Dell XPS 15 (Skylake) 9550 Owner's Lounge

Discussion in 'Dell XPS and Studio XPS' started by T2050, Oct 30, 2015.

  1. Splitframe

    Splitframe Notebook Guru

    Reputations:
    7
    Messages:
    56
    Likes Received:
    20
    Trophy Points:
    16
    Can someone help me? When I last cleaned my 9550 with a very soft brush the little component in the pic was loose.
    It has three legs, two on one side and one on the other, the single one isn't connected to the board anymore.
    It was already soldered when I got the XPS 9550 back from a warranty once.
    It looks like the pad is off as well so I don't know if I can solder it on again.
    The notebook seems to run normal on first glance, but I had a BSOD since and weirdly it really lags over Remote Desktop now.
    Benchmarks run fine though. I am still testing if the BSOD was some one off thing.
    I searched for schematics, but couldn't find what type of component this is.
     

    Attached Files:

    pressing likes this.
  2. pressing

    pressing Notebook Deity

    Reputations:
    257
    Messages:
    1,760
    Likes Received:
    857
    Trophy Points:
    131
    Splitframe,

    The search function is broken at NBR so it is very difficult to search solutions now. I don't remember anyone having any issues with QE5. And I can't find any high definiton motherboard photos right now.

    The apparently failing "QE5" seems to be bipolar transistor related to thermal diode scheme (this throttles your computer when it gets hot). Transistor part number for similar systems might be METR3904WG-SC70-3, although that seems like a Dell part number for different laptop, this seems to be a somewhat generic part.

    I would guess QE5 is relating to temp around the CPU based on location. FYI the three tall silver chokes R22 that are clustered together are for the adjacent CPU (the three chokes lined up nicely are for the nearby GPU). You might find other similar components like QE1, QE2 etc. which might be related to temp around fan, DRAM, skin of laptop, etc.

    Nearby is H56 temp sensor which could be related to QE5. I can't see that in your photo

    EDIT - I don't have the schematic so all this is just speculation.

    Your photo is a bit fuzzy so a few questions:

    - Does QE5 look like it is burnt or broken or with a hole in it?
    - What do the QE5 soldier joints look like? Green, crusty, cracked, black.
    - Is the motherboard around QE5 clean or does it look dirty, burnt, coloured...
    - Look carefully around at other components, their soldier joints, the motherboard. Particularly stuff nearby. Everything tidy and clean?
    - Can you get a high quality clear macro picture of the QE5?

    Without being able to see anything, possible issues are:

    - Failed QE5 or upstream/downstream component
    - Water damage to soldier joints
    - Epoxy damage to soldier joints (from adhering the alu heatsinks to the VRM mosfets)
    - Broken soldier joints (from accidently hitting the QE5 or heat cycling)

    I think it would be best to try to troubleshoot and fix the hardware issue before using laptop anymore.
     
    Splitframe, Bousqi and Schmoo2k like this.
  3. Splitframe

    Splitframe Notebook Guru

    Reputations:
    7
    Messages:
    56
    Likes Received:
    20
    Trophy Points:
    16
    Thank you for your indept answer, it is much appreciated.
    When I first noticed it was loose I took a good look at it.
    It seems the solder pad was taken off.
    I do distinctly remember that I didn't touch any components when installing the heatsinks.
    If I did it would be covered in thermal glue I guess. I believe that the component itself is still intact.
    It still hangs on, just one leg is not connected.
    There is nothing else around it. Nothing is burned, nothing is blackened, it doesn't even smell weird.
    It's just that one disconnected leg.
    Since it was soldered with leaded solder I could imagine that the solder point weakened.
    I do heavily use it with temps going really high, hence the heatsink mod.
     
    pressing likes this.
  4. pressing

    pressing Notebook Deity

    Reputations:
    257
    Messages:
    1,760
    Likes Received:
    857
    Trophy Points:
    131
    Take a good macro photo then to confirm if this is either broken soldier joint or lifted pad. Things are so small it is tough to tell with the naked eye.

    Dell would have used non-leaded soldier for ROHS compliance.

    If this is broken soldier joint, you can fix with a bit of liquid flux, soldier, and something to clean the flux as needed. If this is a lifted pad, there are a few options, including running coated wire either point to point. Louis Rossman has some good YouTube videos repairing these types of damages so will have best practices. I have limited experience fixing lifted soldier pads at microscopic level so can't help more. You might just bring this to the local repair laptop repair shop and ask them to rebond although that is a bit expensive for an old laptop.

    Here are a few pages on testing transistors. For sure computer needs to be off. But I don't know if you will get correct readings if the transistor is tested in a circuit. Once again, I don't recall seeing this QE5 failing in any posts here. I might (or might not) test for fun.

    http://learnabout-electronics.org/transistor_faults_04.php
     
    Splitframe, Bousqi and Schmoo2k like this.
  5. gnbspop

    gnbspop Notebook Guru

    Reputations:
    0
    Messages:
    53
    Likes Received:
    2
    Trophy Points:
    16
    Can your 9550 run Shekiro Shadows Dies Twice? Can You Run It Said no.
     
  6. pressing

    pressing Notebook Deity

    Reputations:
    257
    Messages:
    1,760
    Likes Received:
    857
    Trophy Points:
    131
    As prices of QLC SSDs have become so cheap, some might revisit boosting storage on the 9550. After a terrible launch, these laptops have proven to be quite robust.

