Why is my predator helios 300 suddenly overheating?

24

Answers

  • sri369
    sri369 ACE Posts: 2,763 Pathfinder
    Sharfan said:
    sri369 said:
    Sharfan said:
    I gave it a shot and reinstalled Windows and re downloaded the drivers from the acer site.... still the same issue as before... really don't know what to do apart from trying to re-paste 
    *** shaking my head ***

    Install windows off MS site and let Windows update drivers. Do NOT use Acer drivers.
    Was just following @Red-Sand's instructions from his comment
    That's a option that's worked for some. Didn't work for you. Now try this way :-) Exhaust "soft" options before going to "hard" options that could void warranty.
    -----
    Karma...
    LIKE - if helpful
    ACCEPT - if helped resolve
  • Sharfan
    Sharfan Member Posts: 45 Troubleshooter
    sri369 said:
    Sharfan said:
    sri369 said:
    Sharfan said:
    I gave it a shot and reinstalled Windows and re downloaded the drivers from the acer site.... still the same issue as before... really don't know what to do apart from trying to re-paste 
    *** shaking my head ***

    Install windows off MS site and let Windows update drivers. Do NOT use Acer drivers.
    Was just following @Red-Sand's instructions from his comment
    That's a option that's worked for some. Didn't work for you. Now try this way :-) Exhaust "soft" options before going to "hard" options that could void warranty.
    Tried to reinstall from the USB again and this time I went to the Windows settings and ran the update instead of manually downloading drivers from Acer. However there were no diver updates downloading, just security updates.
  • Sharfan
    Sharfan Member Posts: 45 Troubleshooter
    Sharfan said:
    sri369 said:
    Sharfan said:
    sri369 said:
    Sharfan said:
    I gave it a shot and reinstalled Windows and re downloaded the drivers from the acer site.... still the same issue as before... really don't know what to do apart from trying to re-paste 
    *** shaking my head ***

    Install windows off MS site and let Windows update drivers. Do NOT use Acer drivers.
    Was just following @Red-Sand's instructions from his comment
    That's a option that's worked for some. Didn't work for you. Now try this way :-) Exhaust "soft" options before going to "hard" options that could void warranty.
    Tried to reinstall from the USB again and this time I went to the Windows settings and ran the update instead of manually downloading drivers from Acer. However there were no diver updates downloading, just security updates.
    And the temps were still really high :(
  • Sharfan
    Sharfan Member Posts: 45 Troubleshooter
    I tried elevating the laptop while gaming to improve the air intake from the bottom of the laptop... This managed to help by lowering the temps by about 4 Degrees Celcius... I'm beginning to think that it may be dust inside thats causing the overheating... although both @Vishwas and I reporting the same issue around the same time does seem too coincidental for an issue like dust build up...
  • Queen6
    Queen6 Member Posts: 319 Skilled Practitioner WiFi Icon
    edited June 2018
    Sharfan said:
    I tried elevating the laptop while gaming to improve the air intake from the bottom of the laptop... This managed to help by lowering the temps by about 4 Degrees Celcius... I'm beginning to think that it may be dust inside thats causing the overheating... although both @Vishwas and I reporting the same issue around the same time does seem too coincidental for an issue like dust build up...

    Right now your changing everything and not achieving the goal, you need to identify and isolate the issue.  First let's verify that the windows install is good:
    You can check for corrupt system files by opening the Command  Prompt (under Administrator) type sfc /scannow  This will search and fix some errors if uncorrectable the system will advise.

    The Deployment Image Servicing and Management (DISM) tool can verify & repair deeper issues, however you will need a decent connection and the Windows Update client must be functional.  DISM will make changes / update the system image where it finds difference versus the Microsoft stock Windows image.  

