Extremely Quick Thermal Paste Degradation, 15°C Increase - Bad Batch?

[Wet_Paint]

Approximately a month ago i "upgraded" all my thermal paste from MX-4 to Thermal Grizzly Hydronaut. Temperatures improved by approx. 3-5 degrees °C but after two to three weeks temperatures started to increase to much higher levels.

MX-4: 63-67°C while gaming
Hydronaut initially: 59-63°C while gaming
Hydronaut now: 74-82°C while gaming (Same game as test scenario, in identical place)

MX-4: 78-87°C in Cinebench
Hydronaut initially: 73-85°C in Cinebench
Hydronaut now: 94-97°C in Cinebench (4.7Ghz caused temps over 100°C and throttled, previously would go up to 90°C max)

All tests ran at 4.6Ghz 1.180V and Mesh at 3.2Ghz 1.1V, CPU InputV 1.84V, SysAgent 0.93V.

Even when Windows boots up the temps hover around the 70°C mark for like a minute and eventually goes down to 55-58°C (before 42-46°C).

Liquid in the watercooler is filled to the max, and Pump runs in range of 2930-2980RPM all the time.

Ambient temps 28-30°C, they are always high no matter what, as far as i can tell they didnt change much recently, definitely not a upto Δ15°C.

One of my GPUs is watercooled and has been repasted with Hydronaut too, temps originally ranged from 42 to 45°C max, now they go all the way up to 52°C.

When i first noticed these issues i took off the CPU cooler and the middle part had almost "factory new" look, perfectly clean on both CPU and the waterblock in the center and only small tracers of thermalpaste on the edges of the CPU. After applying more thermalpaste to the center of the CPU temperatures improved to "initial" temperatures but after about a day the temps started rising again so i added new thermal paste again which improved the temps back to initial values again but this time there was thermalpaste spread properly between the IHS and waterblock. I ran the PC in horizontal position for about 3 days and temps stayed abut the same, after putting PC into standard vertical positions the temperature degraded back to much higher levels in about a day and half. So i added smaller amount of thermal paste into the center of IHS again and the temps didnt improve this time at all. The biggest issue is that i dont think ive enough thermalpaste left for more than one re-apply.

I've also tried 3 different motherboard BIOSes (v1201, v1301, v1401) and the temperatures were identical.

Temperatures on my laptop which i repasted too have increased by approx. 8°C over initial temps but my laptop has very low power parts.



Im sorry for a long post but my goal was to give you guys as much information as possible, thanks for any suggestions about how to decrease the temperature.

 

[Final_Fighter]

those temps are already on the high side to begin with. something isnt seated right, i would reapply the thermal paste again.

 

[Wet_Paint]

those temps are already on the high side to begin with. something isnt seated right, i would reapply the thermal paste again.

I did re-seat the waterblock & re-applied thermal paste like one hour before writing this post, literally no difference, maybe 1-2°C which i would consider margin of error since the PC was off while doing it and the liquid couldnt have reached stable temp. But thanks for suggestion, ig I'll have to do re-apply it again if i dont get any other suggestions but i doubt it will help, i also checked the waterblock's screws on the back of the motherboard and they were all evenly screwed. Tried even lifting the PC by the CPU waterblock and it didnt show any sign of being loose at all.

 

[nomdeplume]

When in doubt send an email out? To Thermal Grizzly and see what they reply.

 

[John Naylor]

I'm not sure how 'all" relates to the test scenario.... assuming 1st test if GPU and 2nd CPU ? And one GPU is water cooled, the other isn't ? Even using the same scenario, I wouldn't use a gaming test.

For CPU test I like to use Asus RoG Real Bench. I have had p95 24 hour stable OCs fail under multi tasking stress tests like RB. Most P95 busers will employ the older version of P95 so as to pose no danger to CPU but Im not comfy knowing my OC is stable only if mo modern instruction sets are used. And since it uses real world applications , it actually lets you set the PC higher than a synthetic test would allow.

Since voltage isn't in the equation, anything you can set in MSI AB is safe, I'll use Furmark. These insure that as I test the box at 3 month intervals its an insured apples and apples comparison against the day 1 baseline.

As to the horizontal / vertical thing, this normally points to air in the loop which should not be an issue with a CLC.

 

[cucker tarlson]

The cooler might've moved slightly,that's all it takes. reseat and reapply.Also, please note that ambient temps influence cpu/gpu temps very significantly.

 

[enxo218]

either the thermal paste is of poor quality or method of application is incorrect...I suspect the latter since there was initially an improvement on temperature reads but later began to degrade also this trend applies to all repasted devices even laptop. Clear all the paste off with alcohol and reapply

 

[FireFox]

@Wet_Paint

Did you forget that it's summer time?

 

[Aquinus]

How sure are you that it's the thermal compound? Are you sure there isn't something weird going on with your liquid cooling loop that might impact the flow of liquid? How about the radiator(s), is there any dust? Coincidences happen but, I don't trust them.

 

[crazyeyesreaper]

Question your using an Eisbaer + Eiswolf>

Are GPU and CPU in the same loop with the 360 + 120 rad. If setup that way is the Eiswolf one of the models that includes a secondary pump? Meaning your loop has the original Eisbaer pump + a pump for the GPU block or are you using a single pump to push all that?

If your using the original eisbaer pump its only rated for 70l/hr or 0.30 Gallons per minute. That is the pumps rating now add 2 radiators + 2 blocks its likely a great deal lower than that. With that much strain on the pump I have to wonder if its beginning to break down. There is a reason Alphacool has GPU blocks with built in pumps to help.

 

[Wet_Paint]

@Wet_Paint

Did you forget that it's summer time?

Yeah ik its summer *obviously* but my room is 26-30°C the whole year (multiple PCs in the room and brick walls with thermal isolation). Temps like these have been here literally for 2 months, i dont see reason why would the temp change now after all that time.

Question your using an Eisbaer + Eiswolf>

Are GPU and CPU in the same loop with the 360 + 120 rad. If setup that way is the Eiswolf one of the models that includes a secondary pump? Meaning your loop has the original Eisbaer pump + a pump for the GPU block or are you using a single pump to push all that?

If your using the original eisbaer pump its only rated for 70l/hr or 0.30 Gallons per minute. That is the pumps rating now add 2 radiators + 2 blocks its likely a great deal lower than that. With that much strain on the pump I have to wonder if its beginning to break down. There is a reason Alphacool has GPU blocks with built in pumps to help.

Yes they are both connected to the same loop and only CPU pump is running, i had to cut off the second pump cable bcuz it was interfering with motherboard's "TUF plastic shield" (idk what else to call it), it has been running like this for over a year, before i had this motherboard i tested the temps with both pumps and the temperature was identical.

How sure are you that it's the thermal compound? Are you sure there isn't something weird going on with your liquid cooling loop that might impact the flow of liquid? How about the radiator(s), is there any dust? Coincidences happen but, I don't trust them.

Well I used MX-4, MX-2, Alphacool stock one that came with the watercooler and all of them were "stable" temperature wise. I apply much more paste than average person does, it usually drips on the bracket which holds the CPU. I refill the loop every 90 days with destilled water so there shouldnt be any problems. Radiators were "de-dusted" with pressurized air like 3 weeks before any issues appeared, and they are dust free, just checked.

When i applied the hydronaut for first time i manually spread it with my finger across the whole CPU (which is something i never did before so dont judge) and then added big blob right in the center, and 4 smaller blobs into each corner, then i spread the excess from my finger on the waterblock.




UPDATE: So i didnt move my PC at all for about 24hrs (literally didnt move by even single milimeter) and the temperatures went down by +-5°C, ambient temperature is the same (around 30°C, using the same thermometer).

I cant take the cooler off right now since i'll be lucky if ill be able to re-paste it one more time and i keep that one "try" for the time when my Corsair 750D (the airflow edition) arrives on wednesday/thursday.

//Side-note: When i was re-applying the thermal paste last time i noticed "bulge" in center of the processor in shape of square, very slight one but it was visible mostly thanks to the thermal paste which formed "ring" around it. Also no matter in which order i screw the CPU waterblock screws in one always seems to be screwed in less than the others , which makes me feel like something could be bent between the CPU/waterblock. But this is probably really just me being paranoid.

 

[Athlonite]

I apply much more paste than average person does, it usually drips on the bracket which holds the CPU

By doing that you are defeating the whole purpose of thermal paste which is to fill in the tiny hills & valleys on the mated surfaces use to much and you end up with a thermal blanket

When i applied the hydronaut for first time i manually spread it with my finger across the whole CPU (which is something i never did before so dont judge) and then added big blob right in the center, and 4 smaller blobs into each corner, then i spread the excess from my finger on the waterblock.

That is way to much thermal paste a thinly spread coating is all you need

 

[TheGuruStud]

You're assuming the pump is working...we all know they're THE failure point. Idk about the other stuff. You'll have to reapply everything and see if they drop to your original values to know it's the paste. Damn scientific procedures...

 

[crazyeyesreaper]

Yeah ik its summer *obviously* but my room is 26-30°C the whole year (multiple PCs in the room and brick walls with thermal isolation). Temps like these have been here literally for 2 months, i dont see reason why would the temp change now after all that time.



Yes they are both connected to the same loop and only CPU pump is running, i had to cut off the second pump cable bcuz it was interfering with motherboard's "TUF plastic shield" (idk what else to call it), it has been running like this for over a year, before i had this motherboard i tested the temps with both pumps and the temperature was identical.



Well I used MX-4, MX-2, Alphacool stock one that came with the watercooler and all of them were "stable" temperature wise. I apply much more paste than average person does, it usually drips on the bracket which holds the CPU. I refill the loop every 90 days with destilled water so there shouldnt be any problems. Radiators were "de-dusted" with pressurized air like 3 weeks before any issues appeared, and they are dust free, just checked.

When i applied the hydronaut for first time i manually spread it with my finger across the whole CPU (which is something i never did before so dont judge) and then added big blob right in the center, and 4 smaller blobs into each corner, then i spread the excess from my finger on the waterblock.




UPDATE: So i didnt move my PC at all for about 24hrs (literally didnt move by even single milimeter) and the temperatures went down by +-5°C, ambient temperature is the same (around 30°C, using the same thermometer).

I cant take the cooler off right now since i'll be lucky if ill be able to re-paste it one more time and i keep that one "try" for the time when my Corsair 750D (the airflow edition) arrives on wednesday/thursday.

//Side-note: When i was re-applying the thermal paste last time i noticed "bulge" in center of the processor in shape of square, very slight one but it was visible mostly thanks to the thermal paste which formed "ring" around it. Also no matter in which order i screw the CPU waterblock screws in one always seems to be screwed in less than the others , which makes me feel like something could be bent between the CPU/waterblock. But this is probably really just me being paranoid.

With the second pump removed from the loop your basically limited to 70 l/h which is basically a best case scenario. In reality its probably closer to 50 and you have extra blocks and a non working pump to push fluid through. I am guessing its the single pump just working way to hard. Thats my guess tho having tested multiple AIO including expandable ones. Typically the low L/H and low head pressure designs just don't hold up well. its why as i said before Alphacool makes expansion systems for there loops with a pump built in. It maintains the typical flow rate you would expect.

If temps skyrocket try squeezing the tubing off and on a few times. The pressure from the squeeze should help the pumps flow rate especially when your first boot up. Seen the issue with Alphacool products before when they released low wattage dual bay dual pump reservoirs.

At this point if the pump is working properly your flow rate is likely anemic. If the pump is on its way out then your looking at even further degraded performance.

 

[Beertintedgoggles]

If you still have some MX-4 left, try applying it and verify that you get the same "before" temperatures. That will rule out things such as mounting pressure, dust, fans, ambient, etc.

 

[Athlonite]

With the second pump removed from the loop your basically limited to 70 l/h which is basically a best case scenario. In reality its probably closer to 50 and you have extra blocks and a non working pump to push fluid through. I am guessing its the single pump just working way to hard. Thats my guess tho having tested multiple AIO including expandable ones. Typically the low L/H and low head pressure designs just don't hold up well. its why as i said before Alphacool makes expansion systems for there loops with a pump built in. It maintains the typical flow rate you would expect.

If temps skyrocket try squeezing the tubing off and on a few times. The pressure from the squeeze should help the pumps flow rate especially when your first boot up. Seen the issue with Alphacool products before when they released low wattage dual bay dual pump reservoirs.

At this point if the pump is working properly your flow rate is likely anemic. If the pump is on its way out then your looking at even further degraded performance.

Somehow I don't think it's a pump problem as it's the same setup as before but look at where he admits to overusing thermal compound .... I apply much more paste than average person does, it usually drips on the bracket which holds the CPU & When i applied the hydronaut for first time i manually spread it with my finger across the whole CPU (which is something i never did before so dont judge) and then added big blob right in the center, and 4 smaller blobs into each corner, then i spread the excess from my finger on the waterblock.

I'd say it's more a case of to much thermal compound rather than a flow problem

 

[Wet_Paint]

By doing that you are defeating the whole purpose of thermal paste which is to fill in the tiny hills & valleys on the mated surfaces use to much and you end up with a thermal blanket



That is way to much thermal paste a thinly spread coating is all you need

Im fully aware of that, i just didnt want to bother having to re-paste it if it was too little, so for sake of not wanting to re-do it i applied much more, i just didnt trust the "spread with finger" method.

With the second pump removed from the loop your basically limited to 70 l/h which is basically a best case scenario. In reality its probably closer to 50 and you have extra blocks and a non working pump to push fluid through. I am guessing its the single pump just working way to hard. Thats my guess tho having tested multiple AIO including expandable ones. Typically the low L/H and low head pressure designs just don't hold up well. its why as i said before Alphacool makes expansion systems for there loops with a pump built in. It maintains the typical flow rate you would expect.

If temps skyrocket try squeezing the tubing off and on a few times. The pressure from the squeeze should help the pumps flow rate especially when your first boot up. Seen the issue with Alphacool products before when they released low wattage dual bay dual pump reservoirs.

At this point if the pump is working properly your flow rate is likely anemic. If the pump is on its way out then your looking at even further degraded performance.

I really appreciate your advices, but if the pump was too weak for the loop wouldnt there be some difference after connecting the second pump? I tried this many times and there was simply no improvement. And its basically impossible to bend the tubes to a point where they would stop the flow of the liquid, bcuz the tubes are wrapped in really hard metalic wire and pushing it too much is too risky, im worried about it snapping and cutting up the tube.

Also, would getting "proper" pump with >600L/hour help at this point? I know its easily said for someone who has a lot of experience in this but ive literally never connected any fully custom loop together and im really nervous about doing it especially if the money and risk could give zero improvement.

If you still have some MX-4 left, try applying it and verify that you get the same "before" temperatures. That will rule out things such as mounting pressure, dust, fans, ambient, etc.

I dont have any MX-4 left but ive the alphacool stock one, but it was really bad, i dont know if it would be any different from the "bad" temperatures ive now but Ill try it tomorrow and post results as soon as possible.

Somehow I don't think it's a pump problem as it's the same setup as before but look at where he admits to overusing thermal compound .... I apply much more paste than average person does, it usually drips on the bracket which holds the CPU & When i applied the hydronaut for first time i manually spread it with my finger across the whole CPU (which is something i never did before so dont judge) and then added big blob right in the center, and 4 smaller blobs into each corner, then i spread the excess from my finger on the waterblock.

I'd say it's more a case of to much thermal compound rather than a flow problem

I dont have as much money and time as Steve from GN but his results show that it doesnt matter if its too much.

“Too Much Thermal Paste” – Benchmark of Thermal Paste Quantity

But okay, ill remove all the paste, clean it off properly and ill apply just pea sized amount of thermalpaste in the center of the IHS.

 

[crazyeyesreaper]

if second pump does nothing then you likely have a pressure issue i would clean the thermal paste off, and use cheap stuff remount the blocks and see how the paste spreads via a single pea sized dot method. This will determine how well the paste is spreading and by extention determine how much mounting pressure is being exerted. higher mounting pressure you should see less paste with the copper base plate showing through. Hard to explain but try a mount with cheap paste then remove the block and take a picture.

 

[RealNeil]

//Side-note: When i was re-applying the thermal paste last time i noticed "bulge" in center of the processor in shape of square, very slight one but it was visible mostly thanks to the thermal paste which formed "ring" around it.

This isn't normal. It should be flat with tiny imperfections if any.

 

[Vayra86]

Yeah ik its summer *obviously* but my room is 26-30°C the whole year (multiple PCs in the room and brick walls with thermal isolation). Temps like these have been here literally for 2 months, i dont see reason why would the temp change now after all that time.



Yes they are both connected to the same loop and only CPU pump is running, i had to cut off the second pump cable bcuz it was interfering with motherboard's "TUF plastic shield" (idk what else to call it), it has been running like this for over a year, before i had this motherboard i tested the temps with both pumps and the temperature was identical.



Well I used MX-4, MX-2, Alphacool stock one that came with the watercooler and all of them were "stable" temperature wise. I apply much more paste than average person does, it usually drips on the bracket which holds the CPU. I refill the loop every 90 days with destilled water so there shouldnt be any problems. Radiators were "de-dusted" with pressurized air like 3 weeks before any issues appeared, and they are dust free, just checked.

When i applied the hydronaut for first time i manually spread it with my finger across the whole CPU (which is something i never did before so dont judge) and then added big blob right in the center, and 4 smaller blobs into each corner, then i spread the excess from my finger on the waterblock.




UPDATE: So i didnt move my PC at all for about 24hrs (literally didnt move by even single milimeter) and the temperatures went down by +-5°C, ambient temperature is the same (around 30°C, using the same thermometer).

I cant take the cooler off right now since i'll be lucky if ill be able to re-paste it one more time and i keep that one "try" for the time when my Corsair 750D (the airflow edition) arrives on wednesday/thursday.

//Side-note: When i was re-applying the thermal paste last time i noticed "bulge" in center of the processor in shape of square, very slight one but it was visible mostly thanks to the thermal paste which formed "ring" around it. Also no matter in which order i screw the CPU waterblock screws in one always seems to be screwed in less than the others , which makes me feel like something could be bent between the CPU/waterblock. But this is probably really just me being paranoid.

What the hell man.

Why are you being so creative with your thermal compounds. You are literally creating your own potential issues... and making it MUCH harder to find the real problems.

Either use a big blob in the center (pea sized if you want to 'overdo it', a little more than a rice grain size if you're sensible) or use a plastic card or something like that to evenly spread a THIN layer across the whole surface - which means applying some good pressure during the process - again to prevent air bubbles from getting caught underneath. The latter is mostly useful when the CPU die is not dead center, ie for bigger (HEDT) chips and stuff with multiple chip solutions under the IHS.

There are no sensible other ways to do it without risking a bad mount. After a proper application and mount you can start worrying about your pump and WC solution.

'Using my finger'...lmao

 

[Wet_Paint]

I repasted it with the alphacool paste which came in the box with the cooler, and did some test runs in Cinebench. Ran each test 3 times and waited 20 minutes after boot.

Hydronaut: 89-93°C, 91°C average, 30.0°C ambient
Alphacool #1: 88-91°C, 89°C average, 30.6°C ambient
Alphacool #2: 89-91°C, 90°C average, 31.6°C ambient

if second pump does nothing then you likely have a pressure issue i would clean the thermal paste off, and use cheap stuff remount the blocks and see how the paste spreads via a single pea sized dot method. This will determine how well the paste is spreading and by extention determine how much mounting pressure is being exerted. higher mounting pressure you should see less paste with the copper base plate showing through. Hard to explain but try a mount with cheap paste then remove the block and take a picture.

Hope these pictures are good enough.



Edit: took a picture of the mounting screws, the remaining two look exactly like the one on the left, but the one on the right is impossible to "screw in fully", ive tried and even with a lot of pressure i couldnt move it even little bit, actually i stripped the screw(just a little bit, still can unscrew/screw it as usually but it has some wiggle room).

 

[cucker tarlson]

That's bad coverage, seen almost the same picture on my 4790k when it started to heat up more than it used to before, just put a big round drop in the center and try to apply the pressure slowly and evenly.

 

[Ferrum Master]

shaite... are you replacing TIM on a running machine? The LED are glowing... it is on...

 

[Wet_Paint]

shaite... are you replacing TIM on a running machine? The LED are glowing... it is on...

The LEDs are on even if the PC is off. They usually run for +- 1-2 minutes after disconnecting the PC from the wall.

That's bad coverage, seen almost the same picture on my 4790k when it started to heat up more than it used to before, just put a big round drop in the center and try to apply the pressure slowly and evenly.

Gonna use more thermalpaste then but this will be probably the last application of that alphacool paste since its almost empty and i really need to save up the remaining hydronaut for when the 750D arrives.

 

[Vya Domus]

Im sorry for a long post but my goal was to give you guys as much information as possible, thanks for any suggestions about how to decrease the temperature.

Well first of all these Skylake-X CPU will imply run very hot due to their nature and the fact that they aren't soldered , that 4.7 Ghz overclock doesn't help your cause either.

That been said I have seen the pictures you posted with the TIM applied and nothing looks out of place to me. A blob in the middle is all you need to do , you can spread it however you want in a million different ways it will make no difference , this has been proven time and time again. To me it sounds like there is something wrong with the surface of the block or maybe even the IHS. Maybe the block isn't fully in place.

took a picture of the mounting screws, the remaining two look exactly like the one on the left, but the one on the right is impossible to "screw in fully", ive tried and even with a lot of pressure i couldnt move it even little bit, actually i stripped the screw(just a little bit, still can unscrew/screw it as usually but it has some wiggle room).

I think you have your problem right there.

Gonna use more thermalpaste then

Don't , it's enough.