Worse temps after Delid?

[227167_1454181677]

Hello everyone! 

I'm no stranger to water cooling, but I've always noticed my 4790k has run quite hot, even after delid. The real joke is the fact that the thing is under-volted in the bios as I wanted my system to be silent... Well, it's not. 

I have 2 360mm rads, all for cooling a GTX 1080ti and my CPU... That's it. I mean dare I say I have so much rad, I should practically be able to turn off half these fans and I should still be golden. Well, I just got done playing some Gears of War, checked out my temps as I heard the fans spinning up and was surprised to see the hottest core sitting at 89... My GPU is fine with temps, I can feel the pump running, so I know that's not an issue... I actually ran liquid metal just a few months ago, but I was never impressed with the temps with that either. After 3 applications, I was convinced that I simply sucked at applying the exotic 'paste' and this time around after doing maintenance on my loop, decided to simply throw normal paste on both the die, and the heat-spreader when I put it all back together.

I just.... don't understand how these temps are this bad... Has anyone suffered after de-lidding their CPU? Is it possible there's an air gap between die and heat-spreader now that I've separated them? I'm lost for words...

Pic, not that it'll help you or anything, but there she is. 
 

[CapitalistVN]

Have you checked that the thermal paste between the cooler and the heat-spreader has been applied appropriately? That might cause serious temp issues.

[227167_1454181677]

Just now, CapitalistVN said:

Have you checked that the thermal paste between the cooler and the heat-spreader has been applied appropriately? That might cause serious temp issues.

I really think I'm good in that dept. More than the typical pee shape, but as Linus' video has showed, you can practically put mountains on and it shouldn't suffer that much... I've literally taken all blocks apart and gone through etched channels in the copper, so I know there's no blockage or anything... I mean at this point, I may just have to reapply the past again for like... the 4th or 5th time if I count the liquid metal process that also wasn't good... It's just bizarre. It seems like something must be catastrophically wrong for me to hit 89 with so much rad. 

[CapitalistVN]

2 minutes ago, 227167_1454181677 said:

I really think I'm good in that dept. More than the typical pee shape, but as Linus' video has showed, you can practically put mountains on and it shouldn't suffer that much... I've literally taken all blocks apart and gone through etched channels in the copper, so I know there's no blockage or anything... I mean at this point, I may just have to reapply the past again for like... the 4th or 5th time if I count the liquid metal process that also wasn't good... It's just bizarre. It seems like something must be catastrophically wrong for me to hit 89 with so much rad. 

Yeah 89 with a 360 rad. There must be a problem somewhere. Yeah it's probably the die and the heat-spreader doesn't quite stick to each other.

[wONKEyeYEs]

Haven't done this myself.

Um, did you remove the origi  nal silicone?

[Fasauceome]

4 minutes ago, 227167_1454181677 said:

pee shape

pea shape?

 

Also, did you reseal? GamersNexus says that undoes all the delidding advantage.

Additionally, it looks like your GPU and CPU share a loop. What are your voltages and frequency of CPU?

[Mick Naughty]

Mine hit 90c with a 360+280. No delid. Maybe sand the bottom of the IHS a bit.

[227167_1454181677]

11 minutes ago, fasauceome said:

pea shape?

 

Also, did you reseal? GamersNexus says that undoes all the delidding advantage.

Additionally, it looks like your GPU and CPU share a loop. What are your voltages and frequency of CPU?

LOL! Yes, "Pea" shape. Sorry about that. Sometimes the fingers just type without the mind present. And no, I did not reseal. Pressure from the CPU block is the only thing that holds the heat-spreader to the die.  I'll have to look at my voltages in a sec

[227167_1454181677]

1 minute ago, Mick Naughty said:

Mine hit 90c with a 360+280. No delid. Maybe sand the bottom of the IHS a bit.

No kidding?... And yea, I have 2 360 rads... It just doesn't seem like that could be normal. I mean if I hit 89 with this much rad, i can only imagine what it would be like on an air cooler. 

[Fasauceome]

Just now, 227167_1454181677 said:

No kidding?... And yea, I have 2 360 rads... It just doesn't seem like that could be normal. I mean if I hit 89 with this much rad, i can only imagine what it would be like on an air cooler. 

I hit 4.7GHz with my hyper 212 evo, but I had a very special little 4790k. We still don't know your voltage and frequency.

[Mick Naughty]

1 minute ago, 227167_1454181677 said:

No kidding?... And yea, I have 2 360 rads... It just doesn't seem like that could be normal. I mean if I hit 89 with this much rad, i can only imagine what it would be like on an air cooler. 

Well that was at 4.8ghz 1.30v I think. Only on aida, so isn't a big deal to me. But will be deliding here shortly.

[227167_1454181677]

6 minutes ago, fasauceome said:

I hit 4.7GHz with my hyper 212 evo, but I had a very special little 4790k. We still don't know your voltage and frequency.

