Jump to content
Corsair Community

System runs hot: Sensor readings of AIO (Elite Capellix) get stuck/flatline (this is a post including a lot of research and information)


Recommended Posts

 

Dear community,

Dear @c-attack, @DevBiker, @Corsair Nick, @Corsair Albert@Greybeard,

I'm super frustrated which makes me try different approaches to finally get rid of my issue with iCUE and my Elite Capellix. The issue appears almost every time I use my PC. I'm in contact with Corsair regarding this issue since July 2021 without any fix.

Let me make it clear from the start: I love Corsair products and I love iCUE and all its possibilities. From my point of view Corsair provides the best ecosystem (hardware/software) for RGB. This is what made me choose Corsair and made me pay premium. This post has no intention to insult and is purely based on personal experiences, research and facts.

My system

  • Hardware
    • ASUS ROG Maximus XI Hero (WI-FI) (newest BIOS: 2004)
    • Intel Core i7-9700K
    • 2x Corsair Dominator Platinum RGB (newest Firmware 1.04.28)
    • Nvidia GeForce RTX 3080 Founders Edition (newest BIOS: 94.02.27.00.13)
    • Corsair iCUE H115i ELITE CAPELLIX (newest Firmware 2.6.201)
      • 6x Corsair ML140 RGB-PWM (8-LED Fan Series)
    • be quiet! Dark Power Pro 12 1200W
    • Corsair K70 RGB MK.2 LP (newest Firmware 3.24)
    • Corsair Glaive RGB Pro (newest Firmware 3.31)
  • Software
    • Windows 11 (clean install)
    • iCUE version (clean install): 4.18.209 (newest release)
    • Nvidia driver (clean install): 497.29 (newest driver)
    • HWiNFO: 7.16-4650 (newest release)
      • CorsairLink and Asetek support disabled
    • MSI-Afterburner: 4.6.4.16255 (newest release)
  • Sidenotes
    • There's no other RGB-software installed besides iCUE
    • It happens with and without HWiNFO and/or MSI-Afterburner being installed/running

Settings in iCUE

  • Each components RGB is adjusted based on temperatures
    • Elite Capellix (pump): Based on CPU package temperature
    • Elite Capellix (2 of 6 fans): Based on coolant temperature
    • Elite Capellix (4 of 6 fans): Based on GPU temperature
    • Nvidia RTX 3080 Founders Edition: Based on GPU temperature
    • Dominator Platinum: Based on DIMM temperature
  • Cooling is based on coolant temperature and GPU temperature
    • Pump: Quiet (default cooling preset)
    • 2 of 6 fans: Balanced (default cooling preset based on coolant temperature)
    • 4 of 6 fans: Custom cooling preset based on GPU temperature

The issue

  • While gaming (I believe it didn't happen in idle before) some/all of the sensor readings (coolant temperature and/or fan speeds) of my Elite Capellix randomly get stuck/flatline
    • As a result my PC overheats due to fan speeds won't adjust due to faulty sensor readings
    • RGB is still working

What I/we have tried in order to sort out the issue

  • Software
    • Clean installed Windows 10/11
      • Enabled/disabled fast boot
      • Enabled/disabled USB power saving
    • Clean installed multiple versions of iCUE
      • Enabled/disabled plugins
      • Enabled/disabled software/game integrations
    • Uninstalled HWiNFO
    • Uninstalled MSI-Afterburner
  • Hardware
    • Exchanged Commander Core
    • Exchanged Elite Capellix
    • Installed a powered USB hub (NZXT internal USB Hub [Gen 3])
    • Reset BIOS
      • Enabled/disabled fast boot
      • Enabled/disabled c-states
      • Enabled/disabled legacy USB support
      • Enabled/disabled USB XHCI Hand-off

Debug-Logs

As follows you can find some excerpts of debug logs from iCUE

  • CUELogs/2022-01-02T17-28-34.log
    • (...)

    • 2022-01-02 17:28:38.011 I cue.devices: "Get temperature measurements status failed. Device Commander CORE"

    • 2022-01-02 17:28:38.011 I cue.devices:  Open file with id bragiprotocol::FileId(TemperatureMeasurementsStatus) failed.
        File already opened.

    • 2022-01-02 17:28:38.012 I cue.devices: "Get coolers actual RPMs failed. Device Commander CORE"

    • 2022-01-02 17:28:38.012 I cue.devices:  Open file with id bragiprotocol::FileId(CoolersActualRPMs) failed.
        File already opened.

