e4300 IHS Removed!

[DementeD]

Finally i can post..well after reading this thread for the past couple weeks i went ahead and removed my IHS from my e4300..and well it went really easy took less then 10 mins ..however i ran into the problem of my waterblock(even with retention bracket removed) wouldnt make good enough contact with the cpu so my temps actually rose..the waterblock hit the outside edge of the socket and wouldnt lay properly but i reinstalled my TT blue orb and the temps did drop 10c at load ..it maxes at 57c 3.2ghz 1.44volts(i live in florida so its like 80-90F when i was testing)..im pretty upset i did this because i had 3.4ghz on water stable and i can get it stable on air during the night but not throughout the day its just too hott and itll only get hotter..and i know if i get a different waterblock(or kit altogether its a cheap TT waterkit) that id get a lot better performance..

[I)ickie]

Finally i can post..well after reading this thread for the past couple weeks i went ahead and removed my IHS from my e4300..and well it went really easy took less then 10 mins ..however i ran into the problem of my waterblock(even with retention bracket removed) wouldnt make good enough contact with the cpu so my temps actually rose..the waterblock hit the outside edge of the socket and wouldnt lay properly but i reinstalled my TT blue orb and the temps did drop 10c at load ..it maxes at 57c 3.2ghz 1.44volts(i live in florida so its like 80-90F when i was testing)..im pretty upset i did this because i had 3.4ghz on water stable and i can get it stable on air during the night but not throughout the day its just too hott and itll only get hotter..and i know if i get a different waterblock(or kit altogether its a cheap TT waterkit) that id get a lot better performance..

Can anyone else comment on this. Both my blocks overlap the sides too! perhaps lapping the heatsink and then applying AS5 to the core after removing IHS then sticking IHS back on is the best way.

[menlatin]

You know, there are some older threads that have ALOT of info on this. do a search for "removed IHS". Pics of e6400's, celerons and preslers, and info on mounting.

[DementeD]

also when i mounted my i just used the waterblock to hold it down..so there might be a different way to mount it where contact will be made evenly..that was the main problem..the waterblock sat on it unevenly and caused it to get hotter..

[splat00n]

you think I can do this with the e6300?

[//\\//3W]

Nice.

Is there any other people that tryed this?

Found this:
http://www.legitreviews.com/article/402/2/

[N/A Matt]

I saw some one on here posted about removing the IHS that had solder, but they didn't seem to have an adverse effects. I can't see how simply removing the IHS would damage the chip, it would have to be mounting the heatsink and how you go about that. It was already mentioned, but even pressure is applied, i.e. screws tightened in the right order, there should not be an issue.

[Liquid3D]

One way to safely remove the IHS is to take a piece of very strong string or fishing line. Then find the corner wher there is a slight gap in the seal and begin "sawing" back and forth. This will even work with Epoxy IF your very careful and the line is high tension.

I've also heard placing the chip in the freezer not "boiling" it hehe works. The epoxy used is heat resistant so heat won't make the job easier but cold may.

I used to "pop the top" on all my CPU's but it only gave me 3C ~ 8C. Your better off getting water-cooling which to a Alendale or Conroe is analagous going to phase-change with an old P4.

Your temp drop has me thinking of trying this on my E6400 which is water-cooled and waterblocks are infinately adjustable when you tighten them down so there's really no chance of cracking a core if your careful.

Great job, great thread!!!

[Liquid3D]

also when i mounted my i just used the waterblock to hold it down..so there might be a different way to mount it where contact will be made evenly..that was the main problem..the waterblock sat on it unevenly and caused it to get hotter..

What type of waterblock are you using?

All you need do is find a block with a "minimal" footprint. European blocks follow this design and are usually use less mass. This works better anyway as h more mass that sits atop the CPU the more it is effecetd by the internal cae temp.

With a water block such as the new Sub-Zero by Cool-cases the footprint is minimal and with less meterial between the water and the CPU (base-plate is very thin) you getting the lower temp of the H20 closer to the heat source.

Can you show some pictures I may be able to help.

[g0dM@n]

One way to safely remove the IHS is to take a piece of very strong string or fishing line. Then find the corner wher there is a slight gap in the seal and begin "sawing" back and forth. This will even work with Epoxy IF your very careful and the line is high tension.

