, 2009
- Messages
- 11,267
physics fail
Thank you. The others seem to think they'd contribute by ripping me.
Thermite Paste
Supreme [H]ardness
- Joined
- Dec 17, 2003
- Messages
- 5,650
Have you tried overclocking again? IIRC the main problem with the IB chips was the exponential heat increase with a raised vcore, not the stock specs. Enjoy Diablo III first though
You may find that you will see better heat dissipation using a high quality TIM with the IHS, your going to run into thermal transfer limitations with regards to surface area using the naked die (one of the reasons the old naked Athlon's ran at higher temps). A nice die to IHS bond with some decent paste should give a much greater amount of effective surface area to draw the heat away from.
Crazy Chuckster
2[H]4U
- Joined
- Oct 6, 2001
- Messages
- 3,115
I lol'ed @ this a bit
I was around "back in the day" when cpus did not come with and ihs and to get better temps you slapped a 60mm delta finger ginsu on. As a matter of fact i have a few still floating around here....
I lol'ed @ this a bit
I was around "back in the day" when cpus did not come with and ihs and to get better temps you slapped a 60mm delta finger ginsu on. As a matter of fact i have a few still floating around here....
I've got one too. I'll never throw it away, it reminds me of the days I thought it was cool to have a computer that sounded like it was about to take off.
Random Person: "Why is it so loud?"
Me: "because it's fast!"
lmao.