*OFFICIAL* Alienware Area-51M Owner's Lounge

Discussion in '2015+ Alienware 13 / 15 / 17' started by ssj92, Jan 8, 2019.

  1. HaloGod2012

    HaloGod2012 Notebook Deity

    Reputations:
    631
    Messages:
    1,664
    Likes Received:
    1,349
    Trophy Points:
    181
    Yep, has been happening since the last bios update. Strange


    Sent from my iPhone using Tapatalk
     
  2. Papusan

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

    Reputations:
    21,100
    Messages:
    22,765
    Likes Received:
    39,281
    Trophy Points:
    931
    Not strange at all. Dell's firmware engineers fiddled with fan profiles in last bios version. Probably screwed up pretty well while they tried a new fix to reduce noise... "Reduced fan noise during BIOS update process".
     
    raz8020, Vasudev and Darkhan like this.
  3. HaloGod2012

    HaloGod2012 Notebook Deity

    Reputations:
    631
    Messages:
    1,664
    Likes Received:
    1,349
    Trophy Points:
    181
    Lol, yeah before the fans would go to 100 percent while updating the bios. This is the fix they care about before giving us unlocked ram ?


    Sent from my iPhone using Tapatalk
     
    Papusan likes this.
  4. Papusan

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

    Reputations:
    21,100
    Messages:
    22,765
    Likes Received:
    39,281
    Trophy Points:
    931
    You forgot add in the long promised "increased power utilization from single PSU":vbbiggrin:
     
    raz8020, Vasudev and Darkhan like this.
  5. SacraficeMyGoat

    SacraficeMyGoat Notebook Evangelist

    Reputations:
    44
    Messages:
    512
    Likes Received:
    321
    Trophy Points:
    76
    I know right.. No word on this, no word on 2070/2060 TDP vBios update, etc etc etc.
     
  6. S.K

    S.K Batch 80286

    Reputations:
    274
    Messages:
    806
    Likes Received:
    906
    Trophy Points:
    106

    Before I return one motherboard to Dell tomorrow, I decided to give both motherboards a go and monitor performance / power limits / overclocks and GPU performance. Here is what I found after a full day of testing and after 4 back and forth motherboard swaps. For this context only, I'll call them by their PCB color. Green(with NIKOS inlet mosfets) vs Blue(with AON inlet mosfets) (easier to convey things by sticking to just one name for each).

    The Green one (with NIKOS mosfets) is able to provide more power to the components (both CPU and GPU) at a more stable and sustained rate. The GPU is able to hold it's clocks much better without jumping up and down a lot and I was able to undervolt the CPU at 5 GHz better on it.

    On the other hand, the Blue one (with AON mosfets) appeared to be choking the life out of components in terms of stable power delivery under load conditions (most likely because of it's lower max current ratings and higher thermal resistance? (There are a LOT of other components involved as well so this may or may not be 100% accurate for your board). I was not able to sustain stable overclocks on the CPU at the same undervolt as I was able to maintain on the Green one. Also, the PCIE bus went unstable when I overclocked the CPU to 5 GHz across all cores and I started getting random weird errors and warnings in the windows system event logs for network card and what not. Additionally, the GPU wouldn't stick to the same working Afterburner profile anymore and downclock itself to a much lower (70-75 MHz below given 1950 MHz in the profile) at the same 937mV and was STILL unstable! I kept on receiving crashes for the GPU and it was also running noticeably hotter (around 4-5 degrees hotter), perhaps due to a significant amount of noise on the power lines. (I tried plotting some curves on the oscilloscope but unfortunately couldn't manage as my little girl found those moving lines really interesting and wanted to exchange my oscilloscope with me for her stuffed penguin and I had to give up the idea lol! Plus, it was 4 am!)

    I also noticed that the DGFF contact point (where those 2 ribbon connectors go onto the motherobard) were crisp bright golden colored on the Green one while they were greenish and grey-ish on the Blue board and no, they didn't change color after a prolonged cleaning session with iso-propyl alcohol. They looked like they were made from some really cheap material and that makes me suspect if one of Dell's manufacturers in China cheaped out on the quality of metal that they used to make those PCBs. The contacts on the green board looked gorgeous but the Blue one looked like a corroded car battery terminal! Since that part of exposed metal is actually the same as used inside the motherboard wiring itself, it's safe to assume that it was overall a cheaper quality motherboard itself also which could add extra noise to the circuit under load situations (not talking about the parts here).

    This leads me to a somewhat broad conclusion about these motherboards. The Green one with the parts mentioned seems to be more stable under load and because of higher absolute max ratings, is able to provide more consistent power without any jolts and spikes + better overall PCB build quality while the Blue one apparently suffers from noise on the power lines. On paper, it seem to be not because of the mosfets but from experience, I can say that it's always a good design when the mosfets are beefier than necessary and even under max load, have a large margin to spare before they reach their max specs. The 330W PSU pumps out 17 Amps which could vary and 24A absolute max for inlet mosfets isn't good enough in this case. The NIKOS are 50A max which is really overkill, and since they are running WAY under max specs, they run relatively cooler (and also have lower junction to case thermal resistance (4 vs 4.8 °C/W max) which means they can get rid of the p-n junction heat better).

    I hope this helps someone out in the right direction. I have finally put my laptop back together with the green board and called it a day of successful testing. The blue one will be going back to Dell tomorrow. :)
     
    Last edited: May 16, 2019
  7. custom90gt

    custom90gt Doc Mod Super Moderator

    Reputations:
    6,807
    Messages:
    3,057
    Likes Received:
    3,197
    Trophy Points:
    331
    I'm not surprised by that. For me, Conductonaut has always been only a couple of degrees cooler than Kryonaut. You may have a bad batch of Kryonaut on your hands if it's much worse than MX4.
     
    toughasnails, Vasudev and S.K like this.
  8. SacraficeMyGoat

    SacraficeMyGoat Notebook Evangelist

    Reputations:
    44
    Messages:
    512
    Likes Received:
    321
    Trophy Points:
    76
    Thanks for your in depth analysis!

    Out or curiosity, I'd like to check what board I have (most likely first revision) but what would be the simplest way to go about this?

    Do the mosfets say Nikos/AON?
     
    Fire Tiger likes this.
  9. S.K

    S.K Batch 80286

    Reputations:
    274
    Messages:
    806
    Likes Received:
    906
    Trophy Points:
    106
    The CPU power delivery mosfets on the Lemon are Alpha & Omega Z5038QI5 while on the green one, it's Alpha & Omega Z5038QI without the number 5. I couldn't find a datasheet for the lemon part number ANYWHERE! It's weird, this doesn't happen very often. Plus, the lemon has AON 6380 on it as you can see in my quoted post in the above message.

    Here is a picture of the Z5038QI chip. The other one looks similar apart from the added number 5 to the part number.
    [​IMG]
     
    Last edited: May 14, 2019
    SacraficeMyGoat likes this.
  10. SacraficeMyGoat

    SacraficeMyGoat Notebook Evangelist

    Reputations:
    44
    Messages:
    512
    Likes Received:
    321
    Trophy Points:
    76
    Thanks! Happen to have a picture of each?

    Edit: NVM LOL

    As expected my board has the "lemon" mosfets.
     
    Last edited: May 14, 2019
    S.K likes this.
Loading...

Share This Page