Announcement

Collapse
No announcement yet.

Macbook Pro A1989, 820-00850-A, water damage

Collapse
X
Collapse
 
  • Page of 1
  • Filter
  • Time
  • Show
Clear All
new posts
Previous template Next
    #1

    Macbook Pro A1989, 820-00850-A, water damage

    Hi all

    This laptop was my wife's, a little water dripped onto/into it while it was stowed in her backpack. Unfortunately she couldn't wait and tried turning it on right after she realized what had happened - "It didn't turn on but got hot."
    Since I'm naturally curious and I hate to throw hardware away I figured I'd have a go at and see if I can repair it.
    I've been searching for info on this particular machine but there's so much stuff I'm having a hard time putting it all together.
    The power sequence for example is still unclear to me.

    Here's where I'm at:
    - Took the laptop apart, I have the board out, nothing connected to it.
    - Cleaned all traces of corrosion, mostly around U3200, UB300 and UB400 (all CD3215)
    - U8400 is toast with visible heat damage on the chip right at pin 5 (I've removed the chip)
    - Other than U8400, there are no visibly/obviously burnt components
    - Using the original Apple Power Supply I get 5.17V PPDCIN_G3H, I get this behaviour for all 4 USB-C ports
    - When I supply power (5V) from a bench supply there's approx. 20mA current draw
    - PPBUS_G3H: 0.0V
    - It looks like PPBUS_G3H has is shorted (resistance to GND is below 1 Ohm), however, even with injecting 1V@1A (I started at only 50mA) on PPBUS_G3H nothing gets even remotely warm (using a thermal camera)

    - If I have no PPBUS_G3H, is it likely that U7000 (ISL9240) is dead too?
    - In trying to locate the short on PPBUS_G3H, how much current is safe to insert at what voltage?
    - Am I going at this the right way?

    Any help and insight is much appreciated
    Thanks for reading
    Hanns.G HG281D, 14 years, still going strong.
    Tags: None
    #2
    Re: Macbook Pro A1989, 820-00850-A, water damage

    Resistance to ground lower than 1 ohm doesn't suggest a shorted high-side MOSFET but you can still confirm by measuring resistance between PPBUS_G3H and all large inductors. If there's one that measures lower than 1 ohm then maybe there's a high side MOSFET shorted and you should not inject voltage.
    Otherwise inject voltage and find the component that heats up. No need to limit current (go to the max that the PSU can do, chances are it's not above 5A anyway), just increase gradually from 1V and check if something heats up. Always check large BGAs (CPU/GPU/PCH) first to make sure they don't heat up.
    OpenBoardView — https://github.com/OpenBoardView/OpenBoardView

    Comment

      #3
      Re: Macbook Pro A1989, 820-00850-A, water damage

      Hi piernov, thanks for your reply.
      I measured the resistances from the inductors to PPBUS_G3H. The ones with the lowest resistances were:
      L7210: 2.5 Ohm
      L7220: 2.5 Ohm
      L7270: 9.6 Ohm
      L7410: 2.3 Ohm
      L7420: 2.3 Ohm
      L7430: 2.3 Ohm
      L7702: 25 Ohm
      L8102: 43.5 Ohm
      L8100: 346 Ohm

      The others are in the tens of kOhms range and higher.
      For reference: PPBUS_G3H to GND measures at 0.3 Ohm which is basically the contact resistance of the multimeter probes. The true values of the above measurements are therefore about 0.3 Ohm lower.

      Does this mean the High-Side MOSFETs are ok? I have no idea what kind of values to expect here.

      Thank you for your help
      Hanns.G HG281D, 14 years, still going strong.

      Comment

        #4
        Re: Macbook Pro A1989, 820-00850-A, water damage

        Yes that's ok. Just inject voltage on PPBUS_G3H and find the component that gets hot.
        OpenBoardView — https://github.com/OpenBoardView/OpenBoardView

        Comment

          #5
          Re: Macbook Pro A1989, 820-00850-A, water damage

          Alright, moving on.
          At 2.5A U7600 (CSD58879) started warming up. I've removed it from the board and the short from PPBUS_G3H to GND is now gone.

          Should I check anything else before trying to source a replacement?

          As always, thank you for your help.
          Hanns.G HG281D, 14 years, still going strong.

          Comment

            #6
            Re: Macbook Pro A1989, 820-00850-A, water damage

            I'd recommend replacing at U7650 at the same time. If it's damaged, the replacement U7600 will die again.
            You can measure exact resistance to ground on L7600 to make sure there's no (partial) short to ground.
            OpenBoardView — https://github.com/OpenBoardView/OpenBoardView

            Comment

              #7
              Re: Macbook Pro A1989, 820-00850-A, water damage

              L7600 to GND is 6.06 Ohm. Is that considered a (partial) short?

              Continued thanks.
              Hanns.G HG281D, 14 years, still going strong.

