[Build Log] - Phantom Skies

[GearingMass]

Update: Prime95 Blend Results

The following screenshot is after 20 minutes of Prime95 Blend at 4.4Ghz - 1.200V
Ambient: 24C

 

[ATMINSIDE]

Well, you have a delidding candidate right there... If you want to that is.

 

[GearingMass]

Well, you have a delidding candidate right there... If you want to that is.

Haha Darnit. That is not what I want to hear or do.

 

[GearingMass]

Could it just be a bad TIM application, or are the temps too far gone for that scenario?
It was my first time applying TIM, and I may have not done it perfectly...

 

[ATMINSIDE]

It's possible that it's TIM application.

What TIM did you use and how much?

 

[GearingMass]

It's possible that it's TIM application.

What TIM did you use and how much?

Prolimatech PK-3, and a (probably too long) grain of "rice"

 

[GearingMass]

I'm a pretty careful guy, but delidding still scares me. I really would not want to brick a $300 CPU.

 

[ATMINSIDE]

Prolimatech PK-3, and a (probably too long) grain of "rice"

Go ahead and pull the CPU block (YAY FOR QDC's!!!) and take a picture of the TIM application.

Clean it all up, put a small dot in the middle, and re-mount the block.

About like this size:


In an ideal application that dot will push out to the edges of the IHS, but not quite to the corners.

 

[GearingMass]

Go ahead and pull the CPU block (YAY FOR QDC's!!!) and take a picture of the TIM application.

Clean it all up, put a small dot in the middle, and re-mount the block.

About like this size:


In an ideal application that dot will push out to the edges of the IHS, but not quite to the corners.

Oooohh yeah, I definitely put too much, I think that's what it was (hopefully). Mine was probably almost the length of the cpu die underneath the IHS. Oops.

I had no idea such a small dot would suffice. I bet I have way too thick of a TIM layer.

Yes, yay for QDC's!! I'll try and do that tomorrow if I have the time. Thanks (as usual) for the advice

 

[ATMINSIDE]

Oooohh yeah, I definitely put too much, I think that's what it was (hopefully). Mine was probably almost the length of the cpu die underneath the IHS. Oops.

I had no idea such a small dot would suffice. I bet I have way too thick of a TIM layer.

Yes, yay for QDC's!! I'll try and do that tomorrow if I have the time. Thanks (as usual) for the advice

Yeah, too much TIM will actually hinder the conduction of heat compared to a good application.

It doesn't actually have to cover the full IHS.
The pressure from the block (tighten in a star pattern) will spread the TIM evenly and thinly.

(I would do a mini leak test after popping everything back together, just to make sure you didn't jar anything pulling the QDCs out)

 

[GearingMass]

For the record, what should the blend temps look like with this kind of cooling and clock/volt settings?

 

[ATMINSIDE]

Take a look through this thread, there's a lot of results posted (look for rge in specific) with custom loops.
All P95 testing, various speeds/clocks, pre/post delid.

 

[GearingMass]

Take a look through this thread, there's a lot of results posted (look for rge in specific) with custom loops.
All P95 testing, various speeds/clocks, pre/post delid.

Yeah, I'm still reading through the whole thing, but I just ran 20 minutes of Large ffts for comparison, and had max temps of 78-85C.

I'm almost positive it's just a bad TIM application. Once that's shored up, then I'll be able to see how well of a chip I actually got.

 

[GearingMass]

Ok so potential problem here.
Looked closer at the res again today and found the water to be murky - not good.

I initially put 3-4 drops of PTN in it as well as a silver coil, so I have no idea how/if something is growing in my loop and whether or not I need to start replacing the water.

Pics below - advice welcome.











Edit: Do you think I could just syphon out the water and replace it with new distilled and re-add the PTN, or could gunk have already started to grow in the blocks, and if so, how is that even possible after two weeks with two types of biocide…

 

[GearingMass]

Yeah, too much TIM will actually hinder the conduction of heat compared to a good application.

It doesn't actually have to cover the full IHS.
The pressure from the block (tighten in a star pattern) will spread the TIM evenly and thinly.

Definitely put way too much on. Haha