Jump to content
Corsair Community

H100i V2 loose fit and inconsistent performance


lmn
 Share

Recommended Posts

TL;DR at the bottom

 

I have recently switched motherboards from the Asus E3-Pro-V5 to the TUF Z270 Mk.1. I noticed some strange performance from my H100i V2 with my 6700k which had previously worked perfectly with my old motherboard for a year+. This ranged from hitting 90+ degrees Celsius under load to 50-60, within the course of a day with very few changes.

 

I looked around and found some boards saying to put rubber washers between the backplate and motherboard, which i did using some hard drive shock absorber thingys. I used one layer which significantly reduced the protrusion of the backplate past the motherboard, once the standoffs were installed it moved a little but not as much. I screwed the pump in tightly to make sure the cold plate made good contact, as well as added a liberal amount of thermal paste.

 

My question is- Is adding the washers a real fix or is it more temporary? From what I've seen it seems like the cooler has become more loose over time which might be from the rubber washers. The temperatures have been incredibly random so I dont really know what to make of it, but what I do know is that the temperatures are great directly after a re-install and

re-tightening of the cooler.

 

TL;DR: Inconsistent performance with H100i V2 and 6700k, added rubber washers to reduce movement of backplate as suggested in forum posts and videos but performance generally decreases as time goes on, is this because of the washers? Is the cooler becoming looser?

 

Relevantish Info:

-6700k is at stock clock, 4.2 Ghz, and I manually set the voltage to 1.2V in BIOS. This voltage gives me good thermals and is generally fine for stock so I dont think it's an issue.

-Corsair Link reports that the pump is running at 3000rpm and I can feel the heat when I touch the tubes.

- I have 3 intake fans and the radiator is in a push config so I don't think that airflow is a problem.

- I'm using CPU-Z and RealTemp to monitor temps, voltage, frequency, and load, as well as Prime95 to stress it.

 

Thanks for any help!

Link to comment
Share on other sites

You really shouldn't need to use washers if you are using the correct standoffs. Yes, the backplate is loose against the motherboard but that tightens up when you mount the cooler.

Two comments:

1) You don't need "a liberal amount of thermal paste". You only need about a lentil's worth. Or a grain of rice. Your best connection is direct metal-to-metal (heat spreader to cold plate). The thermal paste is only intended to fill in any small (like microscopic) spaces as well as the space due to the slightly concave heatspreader on Intel CPUs. That said, too much thermal past shouldn't impact your cooling performance more than a couple of degrees and you are seeing more than than.

2) You need to make sure that the fan header that the pump is connected to (usually the CPU Fan header) is set to 100%. On the Asus boards, this means that you disable Q-Fan Control. You can do this in the advanced section of the BIOS. Alternatively, you can just set it to 100% across the board. This can significantly impact your thermal performance since, if you don't do that, the pump won't get the full 12V that it needs to operate. Over time, it can also damage your pump.

Link to comment
Share on other sites

You really shouldn't need to use washers if you are using the correct standoffs. Yes, the backplate is loose against the motherboard but that tightens up when you mount the cooler.

Two comments:

1) You don't need "a liberal amount of thermal paste". You only need about a lentil's worth. Or a grain of rice. Your best connection is direct metal-to-metal (heat spreader to cold plate). The thermal paste is only intended to fill in any small (like microscopic) spaces as well as the space due to the slightly concave heatspreader on Intel CPUs. That said, too much thermal past shouldn't impact your cooling performance more than a couple of degrees and you are seeing more than than.

2) You need to make sure that the fan header that the pump is connected to (usually the CPU Fan header) is set to 100%. On the Asus boards, this means that you disable Q-Fan Control. You can do this in the advanced section of the BIOS. Alternatively, you can just set it to 100% across the board. This can significantly impact your thermal performance since, if you don't do that, the pump won't get the full 12V that it needs to operate. Over time, it can also damage your pump.