    1 - Quick check - Command  Prompt (under Administrator) type: DISM /Online /Cleanup-image /CheckHealth
    2 - Deeper Scan - Command  Prompt (under Administrator) type: DISM /Online /Cleanup-image /ScanHealth
    3 - Will Repair & Modify - Command Prompt (under Administrator) type: DISM /Online /Cleanup-image /RestoreHealth

    Again the 3rd DISM command will change system files so ensure you backup your notebook first, or be prepared to reset Windows.  Some commands will take time to execute so be patient.  If all good leave Windows alone :)

    Questions?
    What revision of Windows, should be 1803
    What revision of Nvidia Driver?
    If using ThrottleStop screen cap Main, FIVR & TPL windows
    Use HWinfo64 to identify if CPU, GPU or both are running hot, use the logging feature and chart results with Excel etc. this will let you see avg temp and spikes.  Avg being the point of interest during the game period.
    What's the CPU core temperature differential under load?
    How old is the notebook

    Q-6

  • Sharfan
    Sharfan Member Posts: 45 Troubleshooter
    Both sfc /scannow and deeper scan using DISM /Online /Cleanup-image /ScanHealth did not bring up any issues...
    I am running version 1803 of Windows
    Version 398.11 for my Nvidia Driver
    I purchased this notebook Mid January



  • Queen6
    Queen6 Member Posts: 319 Skilled Practitioner WiFi Icon
    Sharfan said:
    Both sfc /scannow and deeper scan using DISM /Online /Cleanup-image /ScanHealth did not bring up any issues...
    I am running version 1803 of Windows
    Version 398.11 for my Nvidia Driver
    I purchased this notebook Mid January

    CPU temp is high, 4 degree centigrade CPU core differential is ok, rest nothing wild here, uncheck "Clamp" on ThrottleStop TPL, this will stop the CPU rolling back frequency below the nominal 2.8GHz.  Can you go deeper with the undervolt offset? or does the system incur stability issues.  8.60 TS Bench is very good at revealing issue.  Hit the CPU with the 1024M test repeatedly (x6, x9, x12), no error increase the CPU Core & CPU Cache offset (must be the same).  I look for zero error, then decease the CPU core & Cache offset value by 5% - 10% 

    Need to see the thermals, including aprox ambient, and current draw on the CPU & dGPU.  Step at a time as changing everything presents far too many variables to understand where the problem lies.  At just six months, dust is unlikely to be an issue unless prevalent in the environment.  Look at the limits in ThrottleStop, what's tripping? If flagged yellow after gaming will help to narrow down, if Red means the limit is currently being tripped; thermals, voltage, current EDP etc.

    All silicon is unique, some chips run hotter, some have a higher power demand, some will undervolt better than others, add in software and you can see how complex it can be.  Often described as the "Silicon Lottery" aside my Predator 17 I also have another 17.3" notebook with intel's new 8750H that is pulling 1271CB under Cinebench R15, holding a sustained 3.4GHz - 3.6GHz, boosting to a clean 3.9GHz max Turbo on all cores, insanely fast and stable up to -150mV undervolt, although due to the OEM's power profiles (silent mode - lower power to CPU) it's more prudent to run at a global undevolt of -140mV.  This CPU hardly passes 70C with well over 3GHz on all cores (load dependant 3.2 - 3.6) for literally hours & hours on end.  My notebook's are a function of revenue, therefore they must be optimised for stability and performance.

    We need to understand the limitations of your notebooks CPU, either increasing the undervolt and or limiting Turbo to achieve the best balance and performance.

    FWIW: on the Predator 17 now, Asus S7BS8750 is running Prime95 @ 3.38GHz, CPU temp peaked at 94C, as the cooling system spooled up, 72C, -140.6mV undervolt.  For your Helios 300 we want get the temps down to 85C max and the FPS up :)

    Q-6
  • Red-Sand
    Red-Sand ACE Posts: 1,892 Pathfinder
    edited June 2018
    I'd like to add that Throttlestop can only assist if Windows is optimized properly. 
    This includes removing background services and task schedules that come with installing software such as chrome. 
    Additionally removing windows background processes will also help reduce temps greatly. 

