Alphacool CORE 1 CPU block - bulging with danger of splitting?

[BoggledBeagle]

I had the same block with a mounting frame on my cpu with no problems.

Yeah, but just imagine if you had less thick rubber stuffing sheet in your block, the block could have had a thinner layer of paste under it and the heat could have been flowing so much easier! SO MUCH EASIER...

So you got a different block in the end? Before I discovered that rubber overpressure bulge and those loose screws in aluminium, I had pretty positive feeling from the block.

 

[FoulOnWhite]

Yeah, but just imagine if you had less thick rubber stuffing sheet in your block, the block could have had a thinner layer of paste under it and the heat could have been flowing so much easier! SO MUCH EASIER...

So you got a different block in the end? Before I discovered that rubber overpressure bulge and those loose screws in aluminium, I had pretty positive feeling from the block.

The block was fine even with the seal in it which it needs.

 

[BoggledBeagle]

I found a plastic blister pack for RAM kit and cut an insert from it. It is 0,37 mm thick, the original rubber sheet is 0,64 mm.

The horrible bulge is gone. It seems that the middle of the base has now sufficient support (SUFFICIENT IN COMPARISON TO COMPLETELY REMOVING THE INSERT, as I did before), you can see a tiny amount of bulging still left there. I hope I will be able to build the loop again to compare the performance and paste imprints over the weekend.

 

[Noci]

Yeah, but just imagine if you had less thick rubber stuffing sheet in your block, the block could have had a thinner layer of paste under it and the heat could have been flowing so much easier! SO MUCH EASIER...

So you got a different block in the end? Before I discovered that rubber overpressure bulge and those loose screws in aluminium, I had pretty positive feeling from the block.

It looks like you just not want to hear or accept what the other peeps here want to tell you as well as feedback from companies backed by specialists.

If you look at the following video, please absorb the knowlegde integrated in it and readjust your perspective to reallity.

If after this you are still having doubts about this issue, I seriously think you're only posting to get your message numbers up. However you do seem to entertain forum members with a more than average technical knowledge, I would not want to rob them from their daily dose of humor .

 

[BoggledBeagle]

If you look at the following video, please absorb the knowledge

I rather just do and test things myself, but thanks for the link.

 

[Solaris17]

not a 10 bar pressure resistant seal.

Help me get in my ADS I’m goin deep!

Atmospheric diving suit - Wikipedia

en.m.wikipedia.org

But seriously they replied to you and told you it would be fine. Just use the block or lap it. We already have like a 7 page thread about your ink method this is just one guy and a water block. If your that sketch and it’s not that expensive as you’ve said just return it. The OEM says your fine.

 

[ty_ger]

I found a plastic blister pack for RAM kit and cut an insert from it. It is 0,37 mm thick, the original rubber sheet is 0,64 mm.

The horrible bulge is gone. It seems that the middle of the base has now sufficient support, you can see a tiny amount of bulging still left there.

Reducing pressure increases support? Wha? Are we in the upside-down?

I am surprised you are not now afraid of it imploding.

 

[tabascosauz]

Exactly this.

Once the block has been mounted (properly), there is literally zero chance of it popping open. It isn't going to catastrophically fail because of a piece of rubber. Unless the mounting hardware somehow breaks or is used incorrectly, that in and of itself will guarantee your block won't break.

BTW, inserts are nothing new. Swiftech used them as far back as the early 2000s, EK has used them, Koolance used them etc etc and so on. So yeah, Other manufacturers have in fact stepped there.

Believe it or not, Alphacool really does have a competent R&D team or they wouldn't be nearly as successful as they are. Which is likely why you received the response they gave.
Alphacooleddy is our resident contact iirc (I may have his name wrong) he's easy to find. I'm certain he would be happy to explain how your block is designed to work.

@EddyAlphacool

I don't know if it would do any good, though. I think OP would just start fighting with him instead.

 

[BoggledBeagle]

Reducing pressure increases support? Wha?

Thanks, I fixed my post, I meant sufficient support IN COMPARISON TO COMPLETELY REMOVING THE INSERT, as I did before.

 

[xmanrigger]

I have no idea what overtightening are you guys talking about, I am aware of the loose screws in the alu top and I am very careful with them. The main O-ring is in a groove and the base goes metal on metal on the block body.

What was that "EK-Gate fiasco"?

EK-Gate Fiasco has been explained as above.

"Back in the EK-Gate fiasco, there were thousands of keyboard metallurgist claiming Eddie was using bunk copper. Guess what? He was using the best anyone could get their hands on." quote from me above.