    • 2022-01-02 17:28:38.016 W cue.dev.bragi.feature.channeled_sensors_controller: Failed to get channel id for sensor

    • 2022-01-02 17:28:38.017 W cue.dev.bragi.feature.channeled_sensors_controller: Failed to get channel id for sensor

    • 2022-01-02 17:28:38.017 W cue.dev.bragi.feature.channeled_sensors_controller: Failed to get channel id for sensor

    • 2022-01-02 17:28:38.017 W cue.dev.bragi.feature.channeled_sensors_controller: Failed to get channel id for sensor

    • 2022-01-02 17:28:38.944 I cue.devices: "Get coolers actual RPMs failed. Device Commander CORE"

    • 2022-01-02 17:28:38.944 I cue.devices:  Open file with id bragiprotocol::FileId(CoolersActualRPMs) failed.
        File already opened.

    • (...)

  • corsair_cue_20220102_16_12_03.csv (sensor logs)

    • Timestamp,(...),"H115i ELITE CAPELLIX Coolant Temp","H115i ELITE CAPELLIX Fan #6 (Bottom)","H115i ELITE CAPELLIX Fan #5 (Front Bottom)","H115i ELITE CAPELLIX Fan #4 (Front Top)","H115i ELITE CAPELLIX Fan #3 (Top Middle)","H115i ELITE CAPELLIX Fan #2 (Top Back)","H115i ELITE CAPELLIX Fan #1 (Back)","H115i ELITE CAPELLIX Pump"

    • (...)

    • 2/1/2022 17:25:08,(...),35.20°C,1555RPM,1621RPM,1154RPM,1107RPM,1621RPM,1560RPM,2292RPM

    • 2/1/2022 17:25:13,(...),35.20°C,1532RPM,1578RPM,1155RPM,1109RPM,1579RPM,1513RPM,2288RPM

    • 2/1/2022 17:25:18,(...),35.20°C,1479RPM,1521RPM,1155RPM,1111RPM,1532RPM,1463RPM,2284RPM

    • 2/1/2022 17:25:23,(...),35.20°C,1479RPM,1521RPM,1155RPM,1111RPM,1532RPM,1463RPM,2284RPM

    • 2/1/2022 17:25:28,(...),35.20°C,1479RPM,1521RPM,1155RPM,1111RPM,1532RPM,1463RPM,2284RPM

      • Until rebooting/restarting iCUE the logs continue with the exact same values: Sensor readings got stuck/flatlined

What fixes the issue (temporarily)

  • Reboot the system
  • Restart iCUE

 

What concerns me a lot is that there are lots of people/topics describing the same issue since around 2018 without any fixes eversince. Corsair went silent in most of the topics without providing any more support which makes me think if the issue might be a hardware flaw/general incompatibility. As of now I'm still believing the root cause is a software/driver issue. Here are a few examples, I could provide much more links from this forum, reddit etc.:

 

I am very frustrated with my products not working as supposed, not being able to identify the root cause and not being able to fix it with Corsairs help. This completely kills the fun I'm supposed to have with my PC.

  • All my efforts (reseach, support tickets, replacing hardware etc.) already took lots (days!) of my personal time
  • I am providing debug information and logs to Corsairs support for months now without being asked a single question or feedback from their end
    • Corsairs replies were always very generic and high level which gives me the feeling that my problem is not being taken seriously
  • I offered Corsair my help to identify the root cause and to provide more information in order to get the issue sorted together with their experts but I didn't get the impression that my information are being looked into by the right people

There have to be low level debugging possibilities to identify the root cause of the issue. And overheating systems due to faulty cooling are a serious issue for cooling solutions.

 

This is my final approach, I'm tired of being quiet about this.

Any further help is appreciated.

 

Thank you and kind regards

Tobias

 

211225_Flatlining_Coolant.png

211225_Flatlining_Fans.png

211225_Flatlining_Pump.png

Edited by t0niX
Edited typo
  • Like 2
Link to comment
Share on other sites

I must admit that I‘m very disappointed with not a single reply since Monday. The creation of the topic/post took a lot of time.

Short update from my end: I downgraded all chipset drivers to the Windows 11 default ones. I haven’t had the time to test extensively, but the issue didn’t appear since then.

Maybe a hint to all of you having similar problems.

Link to comment
Share on other sites

On 1/2/2022 at 6:31 PM, t0niX said:
  • I am providing debug information and logs to Corsairs support for months now without being asked a single question or feedback from their end
    • Corsairs replies were always very generic and high level which gives me the feeling that my problem is not being taken seriously
  • I offered Corsair my help to identify the root cause and to provide more information in order to get the issue sorted together with their experts but I didn't get the impression that my information are being looked into by the right people

While I am mentioned in the above post, it seemed clear this was intended for other people.  The period between the holidays and through CES is a tough time to get individual attention, so don't give up quite yet.

 

On 1/2/2022 at 6:31 PM, t0niX said:

As of now I'm still believing the root cause is a software/driver issue.