    This video doesn't show *everything* you need to disable but it shows where to look for those rouge programs. 
    You'll want to search for programs you've installed such as Google chrome (for its updater) in services & task schedules, apple/ipod, antivirus, any auto updater, any telemetry services etc. 

    https://youtu.be/XTPMwQKlfPE
    - Hotel Hero
  • Sharfan
    Sharfan Member Posts: 45 Troubleshooter
    Queen6 said:
    Sharfan said:
    Both sfc /scannow and deeper scan using DISM /Online /Cleanup-image /ScanHealth did not bring up any issues...
    I am running version 1803 of Windows
    Version 398.11 for my Nvidia Driver
    I purchased this notebook Mid January

    CPU temp is high, 4 degree centigrade CPU core differential is ok, rest nothing wild here, uncheck "Clamp" on ThrottleStop TPL, this will stop the CPU rolling back frequency below the nominal 2.8GHz.  Can you go deeper with the undervolt offset? or does the system incur stability issues.  8.60 TS Bench is very good at revealing issue.  Hit the CPU with the 1024M test repeatedly (x6, x9, x12), no error increase the CPU Core & CPU Cache offset (must be the same).  I look for zero error, then decease the CPU core & Cache offset value by 5% - 10% 

    Need to see the thermals, including aprox ambient, and current draw on the CPU & dGPU.  Step at a time as changing everything presents far too many variables to understand where the problem lies.  At just six months, dust is unlikely to be an issue unless prevalent in the environment.  Look at the limits in ThrottleStop, what's tripping? If flagged yellow after gaming will help to narrow down, if Red means the limit is currently being tripped; thermals, voltage, current EDP etc.

    All silicon is unique, some chips run hotter, some have a higher power demand, some will undervolt better than others, add in software and you can see how complex it can be.  Often described as the "Silicon Lottery" aside my Predator 17 I also have another 17.3" notebook with intel's new 8750H that is pulling 1271CB under Cinebench R15, holding a sustained 3.4GHz - 3.6GHz, boosting to a clean 3.9GHz max Turbo on all cores, insanely fast and stable up to -150mV undervolt, although due to the OEM's power profiles (silent mode - lower power to CPU) it's more prudent to run at a global undevolt of -140mV.  This CPU hardly passes 70C with well over 3GHz on all cores (load dependant 3.2 - 3.6) for literally hours & hours on end.  My notebook's are a function of revenue, therefore they must be optimised for stability and performance.

    We need to understand the limitations of your notebooks CPU, either increasing the undervolt and or limiting Turbo to achieve the best balance and performance.

    FWIW: on the Predator 17 now, Asus S7BS8750 is running Prime95 @ 3.38GHz, CPU temp peaked at 94C, as the cooling system spooled up, 72C, -140.6mV undervolt.  For your Helios 300 we want get the temps down to 85C max and the FPS up :)

    Q-6
    I undervolted to -125mV and as you can see from the image it barely made a difference.
    I don't think I was previously able to even undervolt this far without BSOD.
    I was playing Withcer 3 to test temps as that was one of the first games where I noticed these high temps... The temps usually stick around high 80s and as you can see maximum spike up to 94 degrees which never used to happen before all these thermal issues... The regions where there are dips in the temps are due to playing Gwent within Witcher 3 which sits in the 60s as it is barely demanding...
    After gaming, when checking the limits BDPROCHOT was in yellow underneath CPU and GPU. There was also EDP in yellow underneath the RING column and PL1 in yellow underneath CPU.

  • Sharfan
    Sharfan Member Posts: 45 Troubleshooter
    Red-Sand said:
    I'd like to add that Throttlestop can only assist if Windows is optimized properly. 
    This includes removing background services and task schedules that come with installing software such as chrome. 
    Additionally removing windows background processes will also help reduce temps greatly. 

    This video doesn't show *everything* you need to disable but it shows where to look for those rouge programs. 
    You'll want to search for programs you've installed such as Google chrome (for its updater) in services & task schedules, apple/ipod, antivirus, any auto updater, any telemetry services etc. 

    https://youtu.be/XTPMwQKlfPE
    Thank you for the optimization guide :) ... but my Windows is already fairly optimized to my needs and I never used the optimization guide before the heating issues and my ThrottleStop undervolt of -113mV worked perfectly well... I just really wish I could identify whats causing these high temps because I don't remember making any changes to Windows apart from updates... And I noticed these temps before the 1803 update so I don't think its that...
  • gethy
    gethy Member Posts: 3 New User

