Jump to content
Corsair Community

The most insane and hilarious H115i Problem you'll see in 2019


Shayd
 Share

Recommended Posts

Alright boys I know 2019 JUST started but here is the short problem. If I plug in the Mini-USB into USB Header on the motherboard (For Corsair Link) The CPU immediately heats up into 70C Range and keeps rising to 90C within 30 Seconds even though the load is idle, Rebooting with the USB Cable UNPLUGGED Gives me 35-40C Idle temps.

 

Long Version Here:

 

This is NOT a new build, It's been assembled in 2016 or so and the pump has been in place there for a while working very well, but time flies and few days ago when I switched out my 1060 to RTX2080Ti For that Maximum 720P 61FPS in Fortnite and decided to take whole PC apart and clean everything, Everywhere I could I used small blowed to air any dust out, Fans and case was wiped super clean and pump was also taken off to de-dust2 and re-apply new paste because...well the block was taken off. Everything went perfect and system was assembled but the problem showed up immediately that as soon as I boot up the PC even into BIOS the CPU (i7-6850K) Immediately rises to 60-70C and sometimes even higher (90C Max) at 1C per SECOND. I triple Checked the Fans, the pump and cooling over all as well, and indeed the CPU genuinely heats up and the Corsair can NOT keep up with this in any way but if I Unplug the little Micro-USB from the pump or from the motherboard header (NOT SATA POWER) and then boot up I'm welcomed by what is about 35-40C Idle Temps and about 60C When Playing ARK:Extinction. I have confirmed this is the issue. Boot up with pump connected by USB and have heater or boot with it unplugged and have it work perfectly aside from not being able to switch profiles or adjust fan curves. Now it's very important to mention that plugging the Pump IN while system is running makes it appear on Corsair Link BUT has NO effect on the temperature until you reboot! Unplugging it from motherboard while the CPU is overheating does not fix the problem, literally the second where Motherboard posts whether the USB is attached or not decides if the CPU will run or overheat insanely.

 

This is not possible but yet it exists, This has not been a problem before, the system has not been modified in any way including software other than RTX 2080ti was installed before and after the problem and I have plugged everything as it was. I already put clean Windows 10 and Tried USB Header Splitter but did not work. The

fan curve of preset does not change so that's not the issue. Screenshots Below

 

 

With Pump USB Unplugged

https://i.imgur.com/5gY0seW.png

 

After Hot Plugging Pump Back in

https://i.imgur.com/TeQMyaD.png

 

Leaving the Pump Plugged IN and then REBOOTING While it's plugged IN, THIS is the Issue even 7 Minutes after boot up the temp has barely went down from ~78C and will not go down further, attempt to playing game at this state makes CPU reach insane 80-90C UNLESS You unplug the USB and reboot

https://i.imgur.com/TnWy8TT.png

Edited by Shayd
Link to comment
Share on other sites

If you boot into the BIOS and the system is already overheating, you have a hardware problem with the cooler. Assuming this the 2016 era 5th generation Asetek "H115i", this type of problem is frequently mentioned after the removal and cleaning process. Best guess is you stirred it up and a bit and the debris that should not be there is now floating around, clogging the pump strainer. It might be a bit premature to call this one done considering the inconsistent nature. If you can, try booting up with the cooler connected so you can use Link, then toggle the pump back and forth between high and low every 10-15 seconds. See if that dislodges the blockage. You can also try carefully rocking the case a bit while the the pump is running. If it doesn't improve or immediately drop, you can also try things like squeezing the tubes to alter the momentary pressure and see if that does it. I don't really consider these tricks permanent solutions and I would still look at getting a new cooler, but it would offer some evidence as to the source of the problem. The other key thing is when you coolant temp is 36-43C, the exhaust from the H115i radiator should also be that same temp. That should feel pretty warm. If it does not and it's cool, then you know this is a fairly substantial blockage and the game is over.

 

I am not sure what to make of the hot plugging/restart stuff. It could be a coincidence in terms of the whatever physical ailment is affecting the cooler or it may be related to the inability to control the cooler when hot plugged. I would focus on the physical sign first. It is an easier thing to prove it physically dysfunctional to Corsair support and get the RMA rolling than to sort through what may or may not be some type of software/firmware anomaly.

Edited by c-attack
Link to comment
Share on other sites

The radiator exhaust definitely feels warm but only at the spot where the tube pumps the hot coolant in, as you move your hand the air feels cooler. It's not problem of exhausting the heat but maybe absorbing it, The Pump RPM Seems fine but I have no way of telling if something could be gunking the flow reducing it. I though the whole point of AIO Cooler was to avoid the maintenance? Yes I probably moved and rotated it during cleaning but how did dirt get inside?
Link to comment
Share on other sites

Dirt did not get inside and you did not do this. Usually these type of clogs are caused by anti-corrosion elements and adhesive/sealant from the tubes coming together - something not intended and obviously not desirable. It may have been resting somewhere of no consequence, but became mobile during the normal moving and then found its way to a more vulnerable (and narrow) spot. This particular series appears to have a high incident rate and it is not an often reported problem on other Corsair coolers or other across the market. If your assessment is accurate, it suggests the blockage in in the radiator channel, which is no so common. Normally I would suggest you call it day, contact Corsair for a RMA, or go shopping, or both. However, given the apparent location, it might be worth trying to work it out first.

