Question (new) 9900k running on 100°C idle with 360mm radiator

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Dec 6, 2019
20
0
10
So, yeah. The story begins with me buying a new case. I switch my pc into the new case, start the pc - aaand 90°C CPU, back then it was a 8700k. I chose to delid it, which I wanted to do anyway. So after that it hit 98°C, yeah.
That´s when I thought that the CPU might have broke at some point. I´ve bought a 9900k, put it in - aaand 100°C.

For cooling I use a BeQuiet Silent Loop360, that´s why I think those temperatures are pretty ridiculous. I just filled it up again with the new Corsair Coolant, so it´s not like the radiator was empty.

I´m pretty clueless (and broke) as of now. Does anybody have an idea?

edit: The waterblock get´s crazy hot while the tubes are still cold. What is happening . May it be that the system reading is incorrect?

edit: Ok, so right now it´s running on 70°C idle. I don´t do anything but it seems like it got better. However that´s still way too much in idle. It got to 100°C eventually, just had to let it run some time :(
 
Last edited:
Dec 6, 2019
20
0
10
Alright everyone, now it´s getting interesting.

So I´ve moved the radiator from the side to the top, so the water would be flowing down and the air gets up to the back of the rad where it would cause less problems and becomes audible (so I´d notice it).

After booting up, the CPU was at 55°C, then continuing to rise until 100°C, again. But that´s at least an improvement.

The interesting thing is that the lower tube is now completely warm, while the upper one stays cold:
View: https://imgur.com/DAg4t59


I´m not sure why this is. Does anyone know the water flow on the silent loop?

Edit: Oh and I looked on BeQuiets Website and they advertive the SilentLoop as "No noticable vibrations", so I guess there goes that.
 
Last edited:
There really should not be any air in the radiator, but nearly all AIOs have this problem. And minor correction, the reservoirs being mentioned are actually the radiator tanks which cover/connect the tubes in the rad. The only reason I call mention to this is to prevent confusion between these objects vs. a standalone or add-on reservoir possibly seen elsewhere in a liquid cooling loop (although rarely on an AIO).

Fair point, but I did write "reservoirs on the rad" and they are drawn in the picture. They do serve to hold extra fluid, as well as to provide a mounting point for the inlet/outlet tubes, and finally to direct the flow at the end opposite the I/O tubes. Extra fluid allows for some air, provided the fitting is far enough away from the edge of the tank. See this diagram:

49178996638_5810e57eb5_z.jpg


In the above example with the same volume of air, but with smaller tanks, the amount of air would become more significant and the air would be in front of the fitting in the "I/O on Side" and possibly the "I/O on Top" scenario.

Having said that, I do agree that it should be filled as much as possible. That's why I pushed the idea that it's not filled with enough fluid.
 
I´m not sure why this is. Does anyone know the water flow on the silent loop?

The hot tube is where the fluid is going into the rad and the cold tube is where it is coming out. Bottom line is you'll have to recheck your fluid because it seems like it's not filled enough. If the manufacturer has info on refilling it, you should start there so the proper technique is employed in order to bleed out all the air. GamersNexus shows how they did it here:
View: https://www.youtube.com/watch?v=0fobND-PZUg


This is how the coolant flows through an AIO rad (fins not shown, just core tubes):

49179797917_88e4350ece_z.jpg
 

rubix_1011

Contributing Writer
Moderator
Fair point, but I did write "reservoirs on the rad" and they are drawn in the picture. They do serve to hold extra fluid, as well as to provide a mounting point for the inlet/outlet tubes, and finally to direct the flow at the end opposite the I/O tubes. Extra fluid allows for some air, provided the fitting is far enough away from the edge of the tank. See this diagram:

49178996638_5810e57eb5_z.jpg


In the above example with the same volume of air, but with smaller tanks, the amount of air would become more significant and the air would be in front of the fitting in the "I/O on Side" and possibly the "I/O on Top" scenario.

Having said that, I do agree that it should be filled as much as possible. That's why I pushed the idea that it's not filled with enough fluid.

Yes, I was agreeing with you. The only reason I made mention was to prevent someone from confusing what you were describing vs. a standalone watercooling reservoir.