 

Thanks for your response,

I've heard that the movement doesn't matter and that it tightens up, and I didn't have to use washers with my other board so my thoughts were that it was some kind of change between boards that made not fit as well, or at least thats what I saw some people say. Do you think that having the washers could actually hurt the performance?

 

As for the thermal paste, I did apply it correctly, only about the size of a grain of rice, I just wanted to emphasize that I used enough. Also, I checked and made sure it was spread evenly across the cooler.

 

I am using Q-Fan in BIOS right now to set fan curves and have the pump connected to the pump header with nothing connected to the CPU fan header. From what I've seen in the BIOS the pump header is set to %100 all the time. Link reports that the pump is running at a constant 3000+ rpm. Is the %100 stat in BIOS not reliable, because it seems like the pump is functioning correctly if it is reported to be 3000rpm in operation.

Link to comment
Share on other sites

Thanks for your response,

I've heard that the movement doesn't matter and that it tightens up, and I didn't have to use washers with my other board so my thoughts were that it was some kind of change between boards that made not fit as well, or at least thats what I saw some people say. Do you think that having the washers could actually hurt the performance?

Honestly, I have a hard time thinking that it does. However ... I can tell you that I had an issue with my cooler at one point as well ... turned out that the standoffs weren't screwed down tight.

As for the thermal paste, I did apply it correctly, only about the size of a grain of rice, I just wanted to emphasize that I used enough. Also, I checked and made sure it was spread evenly across the cooler.

Cool. You said "liberal" and I had images of cake frosting. :eek:

I am using Q-Fan in BIOS right now to set fan curves and have the pump connected to the pump header with nothing connected to the CPU fan header. From what I've seen in the BIOS the pump header is set to %100 all the time. Link reports that the pump is running at a constant 3000+ rpm. Is the %100 stat in BIOS not reliable, because it seems like the pump is functioning correctly if it is reported to be 3000rpm in operation.

The pump header defaults to 100% power so that's good.

 

So ... another question for you ... you've been telling us about the CPU temps. What about the coolant temperature? What does that look like through all of this?

Link to comment
Share on other sites

Honestly, I have a hard time thinking that it does. However ... I can tell you that I had an issue with my cooler at one point as well ... turned out that the standoffs weren't screwed down tight.

 

Cool. You said "liberal" and I had images of cake frosting. :eek:

 

The pump header defaults to 100% power so that's good.

 

So ... another question for you ... you've been telling us about the CPU temps. What about the coolant temperature? What does that look like through all of this?

 

I seem to have that same problem with the stand offs not being tight, did you just tighten them to fix them? It seems that they become loose, maybe after i tighten down the pump because when I install them at first theyre as tight as possible.

 

I'm not entirely sure how to check the coolant temperature- Link reports that the H100i V2 is around 40C with CPU temps being similar if not lower. The tubes are also hot to the touch after the CPU heats up, I dont think its a flow problem.

 

Also, I played Watchdogs 2 for over an hour last night and had temps in the mid to high 50's, which is really good and what I'm looking for. I slept the computer for the night and this morning played another hour or so of Watchdogs 2 and saw temps in the mid to high 60's. These temps aren't bad at all but are a good 5-10 degrees worse than last night, and I feel that I'm not getting the performance out of the cooler for having my CPU at base clock. I'd also like to overclock in the future and having this kind of performance just won't let me do that. Would it be normal to see performance degradation directly after reinstalling the cooler?

Edited by lmn
Link to comment
Share on other sites

I seem to have that same problem with the stand offs not being tight, did you just tighten them to fix them? It seems that they become loose, maybe after i tighten down the pump because when I install them at first theyre as tight as possible.

May have something to do with the washers? Not sure ... I think mine loosened when I removed the pump header.

I'm not entirely sure how to check the coolant temperature- Link reports that the H100i V2 is around 40C with CPU temps being similar if not lower. The tubes are also hot to the touch after the CPU heats up, I dont think its a flow problem.