    Sorry if cross-posting is a big faux pas. This thread looks relevant to mine, just I with a Nitro 5. The CPU temps are just going wild and really shouldn't.
  • Queen6
    Queen6 Member Posts: 319 Skilled Practitioner WiFi Icon
    edited June 2018
    Sharfan said:
    I undervolted to -125mV and as you can see from the image it barely made a difference I don't think I was previously able to even undervolt this far without BSOD.
    I was playing Withcer 3 to test temps as that was one of the first games where I noticed these high temps... The temps usually stick around high 80s and as you can see maximum spike up to 94 degrees which never used to happen before all these thermal issues... The regions where there are dips in the temps are due to playing Gwent within Witcher 3 which sits in the 60s as it is barely demanding...
    After gaming, when checking the limits BDPROCHOT was in yellow underneath CPU and GPU. There was also EDP in yellow underneath the RING column and PL1 in yellow underneath CPU.


    The CPU is pulling a lot of power, a lot for just the game, I just played Witcher 3 (Ultra) for a 15-20 minutes on the Predator 17 (G9-793) with 7700HQ & GTX 1070.  CPU is undervolted a little deeper at -131.8mV, however the CPU package power didn't exceed 29W, your CPU is pulling over 45W as it's tripping both PL-1 & EDP and or other components are running hot as BDPROCHOT is also tripping.

    Check all the Nvidia Geforce Experience settings and ensure everything is off, same with the Windows Gaming settings as it looks like there's an additional load on the CPU while gaming.  The Witcher 3 is a demanding game and the temperature on the 7700HQ topped out at 90C (auto fans), equally one would expect the power demand to be relative, with your Helios 300 pulling a little more power due the lesser undervolt.  Here we are potentially looking at over a 16W delta so something is amiss.  Try playing the game in windowed mode and see what else is firing up with the game in Task Manager and check the max power consumption with ThrottleStop and or HWinfo64, if not maxing out at >45W then it's possibly the VRM's are running hot.

    For the GPU, try installing 390.77 it may run a little cooler, bigger issue is the CPU as it's clearly pulling a lot of power and throttling.  Even when your playing Gwent the CPU is utilization is still high, although the overall game temp looks ok, just the occasional spike.

    If everything checks out ok with Nvidia & Windows Gaming and no other app is firing up the game, we will need to look at limiting the Turbo with ThrottleStop's FIVR options to get the power budget & temp down.  FWIW my Predator 17 even running Prime95 Small FFT stress test doesn't trip PL-1 pulling just short of 44W.

    Temp although expectedly high is not the point, more the power consumption of the CPU under Prime96 Small FFT, this is way more load on the CPU than game's will continuously apply.  Just note we don't have identical notebooks so this may just be how the Helios 300 is set up.

    Q-6
  • Queen6
    Queen6 Member Posts: 319 Skilled Practitioner WiFi Icon
    gethy said:

    Sorry if cross-posting is a big faux pas. This thread looks relevant to mine, just I with a Nitro 5. The CPU temps are just going wild and really shouldn't.

    Welcome to the party :)

    Q-6
  • sri369
    sri369 ACE Posts: 2,763 Pathfinder
    Installing Witcher 3... will post results in about an hour...
    -----
    Karma...
    LIKE - if helpful
    ACCEPT - if helped resolve
  • sri369
    sri369 ACE Posts: 2,763 Pathfinder
    edited June 2018
    Installed Witcher 3. HD textures mod. Ultra. Minus things like bloom n such. Refresh rate locked at 60 with vsync on.
    CPU undervolted to -0.125
    i7-7700, 1060 with 6 GB. Game installed on a partition on SSD.

    Game was with CPU temps in 80s. Low load temps on my laptops are between 40-60.

    While playing the more recent Ghost Recon Wildlands at full ultra, the temps ranged around 60-75, occasional spike to 85. So well... this game is real hard on processing power.