I agree.  The Corsair.Service is crashing.  That leaves the sensor data used in the control curves in a frozen state.  Unfortunately this is a bit like "I got a BSOD".  There are many possible reasons for the crash and you done a long list of possible fixes.  There is not a lot left for other users to suggest, but this should be a solvable problem.  Thousands of people with that hardware combination have been using CUE and this does not happen.  If the problem returns, there are a few things I would suggest.

1) Leave HWiNFO uninstalled for now or if you need it, use the Portable version and only launch it when CUE is not running and you manually kill the Corsair.service.  It is possible to use HWiNFO along side CUE if you suspend monitoring of the Corsair internal devices (Commander Core, RAM temp monitoring is usually OK but disable to be thorough).  When done you relaunch CUE.  Right now you don't need another variable in the mix.  Afterburner is fine.  A huge percentage of users run this along side CUE and I have never heard of a software collision between those two.  HWiNFO will lock up your fan controller if its polling the Commander Core and that is a known issue.

 

2) If the issue returns, try changing your lighting effects to non-temperature dependent lighting -- normal lighting presets.  While all your hardware and set-up is like everyone else, I believe the number of users who use extensive temp dependent lighting is much smaller.  I almost never do, except when testing a new piece of hardware.  I am not aware of any inherent defect with this, but clearly there must be an interaction between sensor data, software, and lighting control when using this.  It is the one thing that stands out as I read through your long list of attempted remedies.

 

3) You mentioned when the controller locks up, "the RGB still works".  Can you clarify if the LEDs are still changing with temperature as instructed (probably turning red as things get really hot) or it simply means the RGB LEDs are on and not dark.  Typically the lighting locks up as well when the service crashes, but sometimes different devices do different things.  For example, the fans colors might freeze, but the KB continues to run its active effect -- or vice versa.  I am mostly curious about this in relation to Temp dependent lighting effects and if the KB (which does not appear to be on temp dependent lighting) behaves differently.   

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

Posted (edited)

 

Thanks for your reply. Indeed I tagged you, but not by mistake. Your replies were very helpful and valuable to others - and as the support wasn't able to help me out since six (!!!) months, I thought I'd give it a shot.

So - a quick update first: The downgrade of all chipset drivers to the Windows 11 default ones didn’t help.

7 hours ago, c-attack said:

1) Leave HWiNFO uninstalled for now or if you need it, use the Portable version and only launch it when CUE is not running and you manually kill the Corsair.service.  It is possible to use HWiNFO along side CUE if you suspend monitoring of the Corsair internal devices (Commander Core, RAM temp monitoring is usually OK but disable to be thorough).  When done you relaunch CUE.  Right now you don't need another variable in the mix.  Afterburner is fine.  A huge percentage of users run this along side CUE and I have never heard of a software collision between those two.  HWiNFO will lock up your fan controller if its polling the Commander Core and that is a known issue.

Uninstalling HWiNFO didn't fix anything. Question: Does it even lock the controller if "Corsair and Asetek support" is turned OFF in settings?

Quote

2) If the issue returns, try changing your lighting effects to non-temperature dependent lighting -- normal lighting presets.  While all your hardware and set-up is like everyone else, I believe the number of users who use extensive temp dependent lighting is much smaller.  I almost never do, except when testing a new piece of hardware.  I am not aware of any inherent defect with this, but clearly there must be an interaction between sensor data, software, and lighting control when using this.  It is the one thing that stands out as I read through your long list of attempted remedies.

I'll try that, but I doubt it'll help. Will give some feedback on this ASAP.

Quote

3) You mentioned when the controller locks up, "the RGB still works".  Can you clarify if the LEDs are still changing with temperature as instructed (probably turning red as things get really hot) or it simply means the RGB LEDs are on and not dark.  Typically the lighting locks up as well when the service crashes, but sometimes different devices do different things.  For example, the fans colors might freeze, but the KB continues to run its active effect -- or vice versa.  I am mostly curious about this in relation to Temp dependent lighting effects and if the KB (which does not appear to be on temp dependent lighting) behaves differently.   

I can clarify the following things: All my devices are running temperature dependent lighting bug my keyboard and mouse. I'm still able to switch profiles, RGB is not completely locked up (even if the sensor readings are). Temperature dependent lighting on components with flatlined sensor data (all connected to Commander Core) are still changing color if not based on flatlined sensor data such as GPU temperature. The RPM of all fans and/or coolant temp of Commander Core get stuck though.

