Announcement

Collapse
No announcement yet.

Nvidia GTX 1080 (gainward / palit) inductor

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    Nvidia GTX 1080 (gainward / palit) inductor

    Hi,

    Does anyone know what spec the inductor named "Y" is near the 8 pin connector?

    It has failed (although don't know if it was because of something else).

    I notice on other 1080 PCBs where it has L19 printed on the PCB they have larger inductors with either LR33 or 1R0 written on them so maybe that is clue enough.

    Cheers

    Ed
    Attached Files

    #2
    Re: Nvidia GTX 1080 (gainward / palit) inductor

    That looks like it is a fuse. Y=15 amp Q=10 amp. Some mfgr's might install an inductor but those look to be fuses.
    Attached Files
    Last edited by R_J; 05-10-2018, 08:01 PM.

    Comment


      #3
      Re: Nvidia GTX 1080 (gainward / palit) inductor

      The footprint is for an inductor but the part indeed actually looks like a fuse.
      The fuse should not be open, if it is there might be another issue, for example a short to ground after it so check that before replacing the fuse.
      OpenBoardView — https://github.com/OpenBoardView/OpenBoardView

      Comment


        #4
        Re: Nvidia GTX 1080 (gainward / palit) inductor

        Thanks for the info guys. The Q fuse is fine but the Y reads an extremely high resistance.

        Will do some further investigation.

        Comment


          #5
          Re: Nvidia GTX 1080 (gainward / palit) inductor

          There is a short to ground everywhere. What's the most likely culprit? The main GPU core VRMs?

          Comment


            #6
            Re: Nvidia GTX 1080 (gainward / palit) inductor

            GPU will show a very low resistance (2 to 20 ohms) so that's not a short. What is the exact resistance you measure, and where?

            The VRM MOSFET can die, and they will take out the fuse. However, if a high-side MOSFET is shorted, chances are the GPU is dead too (it received 12V directly).
            OpenBoardView — https://github.com/OpenBoardView/OpenBoardView

            Comment


              #7
              Re: Nvidia GTX 1080 (gainward / palit) inductor

              1 ohm, the same as touching the probes together.

              Is there anyway of checking the vrms without removing them from the board? So far I have checked the following to ground:-
              1 to 7
              27 to 31

              They are all comparable with each other. These are apparently quite high specced vrms.

              Comment


                #8
                Re: Nvidia GTX 1080 (gainward / palit) inductor

                I meant to add - The datasheet is here - https://cdn.badcaps-static.com/pdfs/...084433d057.pdf

                This shows the pin outs that I'm referring to.

                Comment


                  #9
                  Re: Nvidia GTX 1080 (gainward / palit) inductor

                  Has anyone got any ideas on what to do next? Was going to try lift 1 leg on each choke to see whether the GPU is OK or not but couldn't get enough heat into it.

                  Comment


                    #10
                    Re: Nvidia GTX 1080 (gainward / palit) inductor

                    If you have a bench power supply... What you could try is suppling a couple volts to where the 12 volts enters the board (but on the other side of the fuse) and see what gets hot. If you don't have a bench supply try using a 9 volt battery, connect the ground to the battery - but just touch the + side to the board (fuse) for a second, don't connect it for a long time. 9 volts is less than 12 volts so it won't harm anything on that line but should supply enough current to heat up the bad component. It could be as simple as a shorted chip cap.
                    You can lightly spray something like isopopyl alcohol on the board and it should evaporate real fast where the shorted component is. or use your finger, Yikes... #^%%% will indicate the bad part. Or use a flir thermal imaging camera
                    Last edited by R_J; 05-15-2018, 03:26 PM.

                    Comment


                      #11
                      Re: Nvidia GTX 1080 (gainward / palit) inductor

                      Originally posted by R_J View Post
                      If you have a bench power supply... What you could try is suppling a couple volts to where the 12 volts enters the board (but on the other side of the fuse) and see what gets hot. If you don't have a bench supply try using a 9 volt battery, connect the ground to the battery - but just touch the + side to the board (fuse) for a second, don't connect it for a long time. 9 volts is less than 12 volts so it won't harm anything on that line but should supply enough current to heat up the bad component. It could be as simple as a shorted chip cap.
                      You can lightly spray something like isopopyl alcohol on the board and it should evaporate real fast where the shorted component is. or use your finger, Yikes... #^%%% will indicate the bad part. Or use a flir thermal imaging camera
                      Or use a sliver of bar soap as a heat indicator.