 

Try the following:

 

1) If you can safely tip your case over on its side, do so, with the system running. Listen for sound changes (bubble rush), weird pump noises, and watch to see what happens with the coolant temp. The hope is the blockage with dislodge and float to a non-critical location and thus restoring normal flow.

 

2) Take the radiator down and shake it. Decidedly non-scientific, no guarantee for success, but if it doesn't work the next step is going to be replacement anyway.

 

If this doesn't work, there is no user acceptable remedy. I know people who have ripped it apart in frustration to find the offending material, but your chances of putting it back together in good working order are slim. There are a number of features that make the process one directional. Obviously you forfeit the warranty at that point and even if you did get it back together, the strongest chance of leak would then always be from the cold plate itself (through the screws). That is not something you can see and the location insures maximum damage. I strongly advise against the tear down, except for purposes of vengeance.

Link to comment
Share on other sites

C-Attack might be onto something here, I did not flip or turn the case here but I noticed the lack of pattern with my USB plugging as today morning I turned the PC On and temps went super high despite the Cable being not plugged in, I tried C-Attack's advice on Turning the Pump between 'Quiet' and 'Performance' but not until I switched between those modes every 2 seconds to cause as much flow disturbance as I could for about 20 seconds before leaving it on Performance mode the temps magically went from about 68 to 39C Which supports his idea about some stuff floating around until you turn off the PC for a while longer and the debris settles. Sadly this does not solve the issue because if you reboot the PC it it still fine but turning it off and on after let's say as soon as 1 Hour the temps go back up and you have to switch the bump 20 times to get the temps back down. I'm nowhere as good with Components to try and take the pump apart so I guess I will simply try to contact Corsair and RMA it. It's funny to be how maintenance to keep temperatures cool made them go up.
Link to comment
Share on other sites

OK, well in some ways this actually good news. Regardless of what's causing it, the end result is a non-satisfactory cooler and it needs to be replaced. Corsair will do that or you can pick something new you prefer, or both. Your description of the heat flow near the radiator gave me pause. Typically the block is at the pump and this causes heat to back up along both tubes near the pump, but not near the radiator at all. That opened the door to the possibility of this being air bubbles created during the move. That would be something resolvable through the case shifting method. However, once you dislodge them, they typically do not come back and you don't see the hot/normal/hot/normal cycling on every boot. That is indicative of the original described issues with some sort of gunk blocking critical flow points, settling, then landing again at the wrong place. It may come and go and you can do these tricks short term, but obviously you can't get it out and thus it will remain a problem. Typically it gets worse, not better. That is the cruel irony with this issue and most people start their post with "I just finished cleaning my radiator and fans and when I put it back on. :( ".
Link to comment
Share on other sites

OK, well in some ways this actually good news. Regardless of what's causing it, the end result is a non-satisfactory cooler and it needs to be replaced. Corsair will do that or you can pick something new you prefer, or both. Your description of the heat flow near the radiator gave me pause. Typically the block is at the pump and this causes heat to back up along both tubes near the pump, but not near the radiator at all. That opened the door to the possibility of this being air bubbles created during the move. That would be something resolvable through the case shifting method. However, once you dislodge them, they typically do not come back and you don't see the hot/normal/hot/normal cycling on every boot. That is indicative of the original described issues with some sort of gunk blocking critical flow points, settling, then landing again at the wrong place. It may come and go and you can do these tricks short term, but obviously you can't get it out and thus it will remain a problem. Typically it gets worse, not better. That is the cruel irony with this issue and most people start their post with "I just finished cleaning my radiator and fans and when I put it back on. :( ".

 

I'm gonna contact Corsair tomorrow once their Offices open I'm just not sure how the hell to even describe this in the ticket.

Link to comment
Share on other sites

You don’t have to explain the underlying cause to them. Just stick to facts. The cooler is alternating between normal and poor performance (high liquid temperature). When the liquid temp is high, the radiator exhaust is cool. The pump speed reads 3000 all the time and the fans are moving. Liquid temp reaches 43C+ with no load. They may ask you some basic questions related to how the pump is powered, but that not relevant on the SATA powered H115i when there is no evidence of power loss. That’s really all you need to say. Upload a copy of purchase invoice to establish the warranty start point. If you prefer, they can do an advance RMA where they send you the new unit on a credit card hold. You ship the one back after and the charge is dropped. That should cut the time in half and keep you semi-functional in the interim.