Plus: The RGB on my RTX 3080 often freeze (LEDs are on doing nothing/showing the RGB state displayed shortly before the freeze) as well. Even switching profiles won't change anything (which differs from the behavior of devices connected to the Commander Core when getting stuck) unless I reboot/restart iCUE (see this topic: https://forum.corsair.com/forums/topic/174396-nvidia-rtx-3080-founders-edition-rgb-control-gets-stuckfreezes-after-some-time

One more info: While playing GTA V and Hitman 3 everything seems to work fine (not playing those games much though). My "one and only" game I'm playing is Call of Duty: Warzone. The issue is almost 100 % appearing during longer sessions playing that one.

I'm so tired of this and I have no clue what to try our anymore.

 

 

Edited by t0niX
Link to comment
Share on other sites

HWiNFO - the disable Asetek/Corsair support was originally intended to limit polling of the AIO units with internal fan controllers. Now we have the Elite units that are really a pump powered through an external controller (Commander Core) and are not Asetek either. That makes it more like my Commander Pro that will lock up immediately. Best practice is to do both - turn on the safety setting and disable the individual device in the data panels. Also errors like one below make wonder about this. 
 

  • 2022-01-02 17:28:38.944 I cue.devices: "Get coolers actual RPMs failed. Device Commander CORE"

It seems like just the data fetching is failing and not the entire service. And also only select sensors lock up?  Not all of them?  This reads like two software programs locking horns, but I am not sure what that is. 
 

Unfortunately I don’t know enough about how CUE accesses the GPU LEDs on the FE cards. However, it physically must be accessed through the motherboard. Most other data must pass through it as well and I am wondering does all CUE data lock up when this happens?  Or select data?  You can add a lot of stuff to the CUE dashboard to get a quick visual about what is or is not updating. 

Link to comment
Share on other sites

Posted (edited)
6 hours ago, c-attack said:

HWiNFO - the disable Asetek/Corsair support was originally intended to limit polling of the AIO units with internal fan controllers. Now we have the Elite units that are really a pump powered through an external controller (Commander Core) and are not Asetek either. That makes it more like my Commander Pro that will lock up immediately. Best practice is to do both - turn on the safety setting and disable the individual device in the data panels. Also errors like one below make wonder about this. 
 

  • 2022-01-02 17:28:38.944 I cue.devices: "Get coolers actual RPMs failed. Device Commander CORE"

It seems like just the data fetching is failing and not the entire service. And also only select sensors lock up?  Not all of them?  This reads like two software programs locking horns, but I am not sure what that is. 

Alright - I get you, thanks for the heads up. If HWiNFO was uninstalled (service was removed as well) and the issue still appeared it should not be the root cause - correct? Same with MSI Afterburner.

The devices connected to Commander Core and/or the Commander Core itself lock up randomly. It's either only the coolant temperature, all the Fan speeds or both. I wonder if there's something with CoD Warzone (you can display FPS, packet loss, CPU/GPU temperature etc. in game) that might be the problem? But even if so neither could I change that, nor would I be the only one to notice.

Quote

Unfortunately I don’t know enough about how CUE accesses the GPU LEDs on the FE cards. However, it physically must be accessed through the motherboard. Most other data must pass through it as well and I am wondering does all CUE data lock up when this happens?  Or select data?  You can add a lot of stuff to the CUE dashboard to get a quick visual about what is or is not updating. 

I don't know either but I created the seperate topic regarding that issue to keep focus on the different issues. To your question: No, it's always ONLY the Commander Core that messes up. All other sensors (like GPU temperature) are working fine. It's only the GPU RGB that is freezing. Sometimes my DRAM and/or my keyboard show a triangle as well.

In all cases restarting iCUE or rebooting do the magic - but this cannot be the solution.

Edited by t0niX
Link to comment
Share on other sites

Just a quick update:

After removing the custom fan curve and removing temperature lighting, my Elite Capellix did not flatline again (thanks for the idea!). BUT: The RGB of my RTX 3080 still got stuck (addressed in another topic, just wanted to mention it in this context).

I haven’t had the time to test extensively, so my comment above regarding the Elite Capellix is not yet verified.

Any more ideas?

Edited by t0niX
Link to comment
Share on other sites

I don't know how CUE accesses the lighting on the Nvidia factory cards.  Is there something in the driver?  I suspect it is the DLL plugin from the CUE/plugins/Nvidia folder.  There have been some issues with fan control in CUE on the FE cards and it was stated this was the result of a driver change.  I don't know if the lighting is affected in the same way, but that should be an answerable question for Corsair.  Sadly, I don't have a FE card to work with.

 

The only other recent discussion I can remember involving the factory cards is below.  This is not something I've worked on so I can't verify any of it, but it suggests the 3080 Ti FE is not supported for RGB.  It seems strange there would be a difference between a 3080 FE and 3080 Ti FE for lighting support.  So is there currently CUE lighting support for any 3000 series FE cards?

