Is My AIO Faulty? (H100i v2)

[Biggs87]

Hi, I just have a quick question regarding my AIO.

 

I'm currently running an i7 7700k at stock and an undervolt of -0.070v with the H100i v2 and I can't seem to keep my temps under control.

 

My idle temps are OK around 37C but under load I'm getting 88-91C

 

I also get quite a big temperature difference between Quiet Pump speed (200RPM) and Performance Pump speed (300RPM). It's around a 10C difference.

 

Quiet Pump Speed = 88-91C

Performance Pump Speed = 78-81C

 

I'm not changing the fan speed, I'm keeping that consistent throughout my testing, I'm only changing the pump speed using Corsair Link.

 

I understand that the 7700k it is a very hot chip, that is why I decided to undervolt it. But I'm a little surprised with an undervolt, at stock speeds and on quiet pump speed, I'm still hitting 90c+.

 

I did a bit of research and this doesn't seem to be a common issue and other people seem to be getting very solid temps with their H100i v2 even when overclocking. They also suggest that there is no more than a few degrees difference when changing pump speed.

 

Is this normal or is my AIO faulty?

[Hockster2]

First thing to check is that the block is mounted correctly.

 

What speeds are the fans running at? Do you have enough airflow?

[Biggs87]

First thing to check is that the block is mounted correctly.

 

What speeds are the fans running at? Do you have enough airflow?

 

Hey, thanks for the reply.

 

My fans run around 1000-1200 (balanced mode) at all times because my Liquid temps are usually around 28-31c at idle with an ambient room temp of 20-23c.

 

My liquid temps do go up to around 39-40C though while gaming if I leave my Pump on quiet mode.

 

For airflow, I have the Define R5 case with 2 140mm bequiet Silent Wings 3 as Intake and 1 bequiet Silent Wings 3 140mm as exhaust.

 

For the AIO, I replaced the default fans with 2 bequiet Silent Wings 3 120mm high speed running in push config.

 

As far as I can tell the block seems to be mounted correctly, there isn't any wobble and the backplate is very tight

[DevBiker]

If your AIO was faulty, you wouldn't need to run a load to experience the issues.

 

So ... a couple of things. First, you should have a fan curve and increase fan speed as the radiator temperature increases. Second, how is your cooler mounted? Is it exhaust? Is it sucking in hot air from the GTX 1080Ti that you have? Because that is one hot GPU and if you don't manage its heat well, it'll actually warm up the coolant. Which is what I suspect is happening here.

[Biggs87]

If your AIO was faulty, you wouldn't need to run a load to experience the issues.

 

So ... a couple of things. First, you should have a fan curve and increase fan speed as the radiator temperature increases. Second, how is your cooler mounted? Is it exhaust? Is it sucking in hot air from the GTX 1080Ti that you have? Because that is one hot GPU and if you don't manage its heat well, it'll actually warm up the coolant. Which is what I suspect is happening here.

 

Hey there

 

The fan curve is currently set to the default Balanced fan curve.

 

Rad is mounted at the top of my case as an exhaust.

 

What would you suggest? Should I move my Rad to the front of the case and have it as an intake?

 

I was considering that, but like you mentioned the 1080ti is a pretty hot Card and I was concerned about those temps increasing too.

[DevBiker]

Putting your radiator as intake should help the issue, definitely.

 

It's also a question of managing the heat from the video card. If you don't do that properly, you're going to have issues and moving the radiator just mitigates the issue that this causes with the radiator - you'll still likely have a really warm case interior. What temperature do your case fans use for their control variable?

[Biggs87]

My Case fans are set to go by my CPU temps using the default fan curve in the bios. They get to about 100% RPM when my CPU hits around 85C

 

Here are some pics if this helps you get a better understanding of my setup, these Pics are taken basically at Idle right now

[Hockster2]

On my previuos build, 8700K, 2 GTX 1080's, I went against convention and used the rear fan as an intake. I had 2 fans in front and the H150i on top exhausting. I didn't have enough intake and found reversing the rear fan helped my temps by 5-8° in typical usage.

 

Edit: Timely, I just had this video popup in my recommended list. Google kinda scares me lol

 

[ame]

[/ame]

[c-attack]

 

I also get quite a big temperature difference between Quiet Pump speed (200RPM) and Performance Pump speed (300RPM). It's around a 10C difference.

 

Quiet Pump Speed = 88-91C

Performance Pump Speed = 78-81C

 

I'm not changing the fan speed, I'm keeping that consistent throughout my testing, I'm only changing the pump speed using Corsair Link.

 

 

This is the part concerning to me. A 2-3C swing between pump speeds? Maybe. 10C usually suggest the coolant is flow restricted in some way. If true, you might see differences at idle as well. I would be interested in your coolant temp changes between idle and load or just normal use. Most people idle with a H100i Temp of +4-7C above their room temp. It looks like that might be true for you. However, I am interested in the load values too. You might see +6-10C depending on load type with mixed CPU/GPU loads often yielding the largest change.

