The Notebook Review forums were hosted by TechTarget, who shut down them down on January 31, 2022. This static read-only archive was pulled by NBR forum users between January 20 and January 31, 2022, in an effort to make sure that the valuable technical information that had been posted on the forums is preserved. For current discussions, many NBR forum users moved over to NotebookTalk.net after the shutdown.
Problems? See this thread at archive.org.

    razer blade 14 temperature fix

    Discussion in 'Razer' started by thp777, Jan 18, 2017.

  1. thp777

    thp777 Notebook Enthusiast

    Reputations:
    2
    Messages:
    30
    Likes Received:
    8
    Trophy Points:
    16
    I ordered myself a razer blade 14 (2014) and even before getting it in I knew I would need to do a little modding to the cooling so I ordered IC diamond thermal compound and some assorted grit wet/dry sandpaper for lapping the heatpipes. Ran aida64 stress test before any modding and was reaching 99c on most cores and throttling slightly below 2793mhz occasionaly during the testing. After dissassembling the laptop and removing the heatsink I cleaned out the old thermal paste from between the heatpipes with alcohol. then i started at 600 grit and worked up to 2000 grit lapping and polishing the areas of the heatpipes that touch the cpu and gpu. After that i wedge the ic diamond thermal paste down under and between the heatpipes in the area the chips touch and smoothed it even with the top of the heatpipes using a blade. Then i applied some past to the top of the gpu and cpu and gave it a little heat with a heatgun for a second to get it more pliable when i set on the heatsink. After reassembly and stress testing again the temps had dropped significantly(see pictures below) and the throttling below 2793-4 mhz had stopped. to further drop temps i also undervolted the cpu afterwards and dropped another 2-3c on each core with a -50mv undervolt on the cpu. I later dropped it another 1-2c bypolishing the heatpipes where the fans blow air over them slightly on the underside. overall i am very happy with the temps and noise now and hope this helps anyone else hoping to cool their blade down to where it SHOULD be if the cooling had been designed properly for a $2000+ laptop.

    If anyone has any questions, ideas or need any help doing this to yours just ask.

    Ive got a second heatsink as well that im in the process of replacing the aluminum fins with copper ones hoping to drop it a bit more but its a bit tricky keeping everything lined up and attached when reflowing the solder everything heats evenly so i need to make a jig of sorts to hold it all in place. Ill also be filling in the gaps with 96.5%tin/3%silver/.5%copper solder and attempting to build up some solder on the contact areas to lap to a perfect surface that contacts 100% of the without gaps.

    Also coming soon.....magsafe mod :D

    Managed to pick this up in perfect brand new condition.

    [​IMG]

    the surgery begins

    [​IMG]

    This is just a mess and unacceptable

    [​IMG]

    terrible contact areas from factory

    [​IMG]

    Before lapping heatsink is very rough and pited

    [​IMG]
    half way thru lapping and polishing

    [​IMG]

    After polishing

    [​IMG]

    Results after just polishing and applying ic diamond compound

    [​IMG]
     
    SimplyJ3sse, Revoluxon and Lauski like this.
  2. Lauski

    Lauski Notebook Consultant

    Reputations:
    85
    Messages:
    249
    Likes Received:
    34
    Trophy Points:
    41
    Great job. There have been plenty of stories that went the other way when trying to improve the cooling.

    I mean, with your technique.. you can now use not only ICD, but gelid GC extreme or even ultra liquid metal.

    And you think you shaved it down by 1-2mm at most? And let me know anything you did to ensure you went down squarely in order to keep contact flush.

    Edit: and can you indicate where you polished near the fins to get that final 1-2C? Not sure exactly where you mean since it sounds like you hadn't yet detached the fins yet at that point.

    Your results put the RB2016 into my consideration list. Thanks for your bravery!
     
    Last edited: Jan 18, 2017
  3. Makyura

    Makyura Notebook Consultant

    Reputations:
    29
    Messages:
    249
    Likes Received:
    118
    Trophy Points:
    56
    Are the pictures showing temperatures "before" and "after"?

    In that case, great job, might give it a try.
     
  4. thp777

    thp777 Notebook Enthusiast

    Reputations:
    2
    Messages:
    30
    Likes Received:
    8
    Trophy Points:
    16
    Id say maybe 0.3mm at most, the pipes are fairly thin so you dont want to go down to far. I used little 1inch glass/mirror squares with the sandpaper glued on to them and alcohol to wet the paper. 

    yes those are before and after tested with room temp of 71F. the after test didnt run the fans at full speed either so theres room for some overclocking for sure.
     
  5. Makyura

    Makyura Notebook Consultant

    Reputations:
    29
    Messages:
    249
    Likes Received:
    118
    Trophy Points:
    56
    Damn man, that's impressive then, congratulations!
    Wish you lived in Italy, I'd absolutely pay you to do this for me on my unit as I'm too scared to ruin the heat pipes, lol.