              Comment

                #8
                Re: Macbook Pro A1989, 820-00850-A, water damage

                On PP5V_G3S 6 ohms is a short to ground yes, it should be in the kiloohms range, so it looks like there's an issue on that rail too.
                You should do another thorough visual inspection because it seems like there are multiple things wrong on this board related to PP5V_G3S. Failure of U8400 is related as well.
                If there's still nothing obvious then (like you did for PPBUS_G3H) check resistance between PP5V_G3S and all the large inductors making sure there's nothing below 1 ohm, and then you can inject voltage.
                OpenBoardView — https://github.com/OpenBoardView/OpenBoardView

                Comment

                  #9
                  Re: Macbook Pro A1989, 820-00850-A, water damage

                  I was able to measure the resistances between the inductors and PP5V_G3S during lunch break. The lowest ones are just above 8 Ohm:

                  L7210: 8.34 Ohm
                  L7220: 8.35 Ohm
                  L7270: 14.74 Ohm
                  L7410: 8.06 Ohm
                  L7420: 8.1 Ohm
                  L7430: 8.1 Ohm
                  L7702: 26.34 Ohm
                  L8102: 18.61 Ohm

                  The rest are even higher.

                  I'll continue tonight with the voltage injection when I have more time.

                  Thanks again!
                  Hanns.G HG281D, 14 years, still going strong.

                  Comment

                    #10
                    Re: Macbook Pro A1989, 820-00850-A, water damage

                    Sorry, this took longer than expected.
                    Here we go:

                    Visual inspection yielded no clues, I had a friend look over it too.
                    So voltage injection. I didn't have a hot-air station last night so instead of removing ICs or Transistors I removed associated resistors (which in all but one case were the components that were getting warm anyway) with a regular soldering station.

                    In the order things started warming up:

                    At 220mA and approx. 1.8V (of course I had to increase the voltage more or less after every round):
                    R7741 and R7747 warmed up -> U7710
                    After removing both resistors, resistance from PP5V_G3S to GND went up to 19 Ohm.

                    RB701 and RB706 got warm next -> UB700
                    After removing these resistors, resistance from PP5V_G3S to GND went up again to 45 Ohm.

                    R3501 and R3506 were next -> U3500
                    After removing these resistors, resistance from PP5V_G3S to GND went up again to 100 Ohm.

                    Last resistors: R8144 and R8177 -> U8110
                    After removing these, resistance from PP5V_G3S to GND went up again to 330 Ohm.

                    By now I was up to about 4.9V but only 15mA and I thought I shouldn't go above 5V as this is a 5V bus.

                    The last thing to warm up was U8500.
                    After removing that today, resistance from PP5V_G3S to GND finally went up to 18.7 kOhm.

                    I measured from the removed resistors pads (IC side) to GND and all of the measured values were in the 2-5 Ohm range. ICs U7710, UB700, U3500, U8110 are still on the board, only the resistors mentioned have been removed.

                    Is it safe to assume U7710, UB700, U3500, U8110 and of course U8500 are all fried?

                    Any other power buses that are bound to be affected?

                    More thanks.
                    Hanns.G HG281D, 14 years, still going strong.

                    Comment

                      #11
                      Re: Macbook Pro A1989, 820-00850-A, water damage

                      Removing the resistors is indeed a better idea than removing the ICs.
                      Yes all these ICs are bad. Something went *very* wrong on the 5V rail so it may not be the end of it.
                      Additionally, 2 buck controllers for CPU power rails are affected, so there's a very high chance that the CPU got damaged as well.

                      You can try to replace all of these, but do not get your hopes up…

                      Btw do not try to power on the board with any of these components removed (U7710, UB700, U3500, U8110 mainly, not U8500 this one wouldn't cause additional problems if missing), this would cause damage to the derived power rails.
                      OpenBoardView — https://github.com/OpenBoardView/OpenBoardView

                      Comment

                        #12
                        Re: Macbook Pro A1989, 820-00850-A, water damage

                        Originally posted by piernov View Post
                        Additionally, 2 buck controllers for CPU power rails are affected, so there's a very high chance that the CPU got damaged as well.
                        Would there be any way to test for this at this point?

                        Originally posted by piernov View Post
                        You can try to replace all of these, but do not get your hopes up…
                        Yeah, I kind of figured that. But since there is no pressure to have it fixed, I'm going to see how far I can take it without too much of an investment in hard to find components. There's also the learning experience.

                        Originally posted by piernov View Post
                        Btw do not try to power on the board with any of these components removed (U7710, UB700, U3500, U8110 mainly, not U8500 this one wouldn't cause additional problems if missing), this would cause damage to the derived power rails.
                        Gotcha.

                        I'll report back when I have the replacement parts installed, probably will be in a few weeks seeing as the sources will be mainly aliexpress.

                        Thanks for your help so far.
                        Hanns.G HG281D, 14 years, still going strong.

                        Comment

                          #13
                          Re: Macbook Pro A1989, 820-00850-A, water damage

                          Originally posted by usbdevice View Post
                          Would there be any way to test for this at this point?
                          Not really, but I assume you get around 2 ohms to ground on CPU VCore which already doesn't sound very good. It's typically higher 4 ohms on this I think.
                          But hard to tell if it's a problem or not, next step is replacing the dead chips and see if the machine turns on and hopefully finishes POST…
                          OpenBoardView — https://github.com/OpenBoardView/OpenBoardView

                          Comment

                            #14
                            Re: Macbook Pro A1989, 820-00850-A, water damage

                            Finally, the last order arrived today and I can get back to work on this.
                            I'll let you know how it goes.
                            Hanns.G HG281D, 14 years, still going strong.

                            Comment

                            Previous template Next
                            Working...
                            X