 

On a side note, are those Silent Wing 2 or 3 fans on the radiator?

[DevBiker]

My Case fans are set to go by my CPU temps using the default fan curve in the bios. They get to about 100% RPM when my CPU hits around 85C

 

Here are some pics if this helps you get a better understanding of my setup, these Pics are taken basically at Idle right now

 

Well, by the time your CPU temp gets to that temp, your thermal environment is already well and truly out of control. In a liquid cooled system, the CPU temp really isn't an appropriate control variable for the case fans as the CPU isn't dumping heat into the case.

 

Like I said, you need to manage the heat inside the case. The CPU isn't the source of heat inside the case.

[Biggs87]

This is the part concerning to me. A 2-3C swing between pump speeds? Maybe. 10C usually suggest the coolant is flow restricted in some way. If true, you might see differences at idle as well. I would be interested in your coolant temp changes between idle and load or just normal use. Most people idle with a H100i Temp of +4-7C above their room temp. It looks like that might be true for you. However, I am interested in the load values too. You might see +6-10C depending on load type with mixed CPU/GPU loads often yielding the largest change.

 

On a side note, are those Silent Wing 2 or 3 fans on the radiator?

 

Hi, I do actually see a slight difference at idle when it comes to pump speed.

 

Quiet pump speed CPU Temps are 37-40C and performance is 33-36C.

 

This seemed to be more consistent with other results i saw while googling. It was the 10C difference at load that had me concerned that it might be my AIO

 

I would say my Coolant temp is always around 8-9C above ambient. Right now Room Temp is 20C and my Liquid Temps are 29.5C with just light web browsing.

 

If I was to start gaming for an hour, it would go up to around 37C.

 

When my Room temps are around 23C it does hit 40C on the coolant, I know because my fans all of a sudden start maxing out.

 

I'm using Silent wings 3, the high-speed variant that goes up to 2,200rpm

[DevBiker]

Let's try something super-silly.

 

Let your system cool down. Then do a gaming session with the case open.

[Biggs87]

Let's try something super-silly.

 

Let your system cool down. Then do a gaming session with the case open.

 

Sure I'll give that a go. I'll get back to you with the results

[c-attack]

I am off the opinion you have a slight blockage forming. CPU temps are extremely dynamic and any background Windows task could cause a blip or two. On the other hand, the coolant temp should stay steady at the desktop level with changes occurring over 10-15 minutes and usually only with environmental changes. So when looking for data, check the coolant temp (H100i v2 Temp) changes between the high/low setting. There should be none at idle and extremely minimal at load over the short duration.

 

OK, good. The SW3 120mm are viable. My testing with SW2 were pretty bad and I wanted to be able to calculate that into any coolant temp change analysis.

[Biggs87]

Ok so I ran most of the test this morning and this is what I got.

 

All tests were run with the side panel off and a room temp of 21.5-22C.

 

Idle Quiet Pump Speed

 

 

Idle Performance Pump Speed

 

 

45Min Gaming Session Quiet Pump Speed

 

 

45Min Gaming Session Performance Pump Speed

 

 

Video Encoding 8Min 1080p60 24mps Quiet Pump Speed

 

Video Encoding 8Min 1080p60 24mps Performance Pump Speed

 

 

I didn't really notice much of a difference with the side panel off than the normal temps I get with it on, maybe a 1-2C difference at most.

 

Under heavy load, I was still getting the 8-10C difference when switching Pump speed

 

Hopefully the screenshots make it a bit clearer whether its a Blockage or bad airflow

[c-attack]

The part that bothers me is the drop in coolant temp at idle when switching from low to high pump speed. I switched my H115i Pro from balanced (2100) to Extreme (2850) 5 minutes ago. Change in coolant temp? 0.0C. Not even a tenth.

 

I am of the opinion you have a slight partial blockage forming. It is something known to happen on that particular model. It is not so bad yet that you are overheating, but there is a noticeable decline in performance. So now what to do about it.

 

You can contact Corsair through the ticket support system and request an RMA. They may drag their feet a little since it has not straight up failed. Typically, these things tend to get worse over time, so I would not expect it to magically get better tomorrow. However, if you want to try, you can take it out and shake it, hoping to dislodge the blockage. It may work. It may not. It could get worse. It could get better. Regardless of the outcome, you may want to also consider simply getting a new cooler of your choice. The v2 series is not on my favorites list and I deliberately avoided it for its entire life cycle, which thankfully is now ending. You can also try to RMA to cooler and go shop for a new one, retaining the replacement H100i v2 as a back up or for future use.

[DevBiker]

That would seem to be indicative of a partial blockage. You really shouldn't be seeing those kinds of temperature differences between Quiet and Performance pump speeds. It's not a full blockage - or the temps would be through the roof - but I'd suggest going down the RMA path.