 

*Be sure to describe the CPU temp slowly, but continually escalating when you entered the BIOS. That is a quick way to eliminate questions about your usage, power plans, settings, other heat sources and the like. When this happen, you essentially need to verify the cooler has power. If so, then there is a pump fail or blockage.

Edited by c-attack
Link to comment
Share on other sites

You don’t have to explain the underlying cause to them. Just stick to facts. The cooler is alternating between normal and poor performance (high liquid temperature). When the liquid temp is high, the radiator exhaust is cool. The pump speed reads 3000 all the time and the fans are moving. Liquid temp reaches 43C+ with no load. They may ask you some basic questions related to how the pump is powered, but that not relevant on the SATA powered H115i when there is no evidence of power loss. That’s really all you need to say. Upload a copy of purchase invoice to establish the warranty start point. If you prefer, they can do an advance RMA where they send you the new unit on a credit card hold. You ship the one back after and the charge is dropped. That should cut the time in half and keep you semi-functional in the interim.

 

*Be sure to describe the CPU temp slowly, but continually escalating when you entered the BIOS. That is a quick way to eliminate questions about your usage, power plans, settings, other heat sources and the like. When this happen, you essentially need to verify the cooler has power. If so, then there is a pump fail or blockage.

 

Corsair Hasn't replied yet but while waiting i also found one interesting thing to look out in the future when diagnosing pump

https://imgur.com/a/QoinTjX You can see two screenshots, first after booting up the second one after me power cycling the pump to 'unclog' it, notice how unstable the speed of it is at first

Link to comment
Share on other sites

Be careful about reading too much into the visual nature of the graph or assessing pump rpm too literally. Pump RPM is an estimate. A polling interval (say 2 seconds) is sampled, then the number of resolution multiplied out (in this case by 30) to make revolutions per minute. Simple enough, except the rather key criteria is what to do with partial revolutions (multiply or round) and how accurately can it measure or assess partial revolutions. So you get a string of 99.875 x 30, 99.95 x 30, 100.05 x30, 99.90 x 30 and on and on --- all implying changes in speed, but the tolerances are so small it is may be below the measurable tolerance with differences multiplied out into something tangible. You can see the opposite approach in the new Pro coolers that round the value prior to multiplying out to make 1 minute intervals. This creates very even RPM numbers, always 2160, 2130, or 2190. Frankly I think I prefer this way and it is less noticeable, but really there is no physical difference between what's happening between the two pumps.

 

Also note the scale in your two shots. Link has a dynamically adjustable range. In one shot you span almost 1300 rpm, so changes of 50 rpm or so are going to look small. In the second shot there is only 120 rpm of range, so even small variances look like mountains. This has caught me out countless times and my eyes leap to a series of jagged peaks, only then to notice the changes are tiny.

Link to comment
Share on other sites

Be careful about reading too much into the visual nature of the graph or assessing pump rpm too literally. Pump RPM is an estimate. A polling interval (say 2 seconds) is sampled, then the number of resolution multiplied out (in this case by 30) to make revolutions per minute. Simple enough, except the rather key criteria is what to do with partial revolutions (multiply or round) and how accurately can it measure or assess partial revolutions. So you get a string of 99.875 x 30, 99.95 x 30, 100.05 x30, 99.90 x 30 and on and on --- all implying changes in speed, but the tolerances are so small it is may be below the measurable tolerance with differences multiplied out into something tangible. You can see the opposite approach in the new Pro coolers that round the value prior to multiplying out to make 1 minute intervals. This creates very even RPM numbers, always 2160, 2130, or 2190. Frankly I think I prefer this way and it is less noticeable, but really there is no physical difference between what's happening between the two pumps.

 

Also note the scale in your two shots. Link has a dynamically adjustable range. In one shot you span almost 1300 rpm, so changes of 50 rpm or so are going to look small. In the second shot there is only 120 rpm of range, so even small variances look like mountains. This has caught me out countless times and my eyes leap to a series of jagged peaks, only then to notice the changes are tiny.

 

I don't mean the scale of the graph I mean how stable the speed is after the pump cycle. Corsair STILL hasn't replied

Link to comment
Share on other sites

I don't mean the scale of the graph I mean how stable the speed is after the pump cycle.

 

I know. That was my point. The speed variability isn't any different between those two shots. The visual difference is a product of the vast difference in the two ranges.

Link to comment
Share on other sites

See here >

 

Basically do you have anything else in iCue (ie psu/keyboard/mouse/headphones - particulary PSU that is plugged into a usb header)

 

 

I too suffered similar problems - however I replaced h100i v2 with h100i plat, and an old psu for a new rm1000i psu - both of which had problems - not some random unplug re plug now not working problem.

 

http://forum.corsair.com/forums/showthread.php?p=985499#post985499

Link to comment
Share on other sites

 Share

×
×
  • Create New...