Legion 5 Pro Thermal Paste FYI it is dry (on my unit)

Discussion in 'Lenovo' started by Spring1898, Jun 4, 2021.

  1. Spring1898

    Spring1898 Notebook Consultant

    Reputations:
    96
    Messages:
    266
    Likes Received:
    53
    Trophy Points:
    41
    Pulled apart the Legion 5 Pro 3070 version today.
    Removal of the back cover is as covered in many videos, plastic pry recommended.
    There are 2 different screw head sizes needed, PH1 and PH000

    To remove the heatsink there are 6 screws on the heat sink, 3 screws around each fan, and 2 additional screws (the third is shared with the fans) on each of the SSD heatsinks that need removed. Don't forget the fan cables

    The Thermal Paste on my unit looked liked standard thermal paste. However it was dried out and crumbly, not at all like putty or pliable.
    The heatsink was stuck on so hard I thought there were additional screws holding it in somewhere. Gentle constant pressure and the entire heatsink lifted off at once.

    Normally my approach is too push / scrape off as much as I can with the plastic pry or a card before the alcohol so as not to smear, but since it was so dried and stuck it was more like scraping paint. Had to be really careful around the edges of the die so as not to damage anything. I ended up leaving a lot around the edges of the PCB because it wasn't worth the effort to get every last bit off after the dies were cleared.

    Arctic MX-4 was the replacement at this time since I had it on hand.

    Replace the heatsink and reconnect the fan cables before putting on the SSD heatsinks.

    Based on what I found I would recommend replacing the thermal paste on arrival.
    Sorry no thermal testings as I don't have time to run benchmarks these days.


    Side note about kryonaut and other pastes:
    A couple years back I purchased kryonaut to test against my older standbys MX-4 and AS5. The best done tests I had seen (multiple real world applications and retests rather than a single application) had these two pastes at or near the top of the pack while being inexpensive and having proven longevity, though none took into account AS5's curing over time which is supposed to improve over time. MX-4 is supposed to remain viable for 7+ years after application.

    When Kryonaut hit the market I bought a tube to try it out a couple years after release given the purported better numbers. After 2 separate applications I did not notice enough of a difference to justify it (results were within margin of error). The remainder of the tube sat for about 2 years. When I went to use it again, the entire tube was dried out, like plumber's putty and was no longer able to be applied. This was sealed with the lid on tight inside their resealable bag. I don't know whether this applies to the thermal paste when under heating/cooling and pressure under a heatsink, but it sapped all my confidence in the longevity of the product. Perhaps they have improved it since then.




    [​IMG]

    [​IMG]
     
    Last edited: Jun 5, 2021
    GrandesBollas, Atma, c69k and 2 others like this.
  2. alaskajoel

    alaskajoel Notebook Deity

    Reputations:
    1,086
    Messages:
    1,026
    Likes Received:
    951
    Trophy Points:
    131
    Do my eyes deceive me, or do those fans say 10v 0.8A? Can't say I've ever seen a 10v fan in a notebook.
     
    Darkhan, Papusan and Vistar Shook like this.
  3. MrMcGibblets86

    MrMcGibblets86 Newbie

    Reputations:
    0
    Messages:
    1
    Likes Received:
    0
    Trophy Points:
    5
    Thanks for sharing. I received my L5Pro a few days ago and the AMD 5800H hits 93-95 C when running Time Spy -- even when I'm using a cooling pad.

    Idle temps are around 58C in a 75F room.

    I too have Kryonaut (from 2 years ago to repaste my Alienware m15) and it has also dried out.
    Just bought some MX-4 three months ago to repaste the Alienware and now I am thinking about doing the same to the L5Pro.

    Here are my benchmark results after replacing the stock RAM:
    https://www.reddit.com/r/GamingLaptops/comments/nxog1t/update_legion_5_pro_usa_rtx_3070_64gb_ram/
     
  4. Muyfa666

    Muyfa666 Notebook Evangelist

    Reputations:
    135
    Messages:
    629
    Likes Received:
    418
    Trophy Points:
    76
    I just repasted my Legion 5 Pro with MX-4. It's ok, not great. I will give some days to see if it cures. Otherwise I'll try Prolimatech PK-3 and see it if makes any difference.