4 GHz, and offset the core voltage to -.045 >.>

[Fasauceome]

2 minutes ago, 227167_1454181677 said:

4 GHz, and offset the core voltage to -.045 >.>

 

Yeah, that's not great at all, since all core turbo should be even higher than that. I think having a single radiator for your CPU and GPU is the culprit, even a 360 rad is basically a dual rad for the GPU and a single rad for the CPU, but not exactly so it's not really cooling all that well.

[227167_1454181677]

Just now, fasauceome said:

Yeah, that's not great at all, since all core turbo should be even higher than that. I think having a single radiator for your CPU and GPU is the culprit, even a 360 rad is basically a dual rad for the GPU and a single rad for the CPU, but not exactly so it's not really cooling all that well.

No no no...  I have TWO 360mm rads... 

[TheDankKoosh]

11 minutes ago, 227167_1454181677 said:

No kidding?... And yea, I have 2 360 rads... It just doesn't seem like that could be normal. I mean if I hit 89 with this much rad, i can only imagine what it would be like on an air cooler. 

What kind of paste are you using, also try liquid metal under the ihs and some good paste on top. Check to see if the ihs is positioned wrongly when applying your cooler, and last of all check to make sure the ihs doesn't have any physicals defects.

[Fasauceome]

Just now, 227167_1454181677 said:

No no no...  I have TWO 360mm rads... 

Ah now that does change things.

Is the front intake and the top exhaust?

[227167_1454181677]

Just now, fasauceome said:

Ah now that does change things.

Is the front intake and the top exhaust?

Yep yep... I'm close to tearing this thing apart, reapplying, and then coming back with identical news. I'm stumped x_x

[Fasauceome]

Just now, 227167_1454181677 said:

Yep yep... I'm close to tearing this thing apart, reapplying, and then coming back with identical news. I'm stumped x_x

What's the loop order?

[227167_1454181677]

3 minutes ago, TheDankKoosh said:

What kind of paste are you using, also try liquid metal under the ihs and some good paste on top. Check to see if the ihs is positioned wrongly when applying your cooler, and last of all check to make sure the ihs doesn't have any physicals defects.

Arctic MX-4

[aezakmi]

wait, are you pumping hot liquid from the graphics card to the CPU?

 

F.

[227167_1454181677]

7 minutes ago, aezakmi said:

wait, are you pumping hot liquid from the graphics card to the CPU?

 

F.

That doesn't matter at all
 

 

[For Science!]

36 minutes ago, 227167_1454181677 said:

-

Pics of your delid process? I always re-sealed mine with silicone and in 3 out of 3 occasions observed 15 degrees improvement. 

 

- Delid

- Clean Silicone off

- Clean Die

- Spread Liquid metal on both sides, I do it fairly generously, too little can be too bad too, too much is also bad

- apply thin strip of silicone

- Release with ample pressure

 

 

 

 

 

 

relidding. 20 minutes.

 

 

[227167_1454181677]

13 hours ago, For Science! said:

Pics of your delid process? I always re-sealed mine with silicone and in 3 out of 3 occasions observed 15 degrees improvement. 

 

- Delid

- Clean Silicone off

- Clean Die

- Spread Liquid metal on both sides, I do it fairly generously, too little can be too bad too, too much is also bad

- apply thin strip of silicone

- Release with ample pressure

 

 

 

 

 

 

relidding. 20 minutes.

 

 

Thank you for the advice, Science. Admittedly, I didn't apply liquid metal under the IHS. I only applied it to the die as it being conductive scares the shit out of me. I will try liquid metal one more time along with silicone. I will get back to you with the results. Thanks for the insight!

[AngryBeaver]

The only place to actually use the LM is under the IHS imo. IF you are worried get either some liquid electrical tape or some clear nail polish and paint over all the short points on top. Then you can also take tape and apply it around the DIE so it covers the rest of the chip.. so anything that spills off the die will land on the tape. Then when you are finished you just pull off the tape and apply the lid. I make sure to use a blade or something shard to make a mark around where the old paste was before I clean it off... makes it easy to apply the LM to the right part of the lid.

 

 

If I was going to use nail polish or the Liquid tape I would 100% paint over the red ones, but if you want to be extra safe you can use it on the yellow too, but I don't feel they are a realistic risk unless you apply the LM wrong.

[227167_1454181677]

Just giving the conclusion to this story, I reapplied some Thermal Grizzly and my temps have now dropped 20c. However, I did discover I was putting on one way too much initially. My system actually refused to even put out a signal to my monitor for the first two applications. Each time I took off the IHS, I'd see little traces of liquid metal around the chips protruding around the die (4790k doesn't look like the picture above at all. There's actually tiny little chips here and there). Anyways, I cleaned it all off twice, and on the third application, applied it -very- gingerly and it went without problems. Thanks for everyone's help / assistance!