                      Comment


                        #12
                        Re: Nvidia GTX 1080 (gainward / palit) inductor

                        Yes, lifting the inductors is the good way to go for now. You don't want to damage the GPU if it's not dead already, don't randomly send 9V into the board.
                        This part of the board will dissipate a lot of heat and the inductors themselves too, so don't try to force it. If you can't get them to come off without overheating the board, you will have to desolder the MOSFETs (and it can be pretty hard to do so too).
                        Last edited by piernov; 05-16-2018, 11:41 AM.
                        OpenBoardView — https://github.com/OpenBoardView/OpenBoardView

                        Comment


                          #13
                          Re: Nvidia GTX 1080 (gainward / palit) inductor

                          Nobody said to "randomly send 9V into the board." What I said was to apply it to where the +12 volts (see 12 volts) would normaly be. How the hell is 9 volts going to harm anything when the normal voltage is 12volts? Use a 1.5 volt battery or supply. even that will show where the short is.
                          Trying to remove the inductors without proper desoldering equipment is going to result in ripped traces or destroyed components.
                          Attached Files
                          Last edited by R_J; 05-16-2018, 12:01 PM.

                          Comment


                            #14
                            Re: Nvidia GTX 1080 (gainward / palit) inductor

                            Originally posted by edless View Post
                            Has anyone got any ideas on what to do next?
                            Remove opened "Y" fuse and solder a thick wire across its spot (or solder the wire directly across the fuse, whichever you prefer). Then use your multimeter and check for short-circuit between the +12V rail and ground pins on the 6/8-pin PCI-E power connectors. If you get a short-circuit, you have a shorted upper MOSFET somewhere, and that could** mean dead RAM or GPU chip.

                            ** on older GPUs and also low-power GPUs, a shorted upper MOSFET doesn't always kill them. I find the survival rate of those about 50-70%. On mid- and high-end chips, the survival rate drops a lot. But your GPU/RAM may still have survived.

                            What to do:
                            - If you have a shorted upper MOSFET, find which one it is (or which set of MOSFETs) - this you will have to do by visual tracing on the board and some help from your multimeter.
                            - If you don't have a shorted upper MOSFET, you can try to power On the video card, provided you have a lower power PSU that will quickly shut down if it detects a short-circuit (I recommend using a high-quality 400-500 PSU at the most for such tests). And of course, if you do this test, be prepared to turn off power to the computer from the PSU, in case it doesn't trip. This is merely to prevent the 6/8 pin PCI-E wires from melting. Otherwise, a faulty card with okay upper MOSFETs but shorted lower MOSFETs will usually just draw excessive current and trip the PSU without doing damage to the GPU chip or RAM.

                            Originally posted by R_J View Post
                            If you have a bench power supply... What you could try is suppling a couple volts to where the 12 volts enters the board (but on the other side of the fuse) and see what gets hot. If you don't have a bench supply try using a 9 volt battery, connect the ground to the battery - but just touch the + side to the board (fuse) for a second, don't connect it for a long time. 9 volts is less than 12 volts so it won't harm anything on that line but should supply enough current to heat up the bad component. It could be as simple as a shorted chip cap.
                            You can lightly spray something like isopopyl alcohol on the board and it should evaporate real fast where the shorted component is. or use your finger, Yikes... #^%%% will indicate the bad part. Or use a flir thermal imaging camera
                            I've tried this method before with a bench PSU, and I can tell you it has very very limited usefulness in scenarios like these. The main problem is that a shorted ceramic cap can become so short-circuited, that it conducts better copper traces themselves. Thus, finding a shorted ceramic cap with this method will be almost impossible. Same goes for MOSFETs, unless the MOSFET has gone partially-shorted (but that rarely happens).
                            Last edited by momaka; 05-17-2018, 01:34 AM.

                            Comment


                              #15
                              Re: Nvidia GTX 1080 (gainward / palit) inductor

                              i had one the the LR33s desolder and is damaged. I cant seem to identify which LR33 it is. does anyone have any advice on this?

                              Comment


                                #16
                                Re: Nvidia GTX 1080 (gainward / palit) inductor

                                or where i can get one

                                Comment


                                  #17
                                  Re: Nvidia GTX 1080 (gainward / palit) inductor

                                  Could we see some pictures of the damaged LR33 inductor? Did it become desoldered by itself from an overload or was that you trying to remove it with a soldering iron? If it's the former, there likely are other shorted components downstream.

                                  Comment


                                    #18
                                    Re: Nvidia GTX 1080 (gainward / palit) inductor

                                    alternative smd inductor
                                    Attached Files

                                    Comment


                                      #19
                                      Re: Nvidia GTX 1080 (gainward / palit) inductor

                                      Anyone have schematics and boardview for Palit GameRock 1080 ?
                                      My main coils are only showing 0.8v - are these meant to be 0.9v ?
                                      All other voltages are up and good, but no display.
                                      There's no short circuit anywhere.

                                      Comment


                                        #20
                                        Re: Nvidia GTX 1080 (gainward / palit) inductor

                                        Originally posted by ermak72 View Post
                                        alternative smd inductor
                                        its just inductor on 12V line. u can use barelly any coil or just run thick wire

                                        Comment

                                        Working...
                                        X