    Very happy with the unit thou.
     
  5. Spring1898

    Spring1898 Notebook Consultant

    Reputations:
    96
    Messages:
    266
    Likes Received:
    53
    Trophy Points:
    41
    Shouldn't be any cure time with MX-4. I am old school when it comes to thermal paste application. Most studies now show that too much thermal paste doesn't insulate or negatively affect thermals as much as was once thought, on dies with heat spreaders. I experimented a long time with ultra thin line applications on bare dies in the ivy bridge generation, and these seemed to do better. But we are talking the bare minimum to get the paste to the edge of the die. This was driven primarily due to the relatively high viscosity of Artic Silver 5

    That said, if you have the thermal head room and stability, the laptop will utilize it and take your temps all the way to the thermal limit (IE no temp change). And the efficiency of these newer chips vs older means that simply a little more thermal headroom doesn't give the small edge in performance it used to. The only way to really tell is if there is an objective change are canned benchmarks or change in average frequency.

    On Blower style cards (without RPM limited fans) I would get at best a 3% improvement over stock thermal paste in terms of performance. Temps would be the same or slightly cooler (91-92c vs 95c).

    EDIT: If you do try the prolimatech I would be interested to see if there is a clinically significant difference over the MX-4
     
    Last edited: Jun 17, 2021
  6. Muyfa666

    Muyfa666 Notebook Evangelist

    Reputations:
    135
    Messages:
    629
    Likes Received:
    418
    Trophy Points:
    76
    As the repaste was very to do, I will try the Prolimatek on monday I think. Going away for a few days now.
     
    Papusan likes this.
  7. Muyfa666

    Muyfa666 Notebook Evangelist

    Reputations:
    135
    Messages:
    629
    Likes Received:
    418
    Trophy Points:
    76
    No real difference with PK-3. Also, I found no obvious application errors from the MX-4.

    Disabeling CPU turbo gives great results thou;

    Tested with 3DMark Time Spy:

    (Turbo Enabled)
    Total: 10866
    GPU: 11045
    CPU: 9953
    Max temp: 99C

    (Turbo Disabled)
    Total: 10770
    GPU: 11034
    CPU: 9489
    Max temp: 80C

    No big performance loss at almost the same performance.

    Also, this is tested in a heat wave, so ambient temps are almost 30C.
     
  8. Spring1898

    Spring1898 Notebook Consultant

    Reputations:
    96
    Messages:
    266
    Likes Received:
    53
    Trophy Points:
    41
    Thanks for the results. It would have been surprising if there was much of a difference. All the tests that I saw almost always had most of the best pastes within a few tenths of a degree from each other, which to me spells margin of error differences.

    So pick whichever is cheaper and has the best longevity.

    Good to know about the turbo also. Unless you need that extra 5% of performance, the temp difference is substantial.
     
    Last edited: Jun 21, 2021
  9. bigspin

    bigspin My Kind Of Place

    Reputations:
    632
    Messages:
    3,952
    Likes Received:
    566
    Trophy Points:
    181

    I have Legion 5 but both models share the same heat sink and M/B. I re-pasted and changed thermal pads as well. I used Cooler Master MasterGel Maker amd gelid extreme thermal pads (You only need 2x 1mm pads)

    I tested system with BF5 Multiplayer. System run at performance state.

    before re-paste CPU average 90C /GPU average 76C (Ambient 22C)
    After re-paste CPU average 88C/GPU average 68C (Ambient 22C)

    i start the temp monitoring software just before game start. What i observed is fans always slow to rev up to maximum speed. So CPU/GPU always show 3-4C higher maximum temperature when game started and once fans started to run at maximum speed temp come down and settled to average value. That is why i always check the average temperature.
     
  10. bigspin

    bigspin My Kind Of Place

    Reputations:
    632
    Messages:
    3,952
    Likes Received:
    566
    Trophy Points:
    181
Loading...

Share This Page