    -----
    Karma...
    LIKE - if helpful
    ACCEPT - if helped resolve
  • Red-Sand
    Red-Sand ACE Posts: 1,892 Pathfinder
    edited June 2018
    sri369 said:
    Installed Witcher 3. HD textures mod. Ultra. Minus things like bloom n such. Refresh rate locked at 60 with vsync on.
    CPU undervolted to -0.125
    i7-7700, 1060 with 6 GB. Game installed on a partition on SSD.

    Game was with CPU temps in 80s. Low load temps on my laptops are between 40-60.

    While playing the more recent Ghost Recon Wildlands at full ultra, the temps ranged around 60-75, occasional spike to 85. So well... this game is real hard on processing power

    If you follow my overclock/undervolt guide it drops temps further for both CPU & GPU.

    For witcher 3 I run max 68/65 C for CPU/GPU at all ultra (hairworks on geralt only) avg 60 fps with dips to 55 & capped fps at 65. (Set from nvidia inspector tool)
    This is a combination with my throttlestop setup.

    https://youtu.be/nYW7ROFx2Kk
    - Hotel Hero
  • Sharfan
    Sharfan Member Posts: 45 Troubleshooter
    Funny thing is I can see a 10 degree decrease in CPU temps when i apply my undervolt and run a CPU stress test... As soon as I stress both the CPU and GPU at the same time using AIDA64... the temps skyrocket into the 90s... Yes I know there will be additional heat from the GPU and i know that stress tests aren't representative of gaming session because the CPU and GPU won't always be running at 100% during a game... But its almost as if the undervolt makes no difference at all... I'm gonna try resetting windows in about a weeks time to see if it makes any difference...
  • Sharfan
    Sharfan Member Posts: 45 Troubleshooter
    Also I removed the undervolt and noticed that when I'm plugged in and I boot up my laptop and check the limits:
    - PL1 is yellow under the CORE column
    - EDP OTHER is yellow under the RING column
    Can someone explain what these mean because I'm new to ThrottleStop in general...
  • Sharfan
    Sharfan Member Posts: 45 Troubleshooter
    Just a quick update... I lowered my undervolt to -110mV and for some reason the temps have decreased while playing Witcher 3...
    Although the Limits Tab shows BDPROCHOT for both the CPU and GPU


  • Queen6
    Queen6 Member Posts: 319 Skilled Practitioner WiFi Icon
    Sharfan said:
    Funny thing is I can see a 10 degree decrease in CPU temps when i apply my undervolt and run a CPU stress test... As soon as I stress both the CPU and GPU at the same time using AIDA64... the temps skyrocket into the 90s... Yes I know there will be additional heat from the GPU and i know that stress tests aren't representative of gaming session because the CPU and GPU won't always be running at 100% during a game... But its almost as if the undervolt makes no difference at all... I'm gonna try resetting windows in about a weeks time to see if it makes any difference...

    Predator Helios 300 is a budget conscious gaming notebook and this is one of the trade off's.  Running both the CPU & GPU at 100% load is not representative of the expected use.  Both CPU & dGPU share the same cooling solution both under full load will easily result in excessive temperatures.  If you follow Red Sand's YouTube guides you wont be far off, temperature will vary due to ambient temp & manufacturing variances.

    It's all about attaining the right balance, if there is no SW issue and the temperature while gaming remains excessive, you need to dial down the performance a little.  It wont effect gameplay even a 5% reduction in CPU performance will make a difference, vast majority of games currently don't need a quad core CPU to be boosting to full Turbo of 3.4GHz which is a massive power draw generating a ton of heat in the process.

    If the system is running at around 80C during gaming with am ambient temp of around 25C it's pretty much ok, less a bonus, realistically unlikely without going to extremes like applying Liquid Metal to the CPU & GPU dies.

    Predator 17: Last run of Witcher 3 with an ambient of 25C, max CPU 80C, max GPU 85C which fine, average temps will be less as these figures are the max values from HWinfo64 as the Predator 17's do not chart plot the temps in PredatorSense.  Your Helios 300 looks to be running at around 85C on the CPU and less on the GPU.  I wouldn't be concerned about the temp spikes as it just the system doing something in the background and of no consequence unless for prolonged periods.

    Q-6