IB De-Lidding Temps before and after

[Ed_1]

are you using negative offset to get 1.15v or your just using fixed . I see your lower now than original 4.5ghz runs .

 

[BlackOmega]

SonDa5, as you were eluding to, I think, at this point, your TIM is the weakest link.

IC Diamond is an awesome TIM (I was part of the test bed for it at OCN before it's initial release and tested it on several rigs). However, it's only good, in my opinion, for CPUs' with non-modified IHS's.
I did a few test approximately 2 -3 years ago with some delidded 939's and a lapped AM3 CPU.
On all the tests I used a lapped Thermaltake big typhoon (lapped to 2000grit then polished out) and here's what I found:

The smoother and flatter the two surfaces are, the less TIM you need, i.e. lower viscosity (thinner). Thicker TIM's such as IC diamond, AS5, OCZ freeze, Tuniq TX-2, etc. are too thick and actually start working as an insulator rather than a conductor.
On both the delidded 939's and the lapped AM3, I found that if I just used the white "pigeon poop" TIM, it actually worked A LOT better than the high dollar TIM's. Reason being is due to it's low viscosity, the thin TIM will spread more evenly and there'll only be enough there to take up what small gaps there are between the two surfaces.

I was seriously thinking of just putting a thin coat of motor oil between the lapped AM3 and the BT just to see what the difference would be.

On a side note, the thinner TIM's also work A LOT better than the thick silver TIM's on GPUs as well. I swapped out the TIM on several "hot" ATI cards (4870's, 5870's, 6870's) and noticed the cooling was a good deal better than before.
Most notably, idle and load temps on the 4870's dropped 13C across the board. All that just from swapping out the TIM.

 

[SonDa5]

SonDa5, as you were eluding to, I think, at this point, your TIM is the weakest link.

IC Diamond is an awesome TIM (I was part of the test bed for it at OCN before it's initial release and tested it on several rigs). However, it's only good, in my opinion, for CPUs' with non-modified IHS's.
I did a few test approximately 2 -3 years ago with some delidded 939's and a lapped AM3 CPU.
On all the tests I used a lapped Thermaltake big typhoon (lapped to 2000grit then polished out) and here's what I found:

The smoother and flatter the two surfaces are, the less TIM you need, i.e. lower viscosity (thinner). Thicker TIM's such as IC diamond, AS5, OCZ freeze, Tuniq TX-2, etc. are too thick and actually start working as an insulator rather than a conductor.
On both the delidded 939's and the lapped AM3, I found that if I just used the white "pigeon poop" TIM, it actually worked A LOT better than the high dollar TIM's. Reason being is due to it's low viscosity, the thin TIM will spread more evenly and there'll only be enough there to take up what small gaps there are between the two surfaces.

I was seriously thinking of just putting a thin coat of motor oil between the lapped AM3 and the BT just to see what the difference would be.

On a side note, the thinner TIM's also work A LOT better than the thick silver TIM's on GPUs as well. I swapped out the TIM on several "hot" ATI cards (4870's, 5870's, 6870's) and noticed the cooling was a good deal better than before.
Most notably, idle and load temps on the 4870's dropped 13C across the board. All that just from swapping out the TIM.

IC Diamond is working out really well for me but I bought some Liquid Metal Pro to try it out. With IC Dimoand I have been able to lower average core temps by around 20C per core.

Ed_1 I am using an MSI Z77 Mpower MB which doesn' have LLC, negative offsetting in the BIOS. For me it is all trial and error. For the most part the MB sticks well with the voltage selected in BIOS but under extreme load there is a slight voltage drop. For 1.145vots in the bios it dips under Intle Burn test load to no lower than 1.13v at 4.5GHZ.

 

[cdawall]

how thefuck

?

Well if somebody were to plan on de-lidding/removing fluxless solder I think there should be a planned TIM to be used. Makes no sense to destroy the factory IHS design if an improved method of cooling isn't possible.

Not all CPUs would benefit from de-lidding but IB does because of Intel's poor IHS design.

Most CPU's benefit from delidding. AMD chips do as well and since most have been soldered people have been sanding through them for a while. The trick is to wrap the sandpaper around a small piece of glass to make sure you are sanding nice and level.

 

[TheoneandonlyMrK]

The trick is to wrap the sandpaper around a small piece of glass to make sure you are sanding nice and level.

isnt that reverse lappingand quite dangerous anyway, put the sheet os sand paper on a piece of glass and rub the cpu on it, id advise more of a guide then that tho

 

[BlackOmega]

IC Diamond is working out really well for me but I bought some Liquid Metal Pro to try it out. With IC Dimoand I have been able to lower average core temps by around 20C per core.