 

https://forum.corsair.com/forums/topic/171431-3080ti-fe-not-showing-in-icue-414179/#comment-1003878

 

***One other note.  It is likely someone had you turn off the "enable plugins" switch in the main CUE app settings during troubleshooting. I would have suggested the same because of the data service errors.  You do need that on so CUE can access the presumed plugin for the 3080 in the Nvidia folder.  

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

On 1/15/2022 at 3:31 PM, c-attack said:

I don't know how CUE accesses the lighting on the Nvidia factory cards.  Is there something in the driver?  I suspect it is the DLL plugin from the CUE/plugins/Nvidia folder.  There have been some issues with fan control in CUE on the FE cards and it was stated this was the result of a driver change.  I don't know if the lighting is affected in the same way, but that should be an answerable question for Corsair.  Sadly, I don't have a FE card to work with.

 

The only other recent discussion I can remember involving the factory cards is below.  This is not something I've worked on so I can't verify any of it, but it suggests the 3080 Ti FE is not supported for RGB.  It seems strange there would be a difference between a 3080 FE and 3080 Ti FE for lighting support.  So is there currently CUE lighting support for any 3000 series FE cards?

 

https://forum.corsair.com/forums/topic/171431-3080ti-fe-not-showing-in-icue-414179/#comment-1003878

 

***One other note.  It is likely someone had you turn off the "enable plugins" switch in the main CUE app settings during troubleshooting. I would have suggested the same because of the data service errors.  You do need that on so CUE can access the presumed plugin for the 3080 in the Nvidia folder.  

A short answer since this is not the content of this topic: The RGB of my RTX 3080 FE works in iCUE (but gets stuck after some time so that it is not changing its effects until iCUE is being restarted or the system is being rebooted).

Back to topic: I've just reinstalled my machine using a clean install with Windows 11. I have not installed MSI Afterburner, RivaTuner Statistics Server and HWiNFO yet. Let's see.

Link to comment
Share on other sites

Hi there,

after a clean install of Windows 11 and iCUE (no data imported at all) the issue instantly appeared after starting Call of Duty: Warzone tonight.

  • Info: No temperature lighting and no custom fan curves set!

This time I was able to catch the missing piece in the debug logs (which was missing in all my debug logs before).

  • Info: I did a forced firmware update to the Elite Capellix before, because I already had problems with it yesterday.

This is what the log file shows: "iCUE > CUELogs > 2022-01-22T21-30-14.log".

  • 2022-01-22 21:44:48.067 C cue.datacollection.user_profile_manager: Cannot get CorsairDevice for sending data collection event
  • 2022-01-22 21:44:48.067 C cue.datacollection.user_profile_manager: Cannot get CorsairDevice for sending data collection event
  • 2022-01-22 21:58:17.990 W qrc:/qml-imports/com/corsair/cue/deviceview/DynamicGeometryView.qml:185: TypeError: Cannot read property 'extensionsScalingRatio' of null
  • 2022-01-22 21:58:18.010 W qrc:/qml-imports/com/corsair/cue/deviceview/DynamicGeometryView.qml:185: TypeError: Cannot read property 'extensionsScalingRatio' of null
  • 2022-01-22 21:58:25.213 W cue.sensors.system_info: Update values already requested for: SensorDataCache::Resolution::OneSecond
  • 2022-01-22 21:58:25.224 I cue.devices.dram: Dram sensor updating skipped: "vid<1b1c>pid<000004>serial<DOMINATORPLATINUMRGB>index<0>"
  • 2022-01-22 21:58:26.054 C cue.devices: Unbind buffer failed. Device: "Commander CORE"
  • 2022-01-22 21:58:26.054 C cue.devices: Exception:  Timeout (100ms) while sending command.
  • 2022-01-22 21:58:26.072 I cue.devices: "Get coolers actual RPMs failed. Device Commander CORE"
  • 2022-01-22 21:58:26.092 I cue.devices:  Describe buffer failed.
    • Timeout (100ms) while sending command.
  • 2022-01-22 21:58:26.830 I cue.devices: "Get temperature measurements status failed. Device Commander CORE"
  • 2022-01-22 21:58:26.830 I cue.devices:  Describe buffer failed.
    • Timeout (100ms) while sending command.
  • 2022-01-22 21:58:26.830 I cue.devices: "Get coolers actual RPMs failed. Device Commander CORE"
  • 2022-01-22 21:58:26.830 I cue.devices:  Open file with id bragiprotocol::FileId(CoolersActualRPMs) failed.
    • File already opened.
  • 2022-01-22 21:58:26.835 W cue.sensors.system_info: Update values already requested for: SensorDataCache::Resolution::OneSecond
  • 2022-01-22 21:58:27.212 I cue.devices: "Get coolers actual RPMs failed. Device Commander CORE"
  • 2022-01-22 21:58:27.212 I cue.devices:  Open file with id bragiprotocol::FileId(CoolersActualRPMs) failed.
    • File already opened.
  • 2022-01-22 21:58:28.298 I cue.devices: "Get coolers actual RPMs failed. Device Commander CORE"
  • 2022-01-22 21:58:28.298 I cue.devices:  Open file with id bragiprotocol::FileId(CoolersActualRPMs) failed.
    • File already opened.
  • [...]

