Announcement

Collapse
No announcement yet.

GTS 250 GPU short circuit

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

    GTS 250 GPU short circuit

    I have a GTS 250 with short that goes all the way to the power connector and shuts down PSU for safety. Fan spins for a while then shuts down. I tracked down one of the mosfets is shorted. Pulled it out and now the short seems to disappear. At least I am getting 12V at the power connector and the PC starts again.

    I also noticed the SMD fuses near the PCI-E slot are blown.

    So 2 questions:
    1) Should I expect any other damaged component since the PSU and the computer start again?

    2) Can someone tell me what are these two fuses details circled in the picture?
    Attached Files

    #2
    Re: GTS 250 GPU short circuit

    Since the shorted MOSFET was a high-side MOSFET, the GPU can be dead (sending 12V to GPU VCore). You can also check the resistance between the GPU VCore inductors and ground to know if there's a shorted low-side MOSFET or shorted GPU. It is expected to be low, but probably not lower than 2ohms.
    Also when replacing a MOSFET it's advised to replace the PWM controller as well, and match the batch numbers of all the high side MOSFETs (or low-side).
    The fuses are between the 12V power rail coming from the PCIe port and the input of a buck converter. Not sure what they are though, but at least several amps. (PCIe port can give 75W on the 12V rail)
    OpenBoardView — https://github.com/OpenBoardView/OpenBoardView

    Comment


      #3
      Re: GTS 250 GPU short circuit

      Originally posted by piernov View Post
      Since the shorted MOSFET was a high-side MOSFET, the GPU can be dead (sending 12V to GPU VCore). You can also check the resistance between the GPU VCore inductors and ground to know if there's a shorted low-side MOSFET or shorted GPU. It is expected to be low, but probably not lower than 2ohms.
      Also when replacing a MOSFET it's advised to replace the PWM controller as well, and match the batch numbers of all the high side MOSFETs (or low-side).
      The fuses are between the 12V power rail coming from the PCIe port and the input of a buck converter. Not sure what they are though, but at least several amps. (PCIe port can give 75W on the 12V rail)
      Thanks for having a look. In the photo I uploaded I have circled in red the MOSFET I removed and tested the GPU to see if the PC will open. The one in green circle I just pointed out so not to bother. It's the original MOSFET I thought it was shorted but then I badly soldered it since I am going to replace all of them either way. I guess this is the Vcore section or high-side as you mention?

      The fuses I found are 5-6A pretty much makes sense what you said about 75W delivery.

      Now.. what I would appreciate from you to do is draw on my pic where the PWM is located in order to change that too.. and where I need to measure the resistance between the GPU VCore inductors and ground as you said. If it's the cubes next to the MOSFETS marked as R80, resistance is 1,2 Ohm
      Does this mean I might got away with a short on the low side?

      This is my first GPU attempt to fix
      Attached Files
      Last edited by Sraros; 12-05-2018, 06:37 PM.

      Comment

      Working...