    Parts to add a 2.5" drive to the 9550 are in the link below (with the smaller battery required). This will run SATA speeds (presumably SATA III @ 6Gb/s but you never know). The m.2 is 4x lanes.

    https://www.dell.com/community/Lapt...ions-And-Upgrade-Requirements-For/m-p/5072834
     
  7. pressing

    pressing Notebook Deity

    Reputations:
    257
    Messages:
    1,760
    Likes Received:
    857
    Trophy Points:
    131
    * Warning on recent wifi bluetooth Drivers on the Dell site.

    Oddly Dell had 5(!) wifi/bluetooth drivers on its site for my laptop service tag (1820 & 1830). This weekend, I downloaded and installed the following Wifi and bluetooth drivers, respectively:

    Communications_Driver_R3D5X_WN32_1.442.0.0_A00.EXE
    Communications_Application_8R3ND_WN32_12.0.1.720_A00.EXE

    I could not get the 1830 wifi card to work after clean install of new SSD. The bluetooth was a bit hit or miss but seemed to work as well as it did previously. The wifi was dead as a doornail. I reinstalled the drivers, removed the wifi card to inspect all the connections, reinstalled windows...no luck. I found a patch on Dell's site by pure luck.

    It seems Dell has a defective wifi driver currently on the downloads site. It seems there is a patch "Dell Wireless 1820A 1830 WiFi Driver 1.566.0.2, A07 (fixes - Corrected WPA2 protocol vulnerabilities) which requires some manual effort to install the inf file.

    Immediately after extracting the file and following Dell's irritating instructions, wifi worked perfectly. Just got lucky looking there:

    "Installation
    1. Browse to the location where the Self-Extractor placed the driver files (C:\DELL\DRIVERS\xxxxx).
    2. Locate the bcmwdidhdpcie.INF and right click on this file.
    3. Choose Install from the contextual menu, and wait for the driver to install. There will be a popup message to confirm driver installed successfully."

    https://www.dell.com/support/home/us/en/04/drivers/driversdetails?driverid=d63k3
    DW1820A_1830_WiFi_Win10_x64_1.566.0.2_SETUP_ZPE.exe

    Not sure if the other two drivers (1820) would work correctly,
     
    Splitframe and maffle like this.
  8. pressing

    pressing Notebook Deity

    Reputations:
    257
    Messages:
    1,760
    Likes Received:
    857
    Trophy Points:
    131
    * I picked up a new 1TB nvme drive (xpg sx8200 pro). For $100 it runs great and CrystalDiskMark results are in line with maker claims and published "reviews".

    * I will install the 1TB HDD for media etc. later this week once laptop is set up.

    * Last year, I installed 3M Super33 electrical tape between the fans and the radiators to improve airflow but some of that tape on the plastic fan housing was a bit loose. So I just removed the offending tape and installed more tape in that area for better adhesion. Also blew some of the dust out of the fans, but they were pretty clean.

    * Also, the FHD screen has some artifacts at the top left corner running towards the bottom right corner. The artifacts are somewhat rectangular but the bottom right side looks like ripped paper. Rectangle is about 5cm x 10cm. Rectangle is not visible straight on, but at an angle we can see the rectangle perimeter, particularly with a white screen.

    To remove the artifacts, I tried removing both the screen and backlight connectors, cleaning with electronics cleaner spray, blowing dry with canned air. Then I gently massaged and straightened the cables, and reconnected. The connections are hopelessly small so maybe there are a few bent pins I can't see. After repair effort, no difference, unfortunately. So this might be physical damage to the panel, although the laptop is in one room virtually all the time. Doesn't bother me much to be frank.
     
    Chang_loo, Splitframe and maffle like this.
  9. pressing

    pressing Notebook Deity

    Reputations:
    257
    Messages:
    1,760
    Likes Received:
    857
    Trophy Points:
    131
    Good video below of UK tech reviewing a dead 9550 with microscope, thermal camera, and DMM.

    Screen test works. Occasional fan spin but will not boot up. Tech noted one of the mosfets to the CPU was shorted (AO ZS019QI TA5S23) which you see in the picture below. My 9550 mosfets have slightly different part numbers FYI.

    In comments, he noted that he replaced mosfet but no change, so concluded failed CPU.

    Not sure how this died. Tech wrote, "In our case it wasn't a bios issue. It's possible that a corrupt firmware can cause this problem."

    I suppose Maybe mosfet failed (overworked and sometimes run very hot) and killed cpu but I don't think the CPU just died and killed the upstream mosfet. I remember people here measuring 120*C+ temps in that area so if the laptop was not measuring and/or regulating temps in that VRM region, the lives of the mosfets would be shortened. Disabling Intel DPTF or shorting out the ambient temp sensor would cause this.

    Potential mitigation is to reduce temps in the laptop (repaste CPU/GPU, run Dell Command-Power Manager on cool, etc.). Maybe BIOS update problem as we know how sloppy dell is.

     
    maffle likes this.
  10. maffle

    maffle Notebook Evangelist

    Reputations:
    100
    Messages:
    423
    Likes Received:
    351
    Trophy Points:
    76
    @pressing is this a 9550 only issue, or can it happen on a 9560 and 9570 too?
     
    pressing likes this.
Loading...

Share This Page