Weird problem with my h100i GTX (purchased about an year ago)

From last 15 days, I have been facing this weird issue. My H100i starts giving me high idle (anywhere between 45-60C) and load temps at any time. I have tried cleaning the rad, fans, re-applied TIM, tried it as case exhaust and case intake. The weird thing is whenever I install it, most of the time, it works fine for about 24-36 hours.

Two days back, I cleaned the rad, installed it back, it worked fine for 36 hours, then after a restart, it started giving me high temps again. Then I just turned off the PC and turned it back on next day morning, it worked fine till the end of the day and then again high temps. For now, I have just installed the stock cooler.

Problem is that if I go for RMA and if it works fine when they check it, I will just be stuck with this faulty H100i.

Processor is i7-4790k and I am from India. Forgot to add that I have tested it on 2 motherboards.

I just posted it on reddit also. Any suggestions or help would be highly appreciated. Thank!

I can see why you would be reluctant to RMA the unit. It's going to be a long trip from India and take some time. However, by removing the H100i GTX, you are no longer in a position to troubleshoot the problem and what you have given us is not enough.

The key value is determining a problem with the cooler from a problem with something else is the H100i GTX Temp in Link. This is your water temperature inside the cooler. It should start off about the same as your room temperature at cold boot, warming up a few degrees as the PC comes to normal idle. After that, it's value should be slow to change. The most common sign you have a problem with the cooler is a slow and continual water temp rise that never cools off.

It's odd the symptoms are sporadic. Usually, a pump or cooler issue is consistent. Look for any unexpected changes in pump speed. Make sure the fan control for the CPU_FAN header (or wherever the H100i GTX is attached) is set to 100%, PWM, disabled, or in some way that is provides 12v at all times. Don't use fan controls on that header.

We need your CPU frequency and voltage to interpret temperatures. Room temperature will also play a large factor in your idle temps, so mention that as well when needed.

I can see why you would be reluctant to RMA the unit. It's going to be a long trip from India and take some time. However, by removing the H100i GTX, you are no longer in a position to troubleshoot the problem and what you have given us is not enough.

 

The key value is determining a problem with the cooler from a problem with something else is the H100i GTX Temp in Link. This is your water temperature inside the cooler. It should start off about the same as your room temperature at cold boot, warming up a few degrees as the PC comes to normal idle. After that, it's value should be slow to change. The most common sign you have a problem with the cooler is a slow and continual water temp rise that never cools off.

 

It's odd the symptoms are sporadic. Usually, a pump or cooler issue is consistent. Look for any unexpected changes in pump speed. Make sure the fan control for the CPU_FAN header (or wherever the H100i GTX is attached) is set to 100%, PWM, disabled, or in some way that is provides 12v at all times. Don't use fan controls on that header.

 

We need your CPU frequency and voltage to interpret temperatures. Room temperature will also play a large factor in your idle temps, so mention that as well when needed.

Yes, that's one of the reasons why I want to be sure before going for a RMA. It is going to take more than a month. I didn't add all the details in the main post, sorry for that. I remember almost all the details and I just removed the water block, the cooler is still inside the case, so putting it back won't be much of a hassle.

Room temperature usually remains 30-31C here. As you said, yes, after a cold boot I see liquid temp rising from 31C. When the cooler works fine, the liquid temp stabilizes at usually 33-35C. But when it doesn't, temp rises faster than usual and remains between 39-42C. I have been running the cooler mostly at performance mode. All the fans and pump readings in CL appear to be fine all the time, i.e., whether the cooler is working fine or not.

Edit: I had to undervolt it to keep the cpu load temps in below 80. So that won't be of much help here. But before the problem started, I was running the i7-4790k on 4.6 GHz at 1.23V on a MSI Z97 Gaming 5 board.

