Page 8 of 11 FirstFirst ... 678910 ... LastLast
Results 71 to 80 of 106

Thread: SF750 batch failure

  1. #71
    Join Date
    May 2020
    Location
    Italy
    Posts
    21
    Thanks Thanks Given 
    13
    Thanks Thanks Received 
    0
    Thanked in
    0 Posts

    Default

    Quote Originally Posted by Jon Gerow View Post
    BOM was correct.

    I'll wait till this blows over before I reveal where the mistake was made.
    Ah hah! You know!
    Can't wait to know the truth!
    Hope you will share it.

    Btw, my SF750 it has been confirmed for delivery on the 4th of June I hope to receive a good one (finger-crossed)

  2. #72
    Join Date
    Aug 2018
    Location
    South Korea
    Posts
    393
    Thanks Thanks Given 
    171
    Thanks Thanks Received 
    40
    Thanked in
    25 Posts

    Default

    Quote Originally Posted by Jon Gerow View Post
    BOM was correct.

    I'll wait till this blows over before I reveal where the mistake was made.
    Didn't anyone(consumers) sent you some blown up ones yet?

  3. #73
    Join Date
    Jul 2015
    Location
    Vista CA
    Posts
    11
    Thanks Thanks Given 
    0
    Thanks Thanks Received 
    6
    Thanked in
    5 Posts

    Default

    Just for reference the heatsink is typically connected to the negative side of the bulk capacitor. If it were connected to earth ground the insulator would have to pass hipot voltage and there would be EMI noise injected into the ground lead.

    It looks like the cathode of the PFC diode is connected to the metal tab. I noticed in the picture upthread the power supplies were being tested in standby mode. There would be about 337 VDC across the insulator under this condition as the PFC circuit is not active in standby mode due to the Energy Star requirements. If the PFC circuit is active there is usually about 380 VDC across the insulator.

    The parts that are typically are used to mount a TO-220 package to a heatsink are an insulated shoulder washer, a silicone thermal pad, and a screw. If the heatsink in not tapped a nut and lock washer may be needed. Here is a link to an example:

    https://www.addicore.com/to-220-heatsink-p/ad285.htm

    Due to the high voltage the hole in the heatsink may need to be counterbored or countersunk. It would then have to be deburred.

    If the mounting hardware is under-torqued, the thermal resistance will by high and the diode could fail under load. The picture shows the power supplies being tested under no load, so this is not the issue.

    If the mounting hardware is way over-torqued the thermal pad would be compressed and possibly cut causing failure.

    In places I have worked in the past we sometimes checked the insulation by using a megohmmeter at 500 VDC or 1,000 VDC after mounting the part to the heatsink before it is installed into the board.

    At the first company I worked at we did not have a megohmmeter so we used a Tektronix 575 curve tracer that could put out up +- 1,500 volts to test the insulation. The advantage of using a curve tracer is one could tell if the insulator was leaky or breaking down.

  4. #74
    Join Date
    Oct 2006
    Location
    jonnyGURU forums, of course!
    Posts
    16,822
    Thanks Thanks Given 
    567
    Thanks Thanks Received 
    494
    Thanked in
    367 Posts

    Default

    Quote Originally Posted by rgbodosk View Post
    Didn't anyone(consumers) sent you some blown up ones yet?
    Yes. Of course.

    Quote Originally Posted by Tim@PSUTest View Post
    If the mounting hardware is way over-torqued the thermal pad would be compressed and possibly cut causing failure.


    VERY close. ;-)


  5. The Following User Says Thank You to Jon Gerow For This Useful Post:

    PIRATA! (05-25-2020)

  6. #75
    Join Date
    Jul 2015
    Location
    Vista CA
    Posts
    11
    Thanks Thanks Given 
    0
    Thanks Thanks Received 
    6
    Thanked in
    5 Posts

    Default

    Quote Originally Posted by Jon Gerow View Post

    VERY close. ;-)

    Could also be a cracked or over compressed shoulder washer I suppose.

  7. #76
    Join Date
    Oct 2017
    Location
    Grand Junction, CO
    Posts
    2,541
    Thanks Thanks Given 
    166
    Thanks Thanks Received 
    93
    Thanked in
    84 Posts

    Default

    Quote Originally Posted by Tim@PSUTest View Post
    Just for reference the heatsink is typically connected to the negative side of the bulk capacitor. If it were connected to earth ground the insulator would have to pass hipot voltage and there would be EMI noise injected into the ground lead.

    I still hate the idea of live heatsinks.

    Actually, I just realized that the anodized coating is an insulator, so maybe the heatsinks are not as dangerous as I at first thought.
    Last edited by ashiekh; 05-26-2020 at 02:36 PM.

  8. #77
    Join Date
    May 2020
    Location
    Italy
    Posts
    21
    Thanks Thanks Given 
    13
    Thanks Thanks Received 
    0
    Thanked in
    0 Posts

    Default

    Just received mine from Amazon Italy, and my serial number is 19534852000067110125

    I sent the email about this to james_logan@corsair.com as suggested for all owners of affected batch 1950/1951/1952/1953

    Hope to have a reply soon. I hope to not have to wait too long for receiving the advanced replacement for having my system up and running with the SF750

  9. #78
    Join Date
    Aug 2018
    Location
    South Korea
    Posts
    393
    Thanks Thanks Given 
    171
    Thanks Thanks Received 
    40
    Thanked in
    25 Posts

  10. The Following User Says Thank You to rgbodosk For This Useful Post:

    PIRATA! (4 Weeks Ago)

  11. #79
    Join Date
    Oct 2017
    Location
    Grand Junction, CO
    Posts
    2,541
    Thanks Thanks Given 
    166
    Thanks Thanks Received 
    93
    Thanked in
    84 Posts

    Default

    Mighty decent of Corsair to step up to the plate on this.

  12. #80
    Join Date
    Oct 2006
    Location
    jonnyGURU forums, of course!
    Posts
    16,822
    Thanks Thanks Given 
    567
    Thanks Thanks Received 
    494
    Thanked in
    367 Posts

    Default

    It's about time.

    Now that that's out, I would like to say that the failure was not due to PSU design or poor QC.

    Not that it matters to someone with a failed unit, but I wanted to put that out there because I realize this incident has soured a lot of people on the SF Series and people are suspect of their quality.

    The problem comes from an impurity in the insulator pads used between the PFC diode and the heatsink. When subjected to moisture (e.g. humidity), the impurties crystallize and once harden again when used in normal humidity environments, causes the insulator to crack and tear. Once the insulator cracks, the diode touches the heatsink and it pops.
    Last edited by Jon Gerow; 4 Weeks Ago at 11:20 PM.

  13. The Following 5 Users Say Thank You to Jon Gerow For This Useful Post:

    ehume (4 Weeks Ago), PIRATA! (4 Weeks Ago), rgbodosk (4 Weeks Ago), Spotty (4 Weeks Ago), vvvpcf (4 Weeks Ago)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •