Ridiculously high temps on the H110i?

[Fwerimuxe]

Hey all,

 

For some time now I've been dealing with increasingly bad temps on my main system. About a three months ago I finally decided to remove my H110i and gave it a full-on cleaning through the rad and the fans with a fresh coat of thermal paste on my CPU. Things were fine until about last month where I noticed the temps got ridiculously high again reaching 100c when rendering. I then proceeded to clean the H110i again several times again whilst also re-applying the thermal compound of course.

 

CPU: i7-6700K

Water cooler: Corsair H110i GTX v1

Old idle temps: < 30c

Current idle temps: > 40c

 

Under load the temperature spikes up really fast, like in 3 seconds from idle temps to 100c. The moment I turn off the load it also drops really quickly back to idle temps.

 

I've been really considering just ripping of my H110i and caving in to a new cooler but I wanna make sure I'm not just doing something real stupid before making that purchase decision. So here are a few things I think might have been the cause of the problem..

1. I've been using the same tube of thermal paste for every application (Mastergel Pro)

2. I just suck at applying thermal compound even though I've watched through many videos to do the "P" method and not over/under coat my CPU

3. My H110i is actually out of commision

 

Some help would really be appreciated, thanks!

 

 

 

EDIT: I realised better late than never to help anyone who does come across this forum post in the future. I solved it.
tl;dr Dead pump. Solution? RMA it. I had the cooler RMA'd with no response from my local vendor for months so I had no idea what was wrong with the cooler during that time.
Anyways, this video from Bitwit demonstrates the exact problem I had so if your cooler happens to act the same way as mine and his.. Dead pump.

 

Thanks to all who helped me out through this!

[Origami Cactus]

3 minutes ago, Fwerimuxe said:

Under load the temperature spikes up really fast, like in 3 seconds from idle temps to 100c

Pump is dead

 

Do you feel liquid flowing around? Do you hear water flowing?

[mxk]

Pump dead, probably.

[Fwerimuxe]

Just now, Origami Cactus said:

Pump is dead

Owh o_o

If that's the case that'd really suck... anyway I can verify this?

[Origami Cactus]

2 minutes ago, Fwerimuxe said:

Owh o_o

If that's the case that'd really suck... anyway I can verify this?

Do you feel liquid flowing around? Do you hear water flowing?

[Fwerimuxe]

3 minutes ago, Origami Cactus said:

Do you feel liquid flowing around? Do you hear water flowing?

Urm as for feeling its hard to say, there is vibration but very subtle..

For water flowing not really.. just some clicking but I think that's just the derpy fan.

[Origami Cactus]

1 minute ago, Fwerimuxe said:

Urm as for feeling its hard to say, there is vibration but very subtle..

For water flowing not really.. just some clicking but I think that's just the derpy fan.

Then the pump is dead. You should be able to rma the cooler to corsair for free, because corsair coolers should have a 5 year warranty

[Fwerimuxe]

5 minutes ago, Origami Cactus said:

Then the pump is dead. You should be able to rma the cooler to corsair for free, because corsair coolers should have a 5 year warranty

Urmh alright, I'll go do that.. Thanks!

[Wr0ngEnemy_]

15 minutes ago, Fwerimuxe said:

Hey all,

 

For some time now I've been dealing with increasingly bad temps on my main system. About a three months ago I finally decided to remove my H110i and gave it a full-on cleaning through the rad and the fans with a fresh coat of thermal paste on my CPU. Things were fine until about last month where I noticed the temps got ridiculously high again reaching 100c when rendering. I then proceeded to clean the H110i again several times again whilst also re-applying the thermal compound of course.

 

CPU: i7-6700K

Water cooler: Corsair H110i GTX v1

Old idle temps: < 30c

Current idle temps: > 40c

 

Under load the temperature spikes up really fast, like in 3 seconds from idle temps to 100c. The moment I turn off the load it also drops really quickly back to idle temps.

 

I've been really considering just ripping of my H110i and caving in to a new cooler but I wanna make sure I'm not just doing something real stupid before making that purchase decision. So here are a few things I think might have been the cause of the problem..

1. I've been using the same tube of thermal paste for every application (Mastergel Pro)

2. I just suck at applying thermal compound even though I've watched through many videos to do the "P" method and not over/under coat my CPU

3. My H110i is actually out of commision

 

Some help would really be appreciated, thanks!

Maybe you plugged it in wrong, fan should go to cpufan on motherboard and ,,water block" (thing that goes on procesor) should go on systemfan1 or you installed it wrong

[Fwerimuxe]

3 minutes ago, Wr0ngEnemy_ said:

Maybe you plugged it in wrong, fan should go to cpufan on motherboard and ,,water block" (thing that goes on procesor) should go on systemfan1 or you installed it wrong

The main unit is connected to CPU_FAN on my Asus Z170 Pro Gaming MB. The two fans are connected to the main unit and the power cable of the main unit is connected to the PSU. Can you verify that this is correct?

[Stormseeker9]

Dead pump imo 

[aezakmi]

28 minutes ago, Origami Cactus said:

Then the pump is dead. You should be able to rma the cooler to corsair for free, because corsair coolers should have a 5 year warranty

Except for the fact that the pump flow meter clearly shows it's running at 3150 RPM

[Origami Cactus]

1 minute ago, aezakmi said:

Except for the fact that the pump flow meter clearly shows it's running at 3150 RPM

But they dont hear or feel water flowing, so the software can say all it wants, but if the water doesn't flow then it doesn't.

Ofcourse it can be a different issue too, but im 90% certain the pump is dead, if you have any other theories then lay them out.

[Wr0ngEnemy_]

39 minutes ago, Fwerimuxe said:

The main unit is connected to CPU_FAN on my Asus Z170 Pro Gaming MB. The two fans are connected to the main unit and the power cable of the main unit is connected to the PSU. Can you verify that this is correct?

im sorry english isnt my first language but fan's should go to cpu_fan and water block should go to systemfan and i installed h110i on my friends pc and he had only 3 of those cables i dont know about any other, but im 100% sure that those 3 cables are correctly installed if you follow my leads, that happened to my liquid cooling, temps go in 1 minute to 100C at idle, did you overclock your procesor?

[AriGee50]

9 hours ago, aezakmi said:

Except for the fact that the pump flow meter clearly shows it's running at 3150 RPM

 

8 hours ago, Origami Cactus said:

But they dont hear or feel water flowing, so the software can say all it wants, but if the water doesn't flow then it doesn't.

Ofcourse it can be a different issue too, but im 90% certain the pump is dead, if you have any other theories then lay them out.

I would agree that it could be the flow sensor just being dumb and not recognizing...except that in the two pictures, the flow rate is measured at two slightly different rates. If it was the sensor in the pump that was also Fked up and therefor reading incorrectly, it would be more likely that the flow rate would always be listed as the exact same number, it probably wouldnt fluctuate. because it IS fluctuating I would think it may actually be reading something. Now it may not be reading correctly, its possible that instead of the pump flowing at 3000 rpm its actually 300, but its unlikely that its not reading anything at all.

 

Also, the coolant temp is measured in the pump head. If there were no flow at all it would be likely that the coolant temp would register as being incredibly hot. But its not, only 3C higher under load than idle. Again, flow rate might be slow, but it is probably flowing at least some.

 

Oh, and I have an H115i (basically the same thing), and I dont hear or feel my coolant flowing at all. And mine is definitely working just fine. And it also reads similar pump RPM, so that lends a little more credence to my theory...seems unlikely that the failure would happen in the exact way that it should function normally...

 

For this reason, and because he said temps rise almost instantly, I would think it would be a thermal compound issue. Even with a dead pump there should be a small amount of time before the coolant in the pump head increases in temp. where cooling should function more or less ok, for at least a few seconds.

 

So possibly your application method sucks (check by seeing what the coverage is like when you take the block off). But since application method doesnt have a HUGE bearing on temps (unless its really way way off), more likely, your thermal compound sucks.

[Origami Cactus]

2 hours ago, AriGee50 said:

 

Well yeah but his thermal paste worked fine for a few months, and then suddenly failed? That seems like an unlikely behaviour for a thermal paste. Also bad thermal paste should be max 20c warmer, not instant 100c territory. 

Maybe when he remounted the cooler he mounted it with not enough strenght, so it came loose.

[AriGee50]

18 minutes ago, Origami Cactus said:

Well yeah but his thermal paste worked fine for a few months, and then suddenly failed? That seems like an unlikely behaviour for a thermal paste. Also bad thermal paste should be max 20c warmer, not instant 100c territory. 