    Guess I might try it in a not too close future.
     
    Last edited: Jan 18, 2017
  6. thp777

    thp777 Notebook Enthusiast

    Reputations:
    2
    Messages:
    30
    Likes Received:
    8
    Trophy Points:
    16
    Where the fins attach to the heatpipes, I polished the opposite side where the heatpipes are exposed. Not alot, just a quick run thru the grits to get a cleaner surface.
     
  7. Lauski

    Lauski Notebook Consultant

    Reputations:
    85
    Messages:
    249
    Likes Received:
    34
    Trophy Points:
    41
    Gotcha, that's what it sounded like in your original post. Are they in contact with anything on the back side there?
     
  8. thp777

    thp777 Notebook Enthusiast

    Reputations:
    2
    Messages:
    30
    Likes Received:
    8
    Trophy Points:
    16
    no but a little bit of air flows over them from the fan and they arent very clean. probably just made a difference getting that oxidized layer of crap off of it.
     
  9. link626

    link626 Asus GL502VM, Lenovo Y580, Asus K53TA

    Reputations:
    209
    Messages:
    1,754
    Likes Received:
    213
    Trophy Points:
    81
    i would have been interested in results of the ICD alone, before the lapping.
    that would have been a more proper test.
     
  10. thp777

    thp777 Notebook Enthusiast

    Reputations:
    2
    Messages:
    30
    Likes Received:
    8
    Trophy Points:
    16
    My goal was to fix it properly. Anyone redoing thermal paste should be lapping it so there arent large areas of thermal compound that it has to move through. You also dont want to reapply ic diamond on exposed dies to many times, diamond is abrasive and really the less you mess with an exposed die the better. Just takes one wrong move to chip a corner and brick the entire thing.

    This is generally what you can expect from ic diamond and similar results from other high end compounds like artic silver etc
    [​IMG]
     
  11. Lauski

    Lauski Notebook Consultant

    Reputations:
    85
    Messages:
    249
    Likes Received:
    34
    Trophy Points:
    41
    Plus there have been several threads in the Razer area of the forums since the 970m Blade that surprisingly have shown several people not getting improvements to temperatures when simply re-pasting with a huge variety of great pastes-- that I know included ICD.

    There even was some amount of discussion the official razer forums where the CEO responded stating that they intentionally put the paste on as they do, and that it's designed for a direct-pipe solution that's in the blade.

    You know, it would sound like a typical response at first glance, but numbers don't lie. One guy after re-pasting 20 some-odd times, sold his laptop to someone else on NBR who picked up the torch trying to get temperatures BACK to stock. (Yes ICD was making temps worse than at stock, I forgot to say that). That's why I was so especially impressed by what thp777 did here.
     
  12. thp777

    thp777 Notebook Enthusiast

    Reputations:
    2
    Messages:
    30
    Likes Received:
    8
    Trophy Points:
    16
    I was really suprised by the results i was expecting around 10c if i was lucky. i need to get fan control so i can redo the test with fans on full load. they wont even spin up as fast as the first test was so im sure at same rpm its even better.
    I got some more ideas to improve the cooling that im playing around with in CAD and some CFD software while i opening it up last night for another mod but heres some eye candy until then.
    [​IMG]
    Built in Razer Naga Epic Chroma & Astro A50 headset Wireless Adapters.
    [​IMG]
     
  13. RageBone

    RageBone Notebook Enthusiast

    Reputations:
    5
    Messages:
    35
    Likes Received:
    4
    Trophy Points:
    16
    @thp777
    I noticed that your RAM chips are different ones then then chips that my two RB 2014 have, Hynix chips that i can find samsung datasheets on.

    "SEC 425 BYK0". Google turns up "K4B2G08460".

    Can you please provide partnumbers ?
    Thank you!

    And to the latest mod, I see what you did there. Nice work! Hope, the USB port still works as intended.

    @Lauski you wrote that you now consider buying a new RB 2016 and moding it this way.

    I feel forced to remind you that opening the RB will void your waranty and i don't know if thats such a good idea!

    Openeing it, might not be noticeable, but moding the cooling, defently is!

    Well, i'd do that because i'm probalby one of the few people who "can" repair RB 2014 motherboards, so i could probably figure out the 2016th boards too, but that is not importainnt right now...

    @Makyura
    I'd might be able to to offer that as a service in the future. Location Germany.

    But before that, i need to try the modification myself.
    I'll report on it!

    EDIT: IC Dimond is on the way.
     
    Last edited: Jan 21, 2017
  14. thp777

    thp777 Notebook Enthusiast

    Reputations:
    2
    Messages:
    30
    Likes Received:
    8
    Trophy Points:
    16
    it uses up a usb port for now. but itll have a 4 port usb3.0 hub soon to add wacom intuos pro wireless reciever and 3dconnexion spacemouse wireless reciever as well.
    the ram chips are micron branded.
    Im tempted to sell my desktop and replace it with a Blade 14 2016 1060 and a razer core now and of course repeat all these mods on it. Id probably keep my 2014 unless i can get back more than what i payed for it, it is one of a kind after all.
     