The sensor log(s) show nothing but the "H115i ELITE CAPELLIX Coolant Temp" even though all available sensors of Elite Capellix were checked in iCUE settings for logging.

After closing iCUE the fans connected to my Elite Capellix go wild until the coolant temp reaches a lower temperature ("hardware mode"). Restarting iCUE will fix the issue instantly until it appears again randomly.

I'm out of ideas and am just annoyed by the product(s).

Tobias

Edited by t0niX
Link to comment
Share on other sites

On 1/22/2022 at 4:38 PM, t0niX said:

 

Hello t0niX, 

I stumbled across this post after looking up some information regarding Corsairs issues surrounding their coolers and the flatlining issue as I am soon to upgrade my PC, so I was seeing if Corsair has resolved this joke.

Unfortunately, this has been an issue since 2018... and has persisted since for many users.

Please see this forum post: 

 

I think I get involved at around page 7 of 20... 

The thread was eventually locked as after 2-3 years of the issue being prevalent they did not fix the issue. 

@Corsair Nick Did lock the thread and you can read the final comment here:

As far as I am aware, the issue has not yet been resolved. I did try posting to Reddit and Corsair Support tickets to which I am met with silence or repeating fixes that have already been attempted with other Corsair staff. The issue is frustrating and at this point, laughable.  

My advice would be too other RMA/return your cooler for anything else, or use Corsair LINK to ensure stability of your system as the cooler will not adjust the fan speeds based on the temps. I ended up sitting at 80c, not realizing my cooler wasn't working until thermal throttling kicked in and my FPS tanked.  

Good luck buddy. 

 

Link to comment
Share on other sites

9 hours ago, BrewMyTeabag said:

Hello t0niX, 

I stumbled across this post after looking up some information regarding Corsairs issues surrounding their coolers and the flatlining issue as I am soon to upgrade my PC, so I was seeing if Corsair has resolved this joke.

Unfortunately, this has been an issue since 2018... and has persisted since for many users.

Please see this forum post: 

 

I think I get involved at around page 7 of 20... 

The thread was eventually locked as after 2-3 years of the issue being prevalent they did not fix the issue. 

@Corsair Nick Did lock the thread and you can read the final comment here:

As far as I am aware, the issue has not yet been resolved. I did try posting to Reddit and Corsair Support tickets to which I am met with silence or repeating fixes that have already been attempted with other Corsair staff. The issue is frustrating and at this point, laughable.  

My advice would be too other RMA/return your cooler for anything else, or use Corsair LINK to ensure stability of your system as the cooler will not adjust the fan speeds based on the temps. I ended up sitting at 80c, not realizing my cooler wasn't working until thermal throttling kicked in and my FPS tanked.  

Good luck buddy. 

 

Hey there - thanks a lot for your reply. Sadly I'm experiencing the same thing.

Corsairs Support already RMAd the Commander Core itself and after that as well the whole Elite Capellix. The issue still persists, so it's either my machine (what I doubt because so many people experiencing the same behavior) or a hardware/software/driver issue which I'm trying to resolve together with Corsairs support.

Sadly I'm getting the same standardized replies with suggested "fixes" (e.g. clean installation of iCUE, forced firmware update of Commander Core etc.). There have to be ways for detailed debugging or direct contact with Corsairs developers (e.g. in a closed beta program). This is the loop I'm in for more than half a year (yep - HALF A YEAR) now.

You're correct with your observation as well: In all threads Corsair went silent without any fix provided. As a customer I feel completly let alone in the dark without any official information if this issue is on the roadmap to be resolved or not. I offered my help to resolve it with whatever data I can supply, but this offer wasn't claimed by Corsairs support (yet).

It's honestly a mess and a no-go for a premium company selling premium products for premium prices.

Edited by t0niX
  • Like 2
Link to comment
Share on other sites

7 hours ago, t0niX said:

Hey there - thanks a lot for your reply. Sadly I'm experiencing the same thing.

Corsairs Support already RMAd the Commander Core itself and after that as well the whole Elite Capellix. The issue still persists, so it's either my machine (what I doubt because so many people experiencing the same behavior) or a hardware/software/driver issue which I'm trying to resolve together with Corsairs support.