I didn't think this would end up being a contest to see who was more right, although 'holding more coolant' in the radiator tanks provides you with only one benefit: holding more coolant. It serves zero purpose for cooling potential.
 

zx128k

Reputable
Mixed Coolant types is a very bad idea without cleaning out the loop. https://www.overclockers.co.uk/mayhems-blitz-cleaning-system-wc-103-mh.html

Should never mix coolants. Also the pump and/or block could be full of crud. Sounds like there is a blockage somewhere or pump issues. If filled with Corsair Coolant, there should be enough coolant if done correctly. You state the water block get´s crazy hot while the tubes are still cold. So there has to be a flow issue. Heat is trapped in the block and you should troubleshoot there first. If you can feel the pump working then ask yourself why is the hot coolant not being displaced by new cold coolant coming from the rad. I would guess that there is no flow from inlet to outlet on the block.

I would replace to cooler or RMA it for repair. Could be the pump is too weak or the block and loop needs cleaned out. Sometimes you get algae in the loop or a lot of crud which could be a sign of corrosion. Just my opinion.

 

rubix_1011

Contributing Writer
Moderator
This is an AIO, not a custom loop. Yes, the pump is terrible, but it should still manage to work for what it is designed for.

This is likely either an airlock problem, or the pump header is not correctly setup or defined to be operational.

Need to tap the pump and shake the tubing to see if there are air pockets.
 
Dec 6, 2019
20
0
10
OP 'might' have to open the block to see what is happening. That could also void the warranty.

With water loops its better to be safe than sorry. I would RMA, just my opinion.
I was thinking about RMA aswell. the problem with that is that it "can take up to 8 weeks" unitl they´re finished.

As I originally stated, I already filled it up like a week ago. However, since, when I shake the radiator, I can hear the water splash (Air inside), I will fill it up in like 30min.

I really don´t want to "waste" my XL5 coolant since I wanted to go custom without having to buy another bottle of coolant. So if the problem´s still present after filling up, I´ll wait for my dark rock pro 4 to arrive and RMA the AiO.
 
Dec 6, 2019
20
0
10
Mixed Coolant types is a very bad idea without cleaning out the loop. https://www.overclockers.co.uk/mayhems-blitz-cleaning-system-wc-103-mh.html

Should never mix coolants. Also the pump and/or block could be full of crud. Sounds like there is a blockage somewhere or pump issues. If filled with Corsair Coolant, there should be enough coolant if done correctly. You state the water block get´s crazy hot while the tubes are still cold. So there has to be a flow issue. Heat is trapped in the block and you should troubleshoot there first. If you can feel the pump working then ask yourself why is the hot coolant not being displaced by new cold coolant coming from the rad. I would guess that there is no flow from inlet to outlet on the block.

I would replace to cooler or RMA it for repair. Could be the pump is too weak or the block and loop needs cleaned out. Sometimes you get algae in the loop or a lot of crud which could be a sign of corrosion. Just my opinion.

Well, to be honest I already mixed coolants. I´ve filled the stock coolant up if the corsiar XL5 since I didn´t think it would be that bad. But it shouldn´t be algae or corrosion since both coolants advertive with "no algae" and "no corrosion".

One of the tubes gets hot and I can feel water flowing, so it´s at least not fully blocked. Maybe the waterblock just gets hot because of the 100°C underneath it but I fully agree that the waterflow´s not right.

Should I rather dump the coolant already in the AiO and go fully XL5?
 

zx128k

Reputable
Well, to be honest I already mixed coolants. I´ve filled the stock coolant up if the corsiar XL5 since I didn´t think it would be that bad. But it shouldn´t be algae or corrosion since both coolants advertive with "no algae" and "no corrosion".

One of the tubes gets hot and I can feel water flowing, so it´s at least not fully blocked. Maybe the waterblock just gets hot because of the 100°C underneath it but I fully agree that the waterflow´s not right.

Should I rather dump the coolant already in the AiO and go fully XL5?