  15. RageBone

    RageBone Notebook Enthusiast

    Reputations:
    5
    Messages:
    35
    Likes Received:
    4
    Trophy Points:
    16
    I repasted my RB 2014 today and IC D 24c did make a difference compared to the old stock, mixed with mx4.

    I messed stuff up a bit, you'll see it on the pictures.

    Stock + mx4
    RB_TP_V1 (2).png

    Ic D 24 carat. No throtteling, like 5 to 10 degrees less, maybe.
    RB_TP_V2.png
    An improvement!

    Test method:
    XTU cpu stress test.
    Furmark 1080p 4x smaa

    @thp777 Im interested in the Magsafe mod that you teased.
    I have also spend time thinking about it, and i don't know if its such a good idea to push 150W through the magsafe connector which normaly does like 80W max.

    When you have the 4port usb hub in your RB, you could mod the USB port that you disabled with your contraption.
    I mean dissconnect the data lines from the port, and hardwire a port from your hub to it.
    Maybe ...
     
    Last edited: Jan 27, 2017
  16. RageBone

    RageBone Notebook Enthusiast

    Reputations:
    5
    Messages:
    35
    Likes Received:
    4
    Trophy Points:
    16
    And more pics

    Topview.
    P_20170127_153849.jpg
    luckily you can't see my work-around for a power delivery probelm.

    Here is the GPU, cleaned it as much as i could, looks like i have a "monday" RB with all that red schmouu on the gpu.
    P_20170127_154715.jpg

    ICD after the first test and before messing it up.
    P_20170127_163843.jpg

    After putting a lot more on there.
    P_20170127_165244.jpg

    Thats it for now..
     
  17. Lauski

    Lauski Notebook Consultant

    Reputations:
    85
    Messages:
    249
    Likes Received:
    34
    Trophy Points:
    41
    @RageBone : Nice on better temps. Did you polish the pipes at all or just simply change out paste?
     
  18. RageBone

    RageBone Notebook Enthusiast

    Reputations:
    5
    Messages:
    35
    Likes Received:
    4
    Trophy Points:
    16
    i just changed paste, but i will polish the pipes later too.
     
  19. thp777

    thp777 Notebook Enthusiast

    Reputations:
    2
    Messages:
    30
    Likes Received:
    8
    Trophy Points:
    16
    @RageBone Glad to see the ic diamond is working for someone else as well.

    Please excuse any typos and bad grammar. Im usually typing before bed.

    In regards to the magsafe im not sure if it will handle that power either. its rated at 4.6amps on the 85w charger so thats as high of amps that there is pushed thru it. Ive seen magsafe mods done before but not on anything above 3 amps.
    taking the connectors apart they are solid pins about that taper down to 24awg on the back side. and then they are connected in pairs so that increases the amps. all i can say is 7amps for sure and at full load(150w) it would need to handle 7.9ish. From testing ive done so far the most ive been able to pull is 100watts with the laptop even in aida 64 and furmark, so id only have to worry about around 6 amps give or take accounting for effeciency etc etc. 24awg solid core is rated at 3.5 amps recommended continous times 2 for the double pin design = 7 amps

    With the 4 port usb hub, yes i will be re-enabling the usb port on the side. ive already laid out where i will be cutting the traces and splicing it in. Im going to have to do something clever with flat pcb or flexible pcb antennas for the wireless receiver to boost them a little. I need to try the spot where there is a mesh above the power switch for an antenna. If that doesnt work ill see if theres room in the plastic section on the screen where the wifi antennas are.

    I also recently ordered a PCB with the touchscreens and the buttons from a switchblade ui out of a blade pro for next to nothing that im going to play around with making into a stand alone pad that will slot into the right usb port just because i can and it was only $13 haha

    Still on the search for some suitable copper fins for the heatsink to replace the aluminum ones. I have a few prospects from china im waiting on
     
    Last edited: Jan 28, 2017
  20. Makyura

    Makyura Notebook Consultant

    Reputations:
    29
    Messages:
    249
    Likes Received:
    118
    Trophy Points:
    56
    Sorry for necroing, @OP:

    How many karats version was the ICD you've used?

    I had no clue there were so many different ICD pastes, damn.

    Thank you
     
  21. Mobius 1

    Mobius 1 Notebook Nobel Laureate

    Reputations:
    3,447
    Messages:
    9,069
    Likes Received:
    6,376
    Trophy Points:
    681


    no different, 7 and 24 is just gram content you receive on the tube
     
    Makyura likes this.
  22. Makyura

    Makyura Notebook Consultant

    Reputations:
    29
    Messages:
    249
    Likes Received:
    118
    Trophy Points:
    56
    Thank you!