Room temperature usually remains 30-31C here. As you said, yes, after a cold boot I see liquid temp rising from 31C. When the cooler works fine, the liquid temp stabilizes at usually 33-35C. But when it doesn't, temp rises faster than usual and remains between 39-42C. I have been running the cooler mostly at performance mode. All the fans and pump readings in CL appear to be fine all the time, i.e., whether the cooler is working fine or not.

 

Edit: I had to undervolt it to keep the cpu load temps in below 80. So that won't be of much help here. But before the problem started, I was running the i7-4790k on 4.6 GHz at 1.23V on a MSI Z97 Gaming 5 board.

More information is needed to see what is causing the actual issue with high temps.

First, you need to remove the overclock and set UEFI/BIOS to default settings before any testing.

Second, make sure the TIM installation has the correct amount material and waterblock was mounted correctly & not make poor contact.

Third, use either Aida64 or Intel Burn Test to obtain system temperatures.

If you need some help with TIM installation or mounting the waterblock, take a look at my guide in the Asetek section. Here's the link: http://forum.corsair.com/v3/showthread.php?t=139165

Report back with the findings

Room temperature usually remains 30-31C here. As you said, yes, after a cold boot I see liquid temp rising from 31C. When the cooler works fine, the liquid temp stabilizes at usually 33-35C. But when it doesn't, temp rises faster than usual and remains between 39-42C. I have been running the cooler mostly at performance mode. All the fans and pump readings in CL appear to be fine all the time, i.e., whether the cooler is working fine or not.

 

Edit: I had to undervolt it to keep the cpu load temps in below 80. So that won't be of much help here. But before the problem started, I was running the i7-4790k on 4.6 GHz at 1.23V on a MSI Z97 Gaming 5 board.

Most of that is as expected. Unfortunately, when your room temps start off at 30C+, you are always going to be 10C higher than what other people post around the web. There is nothing you can do about that part. If your warm idle water temp stabilized around 33-35C in a 30-31C room, then at that moment the cooler is working well and your case set-up is good. Keeping an internal case temperature only 2-4C above the room temp is excellent. You certainly could see water temperatures at 40C with any kind of moderate load, including gaming. I assume from your description the idle water temperature suddenly jumps to 40C. This is certainly possible after extended moderate load running. It can take 5-10-15 minutes for the water to cool back down, depending on how much load you applied and for how long. For now, I will assume this is not the scenario you are describing.

My interpretation of your post is the change in water temp is sudden and after rebooting was right back at the expected temperature. The first thing that comes to mind is an erratic pump. Take note of its normal RPM. Then if this condition returns, check it again.

Are there any other hardware components (GPUs) that could contribute to warmer than expected temperatures? Exactly what conditions/loads are in place when this happens? Random?

Most of that is as expected. Unfortunately, when your room temps start off at 30C+, you are always going to be 10C higher than what other people post around the web. There is nothing you can do about that part. If your warm idle water temp stabilized around 33-35C in a 30-31C room, then at that moment the cooler is working well and your case set-up is good. Keeping an internal case temperature only 2-4C above the room temp is excellent. You certainly could see water temperatures at 40C with any kind of moderate load, including gaming. I assume from your description the idle water temperature suddenly jumps to 40C. This is certainly possible after extended moderate load running. It can take 5-10-15 minutes for the water to cool back down, depending on how much load you applied and for how long. For now, I will assume this is not the scenario you are describing.

 

My interpretation of your post is the change in water temp is sudden and after rebooting was right back at the expected temperature. The first thing that comes to mind is an erratic pump. Take note of its normal RPM. Then if this condition returns, check it again.

 

Are there any other hardware components (GPUs) that could contribute to warmer than expected temperatures? Exactly what conditions/loads are in place when this happens? Random?

As I said earlier, this issue is totally random, i.e., it works fine sometimes. I see high idle and load temps when it doesn't work fine. When I said liquid temps remain 33-35C on 30-31C room temp, that time it works as it should. Liquid temps don't rise easily at that time. But when the problem starts, the liquid stabilize at around 38-42C and cpu idles at 48-60C.

