Page 1 of 2 12 LastLast
Results 1 to 10 of 11

Thread: Help fixing Travla c134 Power Supply

  1. #1
    Raw Wave rando's Avatar
    Join Date
    Mar 2004
    Location
    Redondo Beach, CA
    Posts
    1,973

    Need Help fixing Travla c134 Power Supply

    I just picked up a used c134 with a dead supply. I thought I'd take a shot at fixing it until my Opus gets here (if it ever gets here ).

    Can anyone with one of these supplies post the numbers off the part labeled Q8 (indicated by the yellow arrow in this picture)?
    Attached Images Attached Images  

  2. #2
    Raw Wave rando's Avatar
    Join Date
    Mar 2004
    Location
    Redondo Beach, CA
    Posts
    1,973
    I found a close up pic (by accident) by following a link on a different thread here.

    http://www.carnetix.com/CasetronicsPSU/MVC774X.JPG

    It's an Si4800. The datasheet can be found here.

    http://www.semiconductors.philips.co.../SI4800-02.pdf

  3. #3
    FLAC
    Join Date
    Nov 2003
    Location
    Marietta, GA
    Posts
    1,134
    Rando,

    What failure symptoms is your PSU exhibiting? Do you get anything out of it, or is the whole thing just dead? From what I remember, the SI4800 is used to inhibit the +5V output. It's controlled by the turn-on logic on the input side of the board. There's also a fuse at the input, which could prevent anything on the board from powering up.
    MikeH

  4. #4
    Raw Wave rando's Avatar
    Join Date
    Mar 2004
    Location
    Redondo Beach, CA
    Posts
    1,973
    The supply is 'always on' and is non-responsive to the PWON signal. The board is generating the required voltages (+12, +5, +3.3) up stream. On the output connector though, each of the 5v outputs is getting 2.86v. The other output voltages, including 5vsb, are correct. The Si4800 is visibly damaged (overheated). There are no other visibly defective components. Also, in probing around the board, the 2.86v signal seems to appear at the Si4800.

    I'm guessing there is some hope that the supply might be salvaged. I've already ordered a replacement si4800 (actually a substitute part made by Vishay instead of Phillips, SI4800DY) from Mouser Electronics. It's only $1.44, so no big deal if it just smokes again.

    The Si4800 appears to be a high current switch (TrenchMOS FET) with exceptionally fast switching time. Given that and the failure mode, I too believe it is related to the 5v turn-on/off of the supply.

    Any ideas or insight you can provide is welcome. I'm happy to provide data and/or run any test scenarios that might help.

    Thanks for responding.

  5. #5
    FLAC
    Join Date
    Nov 2003
    Location
    Marietta, GA
    Posts
    1,134
    Rando,

    It's good that your +5Vsb is working. It's fed by the same switching regulator that feeds the +5V output. If it's working, chances are pretty good that replacing the SI4800 will work, at least for +5V outputs.

    The board is generating the required voltages (+12, +5, +3.3) up stream.
    What do you mean by this (up stream)? Where are you able to measure +5V?
    The supply is 'always on' and is non-responsive to the PWON signal
    Are you referring to the PS_ON# input from the motherboard to the PSU connector (not the PWRON from the front panel pushbutton switch to the motherboard)? Remember that this is a TTL low input signal to turn the PSU on. If grounding this pin has no effect on the PSU, there could be a blown TPS3510 (power supply logic controller) since it uses this signal to properly enable the PSU outputs when all voltages are at their proper operating point.

    That's all the guesses I have at this point. When you get the SI4800 soldered in (be careful...steady hands) let me know what you find.

    Good Luck!
    MikeH

  6. #6
    Raw Wave rando's Avatar
    Join Date
    Mar 2004
    Location
    Redondo Beach, CA
    Posts
    1,973
    Sorry, by 'up stream' I meant closer to the 12v input side of the board. +5v can be measured at pin 3 of Q9 (see orange arrow) and pin 2 of Q11.