Sadly I'm getting the same standardized replies with suggested "fixes" (e.g. clean installation of iCUE, forced firmware update of Commander Core etc.). There have to be ways for detailed debugging or direct contact with Corsairs developers (e.g. in a closed beta program). This is the loop I'm in for more than half a year (yep - HALF A YEAR) now.

You're correct with your observation as well: In all threads Corsair went silent without any fix provided. As a customer I feel completly let alone in the dark without any official information if this issue is on the roadmap to be resolved or not. I offered my help to resolve it with whatever data I can supply, but this offer wasn't claimed by Corsairs support (yet).

It's honestly a mess and a no-go for a premium company selling premium products for premium prices.

Unfortunate. There really is nothing you can do. They had access to a multitude of logs in the original post and yet, nothing was used or fixed. 

It's very clear that it is the Corsair Service being the issue, as restarting that also solved my issue and the flatlines became normal again; for myself as well as you and many other users. But this "fix" was only temporary and the issue would reoccur at random.  

My advice is to refund everything you can and stop supporting the company. How this isn't bigger news is beyond me, especially due to the length of time this issue has been occurring for. 

  • Like 2
Link to comment
Share on other sites

33 minutes ago, BrewMyTeabag said:

How this isn't bigger news is beyond me, especially due to the length of time this issue has been occurring for. 

I'm a 100 % with you. Especially with this affecting the main cooling component(s). This is a serious issue as systems run hot without the user noticing.

  • Like 1
Link to comment
Share on other sites

Ive also had a similar experience as you with corsair support as well,they will give you a generic reply and then just go silent,in my case like you the issue remains unresolved sadly enough.At the end of the day we pay a premium price for their products support should come as standard.

I ended up redoing my system with parts and their software and i can tell you the support from them is outstanding,if you can afford to go down that route i recommend it the software is away in another league comparing to icue.

 

Edited by Technobeard
no promotion of competitors
  • Like 1
Link to comment
Share on other sites

On 1/25/2022 at 7:34 PM, GMdoghunter said:

Ive also had a similar experience as you with corsair support as well,they will give you a generic reply and then just go silent,in my case like you the issue remains unresolved sadly enough.At the end of the day we pay a premium price for their products support should come as standard.

I ended up redoing my system with parts and their software and i can tell you the support from them is outstanding,if you can afford to go down that route i recommend it the software is away in another league comparing to icue.

 

I'll second this. I have an open thread from October of 2021 regarding iCUE software causing substantially higher CPU temps (10-15C under water) on Ryzen platforms, as well as links to several other threads with similar information. Unfortunately there has been no reply from anyone at Corsair and at this point I doubt there ever will be.

You can see that I've been bumping every 2-4 weeks to no avail. I simply don't have the patience to cry for help for nearly 4 months for an ecosystem I had paid well over $500 for. Ultimately I chose to swap out and refund as many of my Corsair pieces as I could - 32GB of RAM, H115i RGB Platinum, Commander Pro. I'm still stuck with a keyboard, but I don't mind not having RGB control over it. I replaced my Commander Pro (which turned out to be cheaper as a fan controller, didn't need RGB control) and I've been very impressed with the performance of their program and the service they've provided.

Best of luck to you @t0niX - 6 months of trying to deal with an ecosystem that could substantially damage your valuable PC parts would scare anyone off. You've certainly given them the benefit of the doubt. I don't think they would do the same for you if your computer was damaged by their software.

Edited by Technobeard
no promotion of competitors
  • Like 1
Link to comment
Share on other sites

Honestly how they get away with this i do not know as corsair products are not cheap you pay a premium price for there products,customer service should be the same.

At the end of the day your cooling system in your pc has to be reliable it protects so many expensive parts and delicate parts as well.

Best of luck @t0niX hopefully you get everything sorted out.

I do know what you are going through at the moment,my main reason for going with other products.

Edited by Technobeard
no promotion of competitors
  • Like 1
Link to comment
Share on other sites

Quick update: Corsair support asked me for more system information (created with a specific tool) beside the provided debug logs. I gladly sent it over. The issue instantly appeared after playing Call of Duty: Warzone tonight. Newest version of iCUE (4.20.169) installed.

Edited by t0niX
  • Like 2
Link to comment
Share on other sites

  • 2 weeks later...