The coolants can react with one another. I mixed coolant with distilled water, just enough distilled water to top it up and got a pink/red coloured coolant very quickly. A clear sign of copper in the coolant. Copper is good as a water purifier, algaecide, fungicide, nematocide, molluscicide as well as an anti-bacterial and anti-fouling agent. But it means corrosion of the rad for me which was copper. I quickly replace the rad and flushed/cleaned the loop. Corrosion + time = leaks.

If the block is correctly seated and there is thermal paste on the IHS. You should only see 100c if the cpu is overclocked or vcore has been changed from default. Or if the cooler is not rated for what it is cooling.

Cleaning coolant out requires a complete flush of the water loop with cleaning fluid. You are meant to clean the loop before you add any new coolant. Also you should inspect your loop, you are looking for corrosion, things growing and plaque. The reason for this is the main paranoia with water cooling is always leaks. You can also get a buildup of crud in the block because some other part of the loop has corrosion, there are coolant issues or things growing.

The main rules with water cooling are, never ignore problems or issues and its always better to be save than sorry.
 
Last edited:
Dec 6, 2019
20
0
10
The coolants can react with one another. I mixed coolant with distilled water, just enough distilled water to top it up and got a pink/red coloured coolant very quickly. A clear sign of copper in the coolant. Copper is good as a water purifier, algaecide, fungicide, nematocide, molluscicide as well as an anti-bacterial and anti-fouling agent. But it means corrosion of the rad for me which was copper. I quickly replace the rad and flushed/cleaned the loop. Corrosion + time = leaks.

If the block is correctly seated and there is thermal paste on the IHS. You should only see 100c if the cpu is overclocked or vcore has been changed from default. Or if the cooler is not rated for what it is cooling.

Cleaning coolant out requires a complete flush of the water loop with cleaning fluid. You are meant to clean the loop before you add any new coolant. Also you should inspect your loop, you are looking for corrosion, things growing and plaque. The reason for this is the main paranoia with water cooling is always leaks. You can also get a buildup of crud in the block because some other part of the loop has corrosion, there are coolant issues or things growing.

The main rules with water cooling are, never ignore problems or issues and its always better to be save than sorry.
Alright. So as far as I understood, mixing coolants can lead (over time) to leaks or breaking components, but over a short time period it should still be able to cool? (I´m saying this because it shouldn´t be a concern to cooling performance and. at least right now, I only want to get my CPU to peak at <90 and not 100)

I´ve tested 2 different CPUs (8700k and 9900k). I even underclocked the (already delid) 8700k to have hyperthreading disabled and only let 4 cores run on 3.2GHz with 0.98V vcore and still got 100°C. I didn´t touch anything on the 9900k yet because I want to keep the warranty, for now. In regards of thermal paste, I´ve tested a lot of different variations and looked up some tests on youtube. It looks like it really doesn´t matter how much thermal paste goes on the CPU as long as it´s enough. When I installed the 9900k I kept that in mind and just threw just so much on there that I thought "That´s a little too much", so I´d be on the save side of things.

The AiO should really be able to cool the CPUs I used, since I´ve seen tests using this exact setup and reaching really good temps. And a 360 radiator, even if the fans on it are not pressure optimized, should be able too cool a stock 9900k which is soldered.

As soon as I solved the cooling problem, I´ll try to flush the AiO and only use 1 coolant.
 
Dec 6, 2019
20
0
10
Does anyone know if there is a way to only let my AiO run, without starting the CPU so I don´t have to mount it and waste thermal paste? Because I filled it up now and I can´t seem to get some air out of the radiator no matter how hard I try.
 
Dec 26, 2019
1
0
10
I realize this reply is on the late side, but I had the same exact issue after switching cases. Granted I have a different AiO but they are generally the same anyways.

I noticed my pump kept dropping to 0 rpm and starting itself over at a higher value (I forget exact number) but then immediately dropping again. I concluded that some buildup must have broken loose during the move and got jammed into the pump area.

My solution was simply dismounting the pump/block and hitting it with the butt end of a screwdriver handle a few times around the perimeter of the block base. Theoretically this broke the chunk loose or at least into smaller particles that were able to pass and all has been fine since.

Hopefully this helps someone, OP or not.