IC Diamond is great stuff (great for polishing after too). There's definitely a break-in period with it, however it's not nearly as long as some others (AS5 is 200 hours!).

I was merely suggesting to try the cheapo white TIM, from my testing it has shown to conduct better with smoother flatter surfaces. Couldn't hurt, just give it a try.

 

[SonDa5]

Delidded IB i5-3570k @ 4.5GHZ 1.15v with NO IHS with DT SNIPER Water block
http://valid.canardpc.com/show_oc.php?id=2551498

Loop configuration as tested. Single 120mm slim HWLABS Black ICE radiator.

TIM Liquid Pro
IBT 4.5GHZ Maximum memory tested 10 passes
Ambient Temps 25C
Fresh Install

With IC Diamond 24 average temp of each core was 73C under IBT load. With Liquid Pro average temp of each core is at 61C!! under IBT load.

Taking a look at my IB 3570k progress since delidding.

Before Delidding.

Ambient Temp 26C

TIM IC Diamond 24

Temps under LinX load 83C 93C 89C 84C average of 87.25C per core.


Since delidding I have been able to lower my average core temps under IBT 100% load at 4.5GHZ by 26.25C per core!!!

 

[eidairaman1]

cpus are a lil tough for me to hold onto actually, i might have a better luck with an orbital sander

 

[eidairaman1]

Would you prefer to have a chip that could be delidded, or a chip that had fluxless solder? I personally would go for the fluxless solder because I don't want to take the risk.

that was the idea behind the IHS, reduce core damage

 

[Nordic]

that was the idea behind the IHS, reduce core damage

I know. I still find that to be a illegitimate question for all overclockers. If intel/amd wanted they could sell highend cpu's with a TIM instead of FS all the time.

 

[eidairaman1]

I know. I still find that to be a illegitimate question for all overclockers. If intel/amd wanted they could sell highend cpu's with a TIM instead of FS all the time.

careful you might give them an idea

 

[Peter1986C]

SonDa5, as you were eluding to, I think, at this point, your TIM is the weakest link.

IC Diamond is an awesome TIM (I was part of the test bed for it at OCN before it's initial release and tested it on several rigs). However, it's only good, in my opinion, for CPUs' with non-modified IHS's.
I did a few test approximately 2 -3 years ago with some delidded 939's and a lapped AM3 CPU.
On all the tests I used a lapped Thermaltake big typhoon (lapped to 2000grit then polished out) and here's what I found:

The smoother and flatter the two surfaces are, the less TIM you need, i.e. lower viscosity (thinner). Thicker TIM's such as IC diamond, AS5, OCZ freeze, Tuniq TX-2, etc. are too thick and actually start working as an insulator rather than a conductor.
On both the delidded 939's and the lapped AM3, I found that if I just used the white "pigeon poop" TIM, it actually worked A LOT better than the high dollar TIM's. Reason being is due to it's low viscosity, the thin TIM will spread more evenly and there'll only be enough there to take up what small gaps there are between the two surfaces.

I was seriously thinking of just putting a thin coat of motor oil between the lapped AM3 and the BT just to see what the difference would be.

On a side note, the thinner TIM's also work A LOT better than the thick silver TIM's on GPUs as well. I swapped out the TIM on several "hot" ATI cards (4870's, 5870's, 6870's) and noticed the cooling was a good deal better than before.
Most notably, idle and load temps on the 4870's dropped 13C across the board. All that just from swapping out the TIM.

So that would mean that in case of not only de-lidded CPUs, but as well on laptop CPUs, GPUs and chipsets; something like Coollink Chillaramic (10 gram for about 6 euro/7dollars) would be better than Thermalright Chill Factor III, Arctic Silver 5, Antec Formula 7, IC Diamond etc.? Good to know.

 

[t_ski]

SonDa5, did you ever consider something like this?

http://www.xoxide.com/specialthermalpad-chipsetcooling.html

It's meant for chipsets, but it was supposed to add a little bit of support and help balance the heatsink so you wouldn't crack the edges. You might have to trim it up and make the hole a little bigger inside, but I wonder if it would help.

 

[SonDa5]

SonDa5, did you ever consider something like this?

http://www.xoxide.com/specialthermalpad-chipsetcooling.html

It's meant for chipsets, but it was supposed to add a little bit of support and help balance the heatsink so you wouldn't crack the edges. You might have to trim it up and make the hole a little bigger inside, but I wonder if it would help.

I did think of something like that but my idea was to add a 1mm thick thermal pad around the die. I actually just tried it with the Liquid Pro TIM but it was a little too thick and prevented die from making contact with block. I tried to thin it down but couldn't manage to get a good mount with it and my CPU would not post. Got an error 55. After about 3 on and off mounts with the error 55 I took off the thermal pad and then was able to post but still didn't have a good mount. Took about 5 or 6 mounts to get a good mount so now I am thinking about leaving it alone. I would like to try the Thermal Pad again but it didn't work out the first few attempts and mounting bare die is kind of tricky to get the perfect fit.