Hi @t0niX

I have the exact same problem as yours, also only happened when playing MW2019/Warzone, I've tried everything I can, Including clean install everything( iCue for like million times, five or six times clean reinstall Windows, both Win11 and 10), also replugged commander core, usb header on mobo, I've reached for Corsair support via Skype, cuz I'm not in the US, the troubleshooting step they gave me are just clean reinstall iCue and some useless BS that I've tried it all and didn't fix anything, I've been dealing with the same issue for 2 months since the first time I reached for their support, nothing changed  after trying everything support guy told me to try for 2 months, I called their support again couple of days ago, then I saw your article here, and realized that only MW/Warzone that triggered iCue/Corsair service.exe fail, and I replied the E-mail to the support they asked me to reply and choose RMA or not, and here's what I replied:

"Hi,

I found out that the only game or app is causing the problem is MW2019/Warzone, so I think the Icue/Corsair service.exe is conflicted with whether the game itself or some kind of anti-cheat app that goes with it, so I don’t really think RMA the product will solve the problem since I’m still gonna use Icue with the new one anyway, and plus the dude that has exact same problem that Icue wasn’t detecting coolant temp correctly, also happened when playing warzone, so I don’t really think it’ll be the solution, cuz I’m in Taiwan, and I’ll have to buy a cooler to use my computer since it’ll took long time to get to you, and check, and send back to me, just fix the Icue software /Corsair service.exe, or whatever that causes the problem, cuz if the coolant temp gets too high, I don’t know if it’s going to leak o or something, I don’t want to buy your products like everything in my computer are from Corsair, and  monitor my computer 24/7 to see if it’s not doing good, also don’t want to spend time on RMAing cuz it wastes such a long time, and according to the guy has the same problem as I have, RMA seems to be a waste of time, I probably won’t consider your product next time, since your support are ridiculously useless, just fix the software‘s problem."

and there's no response for 2 days, I don't think they're gonna fix this defected software.

 

 

  • Like 1
Link to comment
Share on other sites

Hey there - sad to hear you're in the same boat. What are your hardware specifications (especially motherboard)?

I'm currently waiting for a reply. In between I've clean installed Windows 11 another time, didn't install HWiNFO at all. MSI Afterburner is not running, so there's no monitoring at all. Yesterday the Coolant Temperature flatlined and made my fans spin like crazy (because of faulty high temperatures).

It's a shame - it took me SO MANY HOURS to type LOTS of e-mails, waiting for replies, assembling, disassembling, assembling the RMA unit and much more steps in between. This goes on for about 8 months now. I almost lost hope.

But the indicator might be a good one: Call of Duty: Warzone might be a/the cause.

Link to comment
Share on other sites

12 hours ago, t0niX said:

Hey there - sad to hear you're in the same boat. What are your hardware specifications (especially motherboard)?

I'm currently waiting for a reply. In between I've clean installed Windows 11 another time, didn't install HWiNFO at all. MSI Afterburner is not running, so there's no monitoring at all. Yesterday the Coolant Temperature flatlined and made my fans spin like crazy (because of faulty high temperatures).

It's a shame - it took me SO MANY HOURS to type LOTS of e-mails, waiting for replies, assembling, disassembling, assembling the RMA unit and much more steps in between. This goes on for about 8 months now. I almost lost hope.

But the indicator might be a good one: Call of Duty: Warzone might be a/the cause.

hi,

Here's my specs

cpu: 9600kf 4.8G

mobo: aorus z390 elite

ram: vengeance rgb pro sl 3200 8g*2

gpu: aorus rtx2080

psu: leadex 750w 80+ gold

os: win10(currently)

commander core that came with the aio

cooler: h100i elite capellix

fans: 4* ll120 rgb

case: 4000d af

I'm listing my specs and thinking, how the hell that I paying all fine quality corsair's product, and end up with a defected software that I'll have to deal with.

and yesterday a new problem came up, icue suddenly not detecting one of the ram stick, cuz one of the rgb are not acting right, the ram also didn't show up in icue control panel and the main page, but in task manager, aida64, device manager, all shows 2 sticks are well installed, so I reinstall the ram and reset my system a coup o times, it shows up in icue agian, I'm just so done with icue. 

and I still not getting their reply yet.

not a native speaker, plus hang over, might be a little bit hard to understand lol

let's just hope corsair stop ignoring this and fix icue for god sake

 

Link to comment
Share on other sites

  • 4 weeks later...

Alright guys - after months, days, hours, and hours, and hours of trial and error, LOTS of testing and even more frustration I've finally found THE solution to my problems with iCUE and can enjoy my PC again ever since.

I'm confident that adjusted CPU priorities are the missing part to the puzzle. It seems that every time a process is taking up very high CPU load (in my case CoD: Warzone) iCUE/some of its services ran into timeouts and have triggered flatlining of sensor data/freezing of RGB effects. Since I've adjusted the priorities (see attached screenshot) I haven't had a single issue.

It is extremely unfortunate that in more than 10 months since I created my first ticket and after over 100 emails no one from Corsairs support has been able to help me out or resolve the issue. I would welcome an official statement on the drama I (and others) have experienced.

I'm happy to share this with you and I'm even happier about everybody I'm able to help out with the above information. Please share your feedback!

Enjoy!! 🙂

220307_iCUE_Solution.png

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...