I was one of them warriors at that time. Yes, I questioned the same without knowing what I was talking about. But hey, I put my money where my mouth was and did a favor for the world. I was proved wrong. What have you done other than make an assumption? I can say I have never had a problem with any Alphacool product I have

 

[FoulOnWhite]

Like i said, mine had the thick black rubber seal in it, i used it with a CPU support frame and it was fine, temps were fine. There is no need to remove the seal.

 

[BoggledBeagle]

Mess has been created, mistakes happened, testing is done.




Here are the test candidates:



The results may surprise you. I will put them together in the evening.

Here is the most beautiful and even thermal paste imprint ever created:



Compare that to the horrible catastrophy the stock bulged block created - see the first post again.

 

[A Computer Guy]

Mess has been created, mistakes happened, testing is done.

View attachment 345342View attachment 345343
View attachment 345344View attachment 345345

Here are the test candidates:

View attachment 345346

The results may surprise you. I will put them together in the evening.

Here is the most beautiful and even thermal paste imprint ever created:

View attachment 345347

Compare that to the horrible catastrophe the stock bulged block created - see the first post again.

One million dollar question: Did you have any temperature improvement?

 

[BoggledBeagle]

One million dollar question: Did you have any temperature improvement?

Yes, a huge improvement, will post it in the evening.

EDIT: It was not a huge improvement. I decided to redo the tests with more rads to lower the water temperature (now it was around 35°C), to properly monitor the temperature of water incoming in the block (I have the water temp sensor that came with the pump and also have the Corsair controller and monitoring box).

The trouble is that I do not have any test computer and I did not want to mess too much with my only functioning PC (now air cooled in a small case).

Anyway, the result was that the best performed the flat block with the plastic insert with the small slot. It seems that concentrating the water flow to the place where it matters most helped.

My one randomly chosen smaller slot size (1,3 x 26mm) means that there is most likely a different size which will perform even better. And such optimal slot will of course perform best when it will have a proper water guide leading to it.

While this block is considered to be one of the best performing blocks available today, I believe there is still some room for improvent:

1) by making a flat version for people who have flat CPUs (easily doable by using thinner inserts)
2) by concentrating the water flow more to the place with the highest heat flow (easily doable by cutting different slot in the insert and redoing the water guide)
3) by possibly rearranging (optimising) the cuts in the base for this new water flow (the cuts are probably done with a CNC machine - just reprogram it)

I bet the engineers in Alphacool will be jumping up and down and shrieking with joy like apes in the ZOO when they get their bananas, after they find out, that they can improve their product even more...

 

[Dr_b_]

Have you considered just getting a different block, all the time and worry you spent on this probably isn't worth it, you are concerned about metal cracking, copper purity at the "macroscopic level", and other neurotic isht. Having used multiple Aqua computer waterblocks, can say their build quality is excellent and may be more in line with your criteria.

 

[ty_ger]

Conversely, I have to wonder how much worse your design overhaul would perform for people with unmodified CPU mounts who needed the bulge in the block. I am sure they chose their design based on more than one data point. I am sure they chose a design which they believed to be the best on average.

 

[BoggledBeagle]

Conversely, I have to wonder how much worse your design overhaul would perform for people with unmodified CPU mounts who needed the bulge in the block.

The answer is easy, it would perform worse for them.

I do not think you need to make one universal version of the product. You can make more versions.

 

[ShrimpBrime]

Here's a formula for clamping force for constant clamping force for each time you mount a water block

This may help aid in variance and ensure proper testing techniques.

this formula is what I use for Peltier (TEC) direct die cooling. It helps 2 things. 1 even clamping force and also so I dont crush the Peltier module.

Maybe this can help you, maybe not. But at least it's something to go by that would make sense around here.

T=((Sa x A)/N) x K x d
where:
T = torque on each bolt
Sa = cycling 25-50 psi, static 50-75 psi.
A = total surface area of module(s)
N = number of bolts used in assembly
K = torque coefficient (use K=0.2 for steel, K=0.15 for nylon)
d = nominal bolt diameter
For steel fasteners, we typically recommend either:

6-32 d=.138 in (.350 cm)
4-40 d=.112 in (.284 cm)

The following recommended torque is calculated for nine 9500/065/018 modules held by four 4-40 steel fasteners:

T=((75 lbs/in.2 x (.44″ x .48″) x 9)/4)x 0.2 x .112 in. = 0.8 in-lbs.

 

[kapone32]

Yes, a huge improvement, will post it in the evening.

EDIT: It was not a huge improvement. I decided to redo the tests with more rads to lower the water temperature (now it was around 35°C), to properly monitor the temperature of water incoming in the block (I have the water temp sensor that came with the pump and also have the Corsair controller and monitoring box).