I've also heard placing the chip in the freezer not "boiling" it hehe works. The epoxy used is heat resistant so heat won't make the job easier but cold may.

I used to "pop the top" on all my CPU's but it only gave me 3C ~ 8C. Your better off getting water-cooling which to a Alendale or Conroe is analagous going to phase-change with an old P4.

Your temp drop has me thinking of trying this on my E6400 which is water-cooled and waterblocks are infinately adjustable when you tighten them down so there's really no chance of cracking a core if your careful.

Great job, great thread!!!

I have an e6400 and really wanna pop the pop if it's something not too hard to do...

[ziddey]

Your better off getting water-cooling which to a Alendale or Conroe is analagous going to phase-change with an old P4.

!?!!?!??!?!?!

[Nightcreeper]

I think he means the Allendale or conroe produces alot less heat, (Core2 based) as compared to the Netburst architecture that you could almost use for a furnace. That is why the moderate to high-end water cooling would allow the same OC % on a Core2 base as the Phase Change on a netburst. Core 2 Duo @ 2.66GHz is 65W (E6700) and the Netburst @ 2.66GHz ( Pentium D 805) is 95W, and the Core2 Outperforms the netburst all the way around.

No puns towards netburst, I have Dual Dempsey's (and one processor has an overheating problem) that I like for the price/performance ratio right now.

[ziddey]

with ambient temps the way they are, I can do the same with air as water.

[Liquid3D]

!?!!?!??!?!?!

I compared cooling low TDP chips on water to cooling something like Prescott on a entry level Phase change system. It was sort of in jest, but all humor makes light of a real situation.

Lets say you overclock a Prescott 950 from it's default settings (1.4V 130W TDP 3.4GHz) to the following 4.4GHz 1.5Vcore, now your TDP is 193W almost 200W

Then you overclock a Conroe E6400 from it's default settings (1.2V 65W TDP 2.1GHz) to the following 3.1GHz, 1.25Vcore, now your TDP is 100W.

If you ran the Prescott as above in a Entry level phase change under full load your temps would be around 0C ~ 10C

If you ran the Conroe as above in a H20 system under full load your temps would be around 33C ~ 40C depending on the system.

I know it's a bit exaggerated, I was trying to raise your sprirts. I know this, since converting to H20 I would never go back to air-cooled. I usually keep dle temps around 28C, which is not extraordinary when yout think about it. Conroe is just 65W, if you mount a dual heater core outside your case in ambeint environment then get a good waterblock TDX, Apex Ultra, Alphacool, Cool-Cases, etc. So my ambeint room temp is 17C / 62F (I like cold but 62F isn't that cold when you think of it) therefore the water entering the block is very close to ambient say 19C (remember dual radiator) and water exiting the block is about 10C higher at 29C interestingly enough this is close to what the CPU temp might be;

[Liquid3D]

I have an e6400 and really wanna pop the pop if it's something not too hard to do...

I advise against this. Once I heard the word epoxy I stopped messing around for two reasons.

1. damage
2. the concern has always been surface contact. I removed many an IHS on single core processors and found obvious problems, Dual Core is another story as the surfce area has a more complex architecture beneath it. In that case an IHS is a good thing because it's epoxied on there. The following statement is undeniably true: if you could solder your IHS or exposed core to your baseplate (albeit a water-block or air-cooler) you would get better temps then removing th IHS and using thermal paste.

When we removed Prescott IHS we found (as expected) a low quality thermal paste (Compared to AS5) and in some cases there was so much paste it was akin to throwing a blanket over the core because it acted as an insulator. Once that was cleaned yes there was better temps, but not in every case because some people had not prepared the surface correctly or the IHS made better contact then the core to the baseplate.

My point is this, it's not worth losing your warranty. I would bet anyone I could take their system and for the same amount of money ($0) improve the cooling without going to that measure. In most cases just removing your dam radiator and washing 3-months of dust out of it will drop temps 10C. I know it's not as fun as playing with tha CPU but it's necessary. Removing your IHS is optional.

Anyay as you can see below the processor is being stress tested at 100% LOAD using S&M, cooled by Swiftech's APEX Ultra system Dual Core radiator my E6400 running at 8x405FSB on the P5W DH 1.30Vcore and only 37C, actually CoreTemp s probably more accurate so we'll say 40C still pretty da goo when some people are at 60C at these speeds under LOAD some even at IDLE!