Connect with us

Amplifier blows power fuse

Discussion in 'Troubleshooting and Repair' started by dataworx, Feb 19, 2016.

Scroll to continue with content
  1. dataworx

    dataworx

    17
    0
    Feb 19, 2016
    I posted an intro, but in case you missed it, I'm a noob at electronics.

    I opened this amplifier (Sansui HT 3000 home theatre job) to check whether it could be repaired. The mains fuse was blown, as per photo below, and knowing that the mains supply here is rather erratic, I suspected that a surge may have caused this. Since the fuse appears to be soldered directly onto the PC board, and not having a suitable replacement immediately available, I soldered a 0.2mm strand of copper wire across the fuse and applied power.

    IMG_0567.JPG

    The temporary fuse disappeared with a most satisfying pop so I started looking around for the cause. All the components look OK, but I'm not sure if the brownish substance around the capacitors (as per photo) could be the villain. It looks as though it has been purposely applied, but lack the experience and knowledge to test this theory.

    IMG_0573.JPG

    Is it worth trying to repair this amp, and if so could someone point me in the direction I should be looking?
     

    Attached Files:

  2. Harald Kapp

    Harald Kapp Moderator Moderator

    10,360
    2,258
    Nov 17, 2011
    The brownish substance could be some kind of glue to fix the capacitors in place as a protection against mechanical shock. However, I think this is rather unlikely, given the erratic structure of the goo.
    Chances are that these are remains from the electrolyte that leaked from one or more of the electrolytic capacitors (2200µF). In that case the respective capacitor will definitely have lost capacity and may have created a short circuit.
    I suggest you de-solder these capacitors and check them for continuity with an Ohmmeter (there should be none). If you have access to a capacitance meter, check the capacity, too.
    Then clean the board using alcohol or demineralized water, dry completely and replace the old capacitors by new ones of the same rating.

    While getting the new capacitors, also get some fuses. You may want to replace the leaded fuse (which may not be so easy to come by) by a suitable fuse holder and loose fuses.

    Never substitute anything else for a true fuse, it may cost you dearly!.

    Check the board also for other visible defects or suspect components. Insert the fuse and before you apply power, check the resistance of the power supply as seen from the input. If the ohmmeter still shows a short circuit, the issue is still present and you'll have to inspect the board further.
     
    Arouse1973 and Martaine2005 like this.
  3. dataworx

    dataworx

    17
    0
    Feb 19, 2016
    Thanks very much for your help Harald.

    From the pattern of the electrolyte on the PC board, (see pic below) it looks like a few, maybe all 6 of the capacitors have leaked. Is that a likely scenario?

    The only test instrument I have is an elderly Fluke 73 multimeter. Can I test the continuity of the capacitors satisfactorily with that?

    IMG_0577.JPG
     
  4. Martaine2005

    Martaine2005

    2,931
    799
    May 12, 2015
    That looks a lot neater from a top view. Looks more like a deliberate application of glue for the larger caps and ferrite chokes, rather than leakage. As Harald suggested against vibration etc.

    Martin
     
    HellasTechn likes this.
  5. dataworx

    dataworx

    17
    0
    Feb 19, 2016
    Yeah, the more I look at it, the more it seems to be deliberate. There is also some similar looking
    goop on the push in power connectors.

    So if the capacitors are OK, where else should I be looking?
     
  6. dorke

    dorke

    2,342
    665
    Jun 20, 2015
    Check the final power transistors in diode test mode.
    There is a good chance some are shorted out.
     
    HellasTechn likes this.
  7. cjdelphi

    cjdelphi

    1,096
    104
    Oct 26, 2011
    2nd from top of photo, the transistor looks cracked from heat stress ...

    I'll bet you'll read a short between 2 of those pins and maybe more
     
  8. cjdelphi

    cjdelphi

    1,096
    104
    Oct 26, 2011
    Infact is it a power transistor? 5 pins
     
  9. dataworx

    dataworx

    17
    0
    Feb 19, 2016
    Thanks dorke

    Are you referring to the components that are fixed to the heat sink - there are two, one is circled below:

    IMG_0580.JPG

    Can these be tested on the board or must I remove them?
     