    I mean the pin labeled PON on the 20 pin ATX connector. If PN is briefly (or continuously) shorted to ground, no voltages change in the circuit.

    Here are the voltages at the ATX connector.

    Label Measured
    ----- ---------
    -12 -11.89
    G 0
    +3.3 3.32
    +5 2.63
    5vsb 5.06
    +12 12.17
    PON 2.68
    PG 0.02

    I don't know about the TPS3510. I'll look it up and see if I can tell if it's doing it's job or not.

    I'll post a follow-up when I get the Si4800 installed.

    Thanks again for the help.
    Attached Images Attached Images  

  7. #7
    Raw Wave rando's Avatar
    Join Date
    Mar 2004
    Location
    Redondo Beach, CA
    Posts
    1,973
    The TPS3510 appears to be doing the right thing, based on its inputs, according to the function table at http://focus.ti.com/lit/ds/symlink/tps3510.pdf.

    It is seeing both and undervolt condition on +5v (2.86v) and PSONBAR is High. PGI and the other voltages appear to be within the expected operatin ranges. It is outputting PGO (Low), and FPOBAR (High).

  8. #8
    Raw Wave rando's Avatar
    Join Date
    Mar 2004
    Location
    Redondo Beach, CA
    Posts
    1,973
    OK, I put in the new Si4800 and the +5v outputs are good now. The entire supply is still stuck in the ON mode though. Shorting PON to Ground does correctly causes PG to go HI.

    From what I can tell, the FPOBAR output on the TP3510 is functioning, almost. It's unclear if the TP3510 is the problem or not though.

    When PON is brought low, FPOBAR goes low as expected. When PON floats, FPOBAR only rises to ~1.56V despite the fact that it is pulled high (+5v) through a 10k resistor. I cannot be sure (without disconnecting the FPOBAR pin) if the current is flowing through TPS3510 (because it's defective) or if the problem is with the transistor network that ultimately switches the +3.3,+5,+12 rails on/off.

    If I remove the TPS3510 and the power supply shuts off, then I belive that points to a defective TPS3510. If the supply stays on, then the problem is in the power on/off transistor switching network.

    Does this make sense?

  9. #9
    Raw Wave rando's Avatar
    Join Date
    Mar 2004
    Location
    Redondo Beach, CA
    Posts
    1,973
    OK, I removed the TPS3510 and was able to get the supply to manually shut-off / turn-on by setting th pin 3 location (where the TPS3510 used to be) to +5vsb / Gnd. I've already ordered the replacement chip (free sample from TI) and this bad boy will be fixed. I feel like I've won the lottery. (ok, not really).

  10. #10
    Raw Wave rando's Avatar
    Join Date
    Mar 2004
    Location
    Redondo Beach, CA
    Posts
    1,973
    Just for kicks, I went ahead a put the old TPS3510 back in (really I had only removed the FPOBAR pin). Oddly, the supply works perfectly now. I'm typing this from my m10000 powered by the supply right now. And I thought only us digital guys had to worry about the bizarre!

    When the new chip gets here, I'll go ahead an replace it -- just in case.

    UPDATE: New chip in and it still works.

Page 1 of 2 12 LastLast

Similar Threads

  1. power supply buzzing help thing i found... read if u have noise
    By TimW in forum General Hardware Discussion
    Replies: 3
    Last Post: 02-25-2012, 12:48 PM
  2. My low cost, easy to build power supply...
    By SkinnyBoy in forum Power Supplies
    Replies: 52
    Last Post: 09-27-2006, 07:10 PM
  3. Replies: 7
    Last Post: 06-15-2004, 10:29 PM
  4. Opus power supply issue
    By dmazyn in forum Power Supplies
    Replies: 2
    Last Post: 10-15-2003, 07:32 PM
  5. HELP!! Power Supply wont supply power
    By Saab in forum General Hardware Discussion
    Replies: 3
    Last Post: 08-28-2000, 10:33 AM

Bookmarks

Posting Permissions

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