*OFFICIAL* Alienware 13 R3 Owner's Lounge

Discussion in '2015+ Alienware 13 / 15 / 17' started by katalin_2003, Oct 24, 2016.

  1. c69k

    c69k Notebook Deity

    Reputations:
    588
    Messages:
    1,188
    Likes Received:
    884
    Trophy Points:
    131
    Hi,

    so I finally got 32x32x0.2mm Thermal Grizzly Carbonaut in my hand for 12 EUR (+6.50 EUR post) from Highflow.nl, took exactly 24 hours to be delivered.

    First impressions ... completely different from Panasonic PGS, there is no plastic, and it (feels) thicker than expected @ 0.2 mm, like it is 0.4 mm.
    There is no material that I ever touched in my life with such properties, but if I really wanted to compare, it would be like very high quality expensive silk.

    It is porous so I believe it can accomodate LM, will try after my nighshift is over tomorrow morning. The size I bought 38x38 should be comfortably enough for 2 re-pads (CPU+GPU), plus a bit for testing with the LM beforehands opening currently perfectly cooled laptop.:vbthumbsup:

    Well, on Grizzly page they say that pressure is not really necessary for carbonaut (?), just the fit and contact has to be good. I so hope this works, as it would negate all disadvatages of LM (being runny, conductive off dies, even pressure requirement and 'under +8C' solidifying problem). Carbonaut will be used as sort of 'structural support' for LM with both having same WM/k.

    I will post my results as soon as I am done.

    Testing:

    Will turn on fans to performance in Bios to rule out fan factor, with 30 minutes of AIDA64 (fpu and GPU(s), with currenttly used LM, and then the same test with mix of LM+Carbonaut.
     
    Last edited: May 22, 2019
  2. propeldragon

    propeldragon Notebook Evangelist

    Reputations:
    117
    Messages:
    536
    Likes Received:
    364
    Trophy Points:
    76
    I hope your quest is successful brother.
     
    c69k likes this.
  3. c69k

    c69k Notebook Deity

    Reputations:
    588
    Messages:
    1,188
    Likes Received:
    884
    Trophy Points:
    131
    Thanx, I really appreciate it.

    I really hope this works, as then I would have the best of both worlds. Perfect contact (LM) and 10 G-force safety against LM spilling (with structural support from Carbonaut) while having highest Wm/k, all in one.
    My die contacts would benefit from better pressure to heat-sink for sure.

    I will take a lot of pics how Gelid pads and LM are looking after one year of heavy use.
     
  4. propeldragon

    propeldragon Notebook Evangelist

    Reputations:
    117
    Messages:
    536
    Likes Received:
    364
    Trophy Points:
    76
    Sounds good! Looking forward to the pics. May the force be with you.

    I might try out the new heatsink before I do the vbios mod to make a more accurate comparison. I would have to do a total repad again...
     
    c69k likes this.
  5. Papusan

    Papusan JOKEBOOK's Sucks! Dont waste your $$$ on FILTHY

    Reputations:
    20,755
    Messages:
    22,502
    Likes Received:
    38,694
    Trophy Points:
    931
    +rep bruh.
     
    Vasudev and c69k like this.
  6. Rei Fukai

    Rei Fukai Notebook Deity

    Reputations:
    702
    Messages:
    756
    Likes Received:
    904
    Trophy Points:
    106
    Hey man ! i had the same idea, but i was afraid that LM and graphene could violate eachother and eat away the CPU die. i have the panasonic one, (not the rebranded, but the original with 1600 wm/k)
    I've also ordered carbonaught to see which one behaves better (and i also had the idea to use a shim for more pressure)

    if you want idea's i got plenty of em, i even though about using them in the heatsink fins (small strips)or to just to spread the heat as even as possible over the die area, or even create a bridge to the GPU coldplate, to help the CPU dissipate heat. My GPU is undervolted, not crossing 75C when gaming with framerate unlocked on 1440P. I'm looking forward to your results. i probably will start my journey next week
     
  7. c69k

    c69k Notebook Deity

    Reputations:
    588
    Messages:
    1,188
    Likes Received:
    884
    Trophy Points:
    131
    Here are results of Carbonaut and TG Conductonaut while trying to combine them. ~~FAIL~~

    I painted a VERY thin layer of LM on a die, placed die-size sheet of Carbonaut on it, all looked perfect, Carbonaut did not want to move a bit, did not 'float' so I thought, OK !!! It absorbs !!!
    Then I started painting on top of the sheet already placed on die, wanting some LM to soak in.
    First it looked that Carboanut absorbs LM, but when the sheet looked all nice painted with LM, suddently half of the pad lost grip on the top layer of LM and exposed naked Carbonaut sheet.
    As I tried to re-paint many times, Carbonaut sheet started twisting (like a thin, heated up peace of plastic sheet), then came apart and tore in half. At this point I naturally aborted this little experiment.
    I looked at the Carbonaut sheet which I placed on LM'd die, and there was also no LM sticking at the bottom side of Carbonaut sheet.