This is what my DIE looked like with the Liquid Pro and Thermal Pad around it.

This is what I used: Fujipoly Extreme System Builder Thermal Pad - 60 x 50 x 1.0 - Thermal Conductivity 11.0 W/mK

FrozenCPU sells it.

http://www.frozencpu.com/products/1...rmal_Conductivity_110_WmK.html?tl=g8c487s1294

 

[t_ski]

I saw that mentioned earlier, which is why I posted the link. This is actually a foam pad, which is going to compress a little as needed. I honestly don't know what the error you had referred to, but if it's due to lack of contact your pad may have been too thick.

 

[DaedalusHelios]

woah. At 1.175 V, my chip hits ~65-70c, under an H100, similar clocks. You have a hot one, yep, but that explains why it's a 3570K, and not a 3770K, like mine is.

Ambient temp and ventilation in your room makes a big difference too.

 

[eidairaman1]

I did think of something like that but my idea was to add a 1mm thick thermal pad around the die. I actually just tried it with the Liquid Pro TIM but it was a little too thick and prevented die from making contact with block. I tried to thin it down but couldn't manage to get a good mount with it and my CPU would not post. Got an error 55. After about 3 on and off mounts with the error 55 I took off the thermal pad and then was able to post but still didn't have a good mount. Took about 5 or 6 mounts to get a good mount so now I am thinking about leaving it alone. I would like to try the Thermal Pad again but it didn't work out the first few attempts and mounting bare die is kind of tricky to get the perfect fit.

This is what my DIE looked like with the Liquid Pro and Thermal Pad around it.

http://minidriven.com/BlacknBlueforIB/DeLiD/3570kwithTParoundDIE.jpg

This is what I used: Fujipoly Extreme System Builder Thermal Pad - 60 x 50 x 1.0 - Thermal Conductivity 11.0 W/mK

FrozenCPU sells it.

http://www.frozencpu.com/products/1...rmal_Conductivity_110_WmK.html?tl=g8c487s1294

akasa shinetsu

 

[SonDa5]

akasa shinetsu

The Fujipoly is rated for better thermal heat transfer.

 

[eidairaman1]

The Fujipoly is rated for better thermal heat transfer.

reason I used them is because of the stackable factor of them.

I mean you were griping about the code 55 right, just giving as a solution.

 

[SonDa5]

reason I used them is because of the stackable factor of them.

I mean you were griping about the code 55 right, just giving as a solution.

Not really griping just sharing my experience.

The Fujipoly comes in different thickness sizes and could be stacked but I'd rather just use proper single layer size to begin with.

Big thing about my choice for using Fujipoly is the high heat transfer rating and pricing and availability at Frozen.cpu

With the Fujipoly 1mm size in hand I think I could get a bakers roller and roll it down to the thickness that works best.


Right now I think my temps are great so continuing to experiment with the Thermal Pad layer around the PCB does interest me but it just takes time.

I will continue to experiment when I get time and share. I hope I don't screw up my mount performance that I have right now.

I think this is full specification data on the fujipoly thermal pad I am using.

http://www.fujipoly.com/usa/product...-filler-pads/gap-filler-pads-putty/xr-pe.html

 

[SonDa5]

Just got done tweaking my delidded 3570k NO IHS mount.

I added Fujipoly Extreme .5 mm in thickness around the die.

DT SNIPER nice and snug.

Added some lock washers to my bolts on the front side of the mother board to keep my DT SNIPER mount bolts nice and firm when I put it back in the case. Also I am using a EK S115X TRUE back plate which works out great for this mount.

Still haven't booted. Once I get my loop back together I will see if it boots. I tried the Fujiploy Extreme last week with 1 mm thickness and the CPU would not post. So hopefully .5mm thickness on the sides will work out better.

 

[eidairaman1]

hey man keep it up

 

[DaedalusHelios]

That is too dangerous for me to be honest. I have crushed a core before on a northbridge from over tightening and I felt horrible. It wasn't about the money of course. I just hate to break things.

 

[SonDa5]

That is too dangerous for me to be honest. I have crushed a core before on a northbridge from over tightening and I felt horrible. It wasn't about the money of course. I just hate to break things.

Adding just the right thickness of thermal pad around the die area helps equalize and stabilize the block pressure. Also with Coollaboratory Liquid Pro high pressure mount isn't needed. Just soft even pressure. My block is gently hugging the die.

 

[mediasorcerer]

I admire your innovation, hows it running temps wise after the pads in place etc?