Maybe when he remounted the cooler he mounted it with not enough strenght, so it came loose.

thats possible. happened to me fairly recently actually. I reseated the cooler and must not have put it tight enough then later I moved some stuff  in the case and it must have been enough to move the pump head a bit and suddenly I saw a 15C increase in temps even though I had a cleaner radiator...

[Fwerimuxe]

2 hours ago, AriGee50 said:

 

I would agree that it could be the flow sensor just being dumb and not recognizing...except that in the two pictures, the flow rate is measured at two slightly different rates. If it was the sensor in the pump that was also Fked up and therefor reading incorrectly, it would be more likely that the flow rate would always be listed as the exact same number, it probably wouldnt fluctuate. because it IS fluctuating I would think it may actually be reading something. Now it may not be reading correctly, its possible that instead of the pump flowing at 3000 rpm its actually 300, but its unlikely that its not reading anything at all.

 

Also, the coolant temp is measured in the pump head. If there were no flow at all it would be likely that the coolant temp would register as being incredibly hot. But its not, only 3C higher under load than idle. Again, flow rate might be slow, but it is probably flowing at least some.

 

Oh, and I have an H115i (basically the same thing), and I dont hear or feel my coolant flowing at all. And mine is definitely working just fine. And it also reads similar pump RPM, so that lends a little more credence to my theory...seems unlikely that the failure would happen in the exact way that it should function normally...

 

For this reason, and because he said temps rise almost instantly, I would think it would be a thermal compound issue. Even with a dead pump there should be a small amount of time before the coolant in the pump head increases in temp. where cooling should function more or less ok, for at least a few seconds.

 

So possibly your application method sucks (check by seeing what the coverage is like when you take the block off). But since application method doesnt have a HUGE bearing on temps (unless its really way way off), more likely, your thermal compound sucks.

The reason why the coolant was still at 37c ish was because the temp rised too fast so I turned it off b4 it fried my CPU. If it was left runnung for longer like when I first discovered it, it'll reach abt 44c +

[Fwerimuxe]

24 minutes ago, Origami Cactus said:

Maybe when he remounted the cooler he mounted it with not enough strenght, so it came loose.

I've remounted the cooler multiple times with varying tightness. Currently its firm (not loose no shaking) but not to say overtightened till the board feels pressured.

[Origami Cactus]

48 minutes ago, Fwerimuxe said:

The reason why the coolant was still at 37c ish was because the temp rised too fast so I turned it off b4 it fried my CPU. If it was left runnung for longer like when I first discovered it, it'll reach abt 44c +

Then i would sill remain with the dead pump theory.

[Fwerimuxe]

Here's the coolant temp slowly rising up just by using Wallpaper Engine lmao

@Origami Cactus @mxk. @aezakmi @AriGee50

Further proof of dead pump or does this mean smth else

 

[wildgg]

i've seen from other people and had personal pumps die slowly on me from corsair units.

 

another one bites the dust i guess

[AriGee50]

40 minutes ago, Fwerimuxe said:

Here's the coolant temp slowly rising up just by using Wallpaper Engine lmao

@Origami Cactus @mxk. @aezakmi @AriGee50

Further proof of dead pump or does this mean smth else

 

 

when it was running crazy hot, do you know what kind of coolant temps you were getting after a couple minutes? or did you not let it stay like that for that long?

 

with that kind of coolant temp i think it probably suggests a slow moving pump if not a dead one

 

Look. a tube of arctic MX-4 is cheap and its pretty good. or a tube of arctic silver 5 is a few bucks more but still under $10. you can have either delivered in a couple days from amazon. you could try that (you probably want something better/newer than what you have anyway) and you dont lose much if it doesnt work. its easy to reapply and will take less time and be less of a headache than RMAing your unit. if it doesnt work, you can always RMA it after that.

[Fwerimuxe]

1 hour ago, AriGee50 said:

when it was running crazy hot, do you know what kind of coolant temps you were getting after a couple minutes? or did you not let it stay like that for that long?

Here, I recorded a video of the temperature rising when just running CSGO. It rises slowly and drops slowly. The temperatures of the CPU itself are almost as hot as usual.

 

[AriGee50]

Yea a coolant temp of nearly 50 degrees, that doesnt seem affected by fan speed(though fans werent running high that long, maybe not long enough to be affected anyway, but still), in a case with an ambient temp of 28C (or 22...or 15...idk your board is weird...) doesnt make sense. I doubt its not moving at all but its probably moving very slowly...RMA it...