    I wanted to re-use the LM that was left on the die, but it would not stick at all, probably some chemical in Carbonaut simply made it impossible to spread. I removed this 'Carbonauted LM', cleaned the die with alcohol and applied new LM, all good.


    Anyways, I stay with LM and Gelid pads.

    Before openeing, I left computer to completely cool down over night, but still, 2 pads around memory chips got torn apart on heat-sink removal, nothing major, could have just re-use them easy, as they tore clean.
    After a year Gelids looked healthy as new. LM was all liquid on dies, and also on facing part of heat-sink. What I could not understand was that LM on heat-sink that had absolutely no contact with dies stayed liquid too.

    I was a bit worried about an extra 0.2 mm of Carbonaut sheet tripping the whole balance of heat-sink, so I changed all Gelid pads for new.
    I also opened the fan casing, to check for dust build up, but there was absolutely no dust, only 3-4 cat hair LOL I vacuum clean fans sucking the air from bottom once a month, so I better keep doing that.

    Right after assembly, I removed all undervolts and now I run AIDA64 fpu+Gpu(s) so any Gelid pads that stick out will flatten by heat and fresh heat-sink arm tension pressure.
    Luckily temperatures are good, CPU cores are the same temperature.

    I did not try 'Carbonaut only' on dies, as it surely can't beat proper LM.

    :::::EDIT:::::

    I promised pictures of 1 year LM and Gelid pads. I hit door frame HARD with left bottom corner of laptop 4 months ago while walking to another room, see the LM line near processor die? 1 core got 4C hotter than the rest within a week. LM is missing on the bottom part of CPU.

    IMG_20190524_125939.jpg
    Annotation 2019-05-25 003934.jpg
     
    Last edited: May 24, 2019
    Ashtrix, hans12321, Papusan and 5 others like this.
  8. Papusan

    Papusan JOKEBOOK's Sucks! Dont waste your $$$ on FILTHY

    Reputations:
    20,755
    Messages:
    22,502
    Likes Received:
    38,694
    Trophy Points:
    931
    +rep

    Maybe you'll test it next time:)

    Nothing is equal as Liquid metal but would be nice got some results from the way Carbonaut is mean't to be used.
     
    Vasudev likes this.
  9. Wilhelm_SI

    Wilhelm_SI Notebook Enthusiast

    Reputations:
    15
    Messages:
    23
    Likes Received:
    18
    Trophy Points:
    6
    Hi guys, i was wondering how good your 7700hq and gtx1060 performance is?

    After hours of repasting, re-padding and tweaking i think, i´m hitting a wall now. Perhaps some of you have additional advise

    What did i do:

    - repaste LM Thermal Grizzly on CPU, GPU and PCH (little heatsink for pch is coming)
    - new pads from arctic silver (0,5mm and 1mm)
    - uv -115,2 (Cache and Core), -35,2 IntelGPU
    - active cooler with Noctua NF-P12 and pwm cudos to "Sk0b0ld" from the aw-community forum. He came up with this idea
    [​IMG]
    [​IMG]

    Which test´s did i do:
    Ghost Recon Wildlands (GRW) and Prime 95 (small FFTs) + Heaven Benchmark 4.0 - 20 minutes warm up, then reset of hwinfo + 45 Minutes
    - Laptop lifted 5cm
    - U2 cooler + lifted 5cm
    - U2 cooler + alien plain
    - U2 cooler alien plain without backplate
    Modded U2 cooler always on 100%

    [​IMG]

    I tried to do another test with locked fans, but if i set cpu fan, the same value is applied for gpu. But i wanted to have different values for cpu and gpu fans.
    Tried it with NBFC, but failed to set up a fan profile.

    Today i oc´ed my gtx1060 +180mhz gpu clock and 330mhz mem clock. Testing is imminent.
     
  10. c69k

    c69k Notebook Deity

    Reputations:
    588
    Messages:
    1,188
    Likes Received:
    884
    Trophy Points:
    131
    @Wilhelm_SI

    Your cooling is already perfect. All I can think of is, maybe you would like to have max. TDP of the GPU increased to, say 110W=no more power throttling?
     
    Last edited: May 30, 2019
Loading...

Share This Page