If I leave the PC off overnight and turn it back in the morning, then it again starts working normal for atleast 10-12 hours or may be longer sometimes.

Anyway, I just installed the cooler back and here are idle temps and load temps with wPrime. Package is idling at 48-51C. Note take I took idle temps after leaving the PC idle for more than half an hour, although that was not required. Check the images on the link below.

http://imgur.com/a/sQSwJ

Same wPrime test used to give me 61C load temp earlier at the same room temperature. Since this issue is random, I am a little bit hesitant.

Edit- Forgot to add that the processor was running at 4GHz@1.03V. My stock cooler gives 43-45C idle temps at same room temp whereas this H100i GTX used to give me 39-42C idle temps earlier.

More information is needed to see what is causing the actual issue with high temps.

 

First, you need to remove the overclock and set UEFI/BIOS to default settings before any testing.

 

Second, make sure the TIM installation has the correct amount material and waterblock was mounted correctly & not make poor contact.

 

Third, use either Aida64 or Intel Burn Test to obtain system temperatures.

 

If you need some help with TIM installation or mounting the waterblock, take a look at my guide in the Asetek section. Here's the link: http://forum.corsair.com/v3/showthread.php?t=139165

 

Report back with the findings

Since, I started having this issue, I have been running the processor at 3.6 or 4 GHz. Just tried Aida64 while keeping my processor at 4GHz@1.03V, temps reached 94C in about 20-25 seconds.

How fast did it get to 94C? In a second or two? Or did it start at 60C or so and ramp up over the 20-25 seconds?

Did the water temperature (H100i GTX Temp) change at all? It should only move 1-2C max in 20 seconds and then should have dropped down as soon as you stopped the test. If it went up quite a bit and stayed, that does suggest a flow problem.

How fast did it get to 94C? In a second or two? Or did it start at 60C or so and ramp up over the 20-25 seconds?

 

Did the water temperature (H100i GTX Temp) change at all? It should only move 1-2C max in 20 seconds and then should have dropped down as soon as you stopped the test. If it went up quite a bit and stayed, that does suggest a flow problem.

It started with 69C and ramped up to 94C in about 25 seconds just like temps during Aida64 stability test rise usually. Liquid temp reached ~39C.

I am afraid this means there is a problem with the cooler. My guess is the pump is shutting off for random intervals. This will get worse, so I would put your temporary cooler on and start the RMA process. Sorry.

If it works fine when they receive it and test it, I will probably be stuck with this one, right?

Usually, they ship you a replacement model as soon as they receive yours -- not fix your cooler and return it to you. There might be exceptions for older or discontinued products, but the the H100i v2 is not on that list.

When you fill out the RMA information, specifically state that "when load is applied, the CPU temps start of at one temperature and continually climb until the throttle point". This is the clear mark of a blockage or failing pump. Aggressive voltage and tough stress test programs don't make your temps slowly ramp up. The voltage is on or off, hot or not. Corsair is very friendly about this stuff and are not trying to make this hard.

Usually, they ship you a replacement model as soon as they receive yours -- not fix your cooler and return it to you. There might be exceptions for older or discontinued products, but the the H100i v2 is not on that list.

 

When you fill out the RMA information, specifically state that "when load is applied, the CPU temps start of at one temperature and continually climb until the throttle point". This is the clear mark of a blockage or failing pump. Aggressive voltage and tough stress test programs don't make your temps slowly ramp up. The voltage is on or off, hot or not. Corsair is very friendly about this stuff and are not trying to make this hard.

Alright! Thanks for the help. RMAs are handled by the distributor here. I am not sure if they handle/replace them in India themselves or send them to Corsair.

Alright! Thanks for the help. RMAs are handled by the distributor here. I am not sure if they handle/replace them in India themselves or send them to Corsair.

OK, that may be different and it sounds like you are returning through your retailer, not Corsair directly. I am not sure how that will affect the process. Read the retailer's terms carefully.