The trouble is that I do not have any test computer and I did not want to mess too much with my only functioning PC (now air cooled in a small case).

Anyway, the result was that the best performed the flat block with the plastic insert with the small slot. It seems that concentrating the water flow to the place where it matters most helped.

My one randomly chosen smaller slot size (1,3 x 26mm) means that there is most likely a different size which will perform even better. And such optimal slot will of course perform best when it will have a proper water guide leading to it.

While this block is considered to be one of the best performing blocks available today, I believe there is still some room for improvent:

1) by making a flat version for people who have flat CPUs (easily doable by using thinner inserts)
2) by concentrating the water flow more to the place with the highest heat flow (easily doable by cutting different slot in the insert and redoing the water guide)
3) by possibly rearranging (optimising) the cuts in the base for this new water flow (the cuts are probably done with a CNC machine - just reprogram it)

I bet the engineers in Alphacool will be jumping up and down and shrieking with joy like apes in the ZOO when they get their bananas, after they find out, that they can improve their product even more...

By your reasoning the Eisbaer should perform better than the C1 for point number 2. That has a res unit on top of the block. The Inwin Sr360 also comes to mind. The TPU review though showed that the C1 is the best performing block in their tests.

 

[BoggledBeagle]

By your reasoning the Eisbaer should perform better than the C1 for point number 2. That has a res unit on top of the block. The Inwin Sr360 also comes to mind. The TPU review though showed that the C1 is the best performing block in their tests.

I am afraid I do not understand why you talk about reservoir. In point 2 I meant heat flow density in various places of the block base layer in contact with the CPU. When the block layer is thin, you are not getting too much planar redistribution of heat in the block base, the block base acts more or less like an extention of the CPU IHS.

The small slot in the plastic insert has more than twice smaller area than the slot in the stock plate, and in spite of this causing worse water flow and additional resistance to water flow by the water guide not matching the small slot, just concentrating the water flow more on the spot with higher heat flow density improved performance of the block.

I hope it is now more clear what I meant.

 

[ShrimpBrime]

I hope it is now more clear what I meant.

Nobody really knows what you meant. Lol.

None of it has anything to do with a block bending to the point it has fractured with an undetermined amount of clamping force and repeated constant testing to perhaps show a gradual bending of said material.

Or perhaps because it's nickel dipped, there's a hump from being formed during the solidification process when cooling off. Who's to say the copper under the plating isn't or is flat? Or cracked for that matter.

 

[thesmokingman]

EK-Gate Fiasco has been explained as above.

"Back in the EK-Gate fiasco, there were thousands of keyboard metallurgist claiming Eddie was using bunk copper. Guess what? He was using the best anyone could get their hands on." quote from me above.

I was one of them warriors at that time. Yes, I questioned the same without knowing what I was talking about. But hey, I put my money where my mouth was and did a favor for the world. I was proved wrong. What have you done other than make an assumption? I can say I have never had a problem with any Alphacool product I have

Nope, we all knew the plating was shite. Some question the copper, ppl questioned everything. Putting it all on the copper is BS.

 

[xmanrigger]

Nope, we all knew the plating was shite. Some question the copper, ppl questioned everything. Putting it all on the copper is BS.

Yes, it was the plating process. I had mentioned that in a previous post. But the general consensus back then was that Eddie was using inferior materials. It was the plating process itself. And that is what I had said.

 

[BoggledBeagle]

I redid the measurement with larger rad and temperature monitoring (measuring water temperature in the rad output feeding the block), this time with 200W heat load. ( I do not want to degrade my CPU with 253W anymore... )





Results (water temperature / typical momentary average temperature over all cores / max temperature over all core in during the whole meaurement)

Plastic insert with large slot: 24,9/55/67
Plastic insert with small slot: 25,1/54-55/64 - 54 and 55 alternating regularly
Original rubber insert: 25,1/54-55/65 - 54 displayed more often than 55

All results are nearly identical, the paste imprints looked as before - very thin homogenuous layer of paste with the plastic insert and thin layer in place of the bulge and thick layer around it with the original rubber insert.

So my conclusions:

If you have a flat CPU, you do not need the block to have a pressure bulge of unhappy stressed base on it, you can put a thinner insert in it and have a nice relaxed flat block instead with the same performance.

While it is true that the block with the plastic insert with a small slot performed a little bit better than that with the large slot, I will use the one with large slot to lower the water flow restriction, if I choose to keep and use the block.

 

[Assimilator]

This thread is a beautiful demonstration of the Dunning-Kruger effect in action.