That's the coolant temperature.

Also, I played Watchdogs 2 for over an hour last night and had temps in the mid to high 50's, which is really good and what I'm looking for. I slept the computer for the night and this morning played another hour or so of Watchdogs 2 and saw temps in the mid to high 60's. These temps aren't bad at all but are a good 5-10 degrees worse than last night, and I feel that I'm not getting the performance out of the cooler for having my CPU at base clock. I'd also like to overclock in the future and having this kind of performance just won't let me do that. Would it be normal to see performance degradation directly after reinstalling the cooler?

Hard to say if you are actually seeing a performance degradation as there are several variables that can come into play - the largest of these is the ambient temperature in the room as well as the interior case temperature (this is actually critical if your cooler is set to exhaust).

So many folks focus on one individual temperature but that's incredibly myopic. You need to view it as an integrated system of interrelated temperatures; no single temperature can be evaluated one its own for "good" or "bad". Cooler performance should be measured on deltas ... the delta between ambient (or interior) and coolant temperature and the delta between CPU and coolant temperature. Consider: if your cooler is set up as exhaust, the GPU temperature will directly impact the case interior temperature, which directly impacts the radiator's ability to cool the liquid coolant, which directly impacts its ability to cool your CPU.

Link to comment
Share on other sites

May have something to do with the washers? Not sure ... I think mine loosened when I removed the pump header.

 

That's the coolant temperature.

 

Hard to say if you are actually seeing a performance degradation as there are several variables that can come into play - the largest of these is the ambient temperature in the room as well as the interior case temperature (this is actually critical if your cooler is set to exhaust).

So many folks focus on one individual temperature but that's incredibly myopic. You need to view it as an integrated system of interrelated temperatures; no single temperature can be evaluated one its own for "good" or "bad". Cooler performance should be measured on deltas ... the delta between ambient (or interior) and coolant temperature and the delta between CPU and coolant temperature. Consider: if your cooler is set up as exhaust, the GPU temperature will directly impact the case interior temperature, which directly impacts the radiator's ability to cool the liquid coolant, which directly impacts its ability to cool your CPU.

 

Thanks for all your responses, I'm pretty sure that the problem has to do with the contact between the coldplate and the heatspreader, I think I'm just going to have to mess around with the mounting or maybe get another cooler. The radiator is set to pull air into the case from the room, which is around 68-70F so I don't think thats necessarily the problem. For perspective, my GPU radiator is set to exhaust from air in the case and is performing fine.

 

My temperatures have been mostly fine after remounting so I think I've mostly fixed the problems.

Link to comment
Share on other sites

It's the next day and I launched Watchdogs 2 again, now I'm getting temperatures in the 70's and my fans are getting loud. I'm going to try and remount it again but I'm getting annoyed and dissapointed.
Link to comment
Share on other sites

Just reinstalled the pump after getting mid 70's in Watchdogs 2, now I'm seeing barely a degree over 50, which is better than I expect and I wouldn't be surprised to get up to 60 in the future. The problem is that I'm afraid I'll get the same kind of performance degradation as last time, and I'll see temperatures in the 70's again as soon as I launch the game.
Link to comment
Share on other sites

OK ... so that's good. Did anything appear loose?

 

Nothing was loose in the sense that it had a lot of movement but I was able to tighten the standoffs semi-easily by hand, which I guess is looser then when I installed them a few days ago. I tightened down the pump only by hand this time since I think I might have over-tightened it with the screwdriver. I'll update this if anything changes but if I don't, just assume it's all running perfectly:biggrin:.

Thanks for all the help.

 

Edit- Just saw that Corsair released two new coolers... Pretty interested in those if I need to get a new one. The H150 looks cool with the triple radiator but I dunno if I'd be concerned about accommodating it in future cases, I know it fits in my current one since it's the one that corsair uses in all of the glam pictures.

Edited by lmn
Link to comment
Share on other sites

 Share

×
×
  • Create New...