  10. Harald Kapp

    Harald Kapp Moderator Moderator

    10,360
    2,258
    Nov 17, 2011
    I agree with my peers about the glue, the new photos show it better.
    You'll have to remove them, otherwise other components in the circuit will distort your measurements.
     
  11. dorke

    dorke

    2,342
    665
    Jun 20, 2015
    If it is a home theater amp it should have 5 channels or more(how many are there?)
    so,
    If there are only 2 then these aren't them.
    Could you post a picture showing "all the amp parts".

    What is written on the 5 pin black devices on the heatsink in post#3 ? they could be the PA ICs.
     
  12. HellasTechn

    HellasTechn

    1,537
    213
    Apr 14, 2013
    I concur !

    Also check for voltage regulators nearby.

    And yes the brown substance is more likely glue !
     
  13. Arouse1973

    Arouse1973 Adam

    5,164
    1,081
    Dec 18, 2013
    Hey Martin good to see you back.
    Adam
     
    Martaine2005 likes this.
  14. Martaine2005

    Martaine2005

    2,931
    799
    May 12, 2015
    Thanks Adam.
    My tablets only give me half hour of sanity.. I sound like a walking pharmacy!!
    But definitely on the mend..

    Martin
     
  15. dataworx

    dataworx

    17
    0
    Feb 19, 2016
    Yes, this is a 5 channel amp. The 5 pin devices in post #3 are marked D1875, there are 10 of them on the heat sink. Post #9 shows the heatsink that has 2 other devices (only 3 legs) attached to it - I took these to be power transistors. Here are some more pics to help clarify:

    IMG_0584.JPG
    IMG_0585.JPG
    IMG_0586.JPG
    IMG_0587.JPG
     
  16. 73's de Edd

    73's de Edd

    2,922
    1,235
    Aug 21, 2015
    dataworx . . . . .

    Your referenced "transistors" are actually being a stringed out series of ten, 5 pinned, LM1875 . . . . . 20-30 watt power amplifier I.C.'s.
    I would initially be looking for failure fault at the 4 diodes just past the fuse involved . . .D40 and its 3 companions .

    Then there is the switch mode power supply for the unit, with that UM1953 being the main power transformer, and the smaller UM 1203 beside it probably being the units standby supply transformer .
    Sooooo the larger frontal heatsink should have one or possibly two power FETS . . . . or . . . . . a custom Power IC mounted on it.
    Pass us its/their part numbers .
    The smaller rear heat sink should be cooling up to four large case dual schottky rectifiers mounted to it . . . pass on their numbers to us also.
    After the four diodes earlier mentioned, the next most likely suspect to be having blasted parts is those on the larger heat sink.

    Yes the reddish brown "pookie" is a retentive adhesive applied by the manufacturing line.

    Also needed is the controller IC near the UM1203, I see its Texas Instruments logo, but confirm the TL493 or 4 or 5 series number .

    Thassit . . . . .

    73's de Edd


    Aside:
    Sir Martin . . . . . .Gomer sez hey ! ! !
     
    Last edited by a moderator: Feb 20, 2016
  17. dataworx

    dataworx

    17
    0
    Feb 19, 2016
    73's de Edd

    Many, many thanks for your assistance. I need to get the PC board out of the chassis so I can get the details you need.

    First suspect- the diodes. Can I loosen one leg of each from the board and test them conclusively with a DMM?
     
  18. 73's de Edd

    73's de Edd

    2,922
    1,235
    Aug 21, 2015
    Yes

    73's de Edd
     
  19. dorke

    dorke

    2,342
    665
    Jun 20, 2015
    O.K so we have 5 channels and 10 LM1875 devices.
    That means each channel is most probably made of
    2 bridged LM1875.
    I would check in ohm test(not diode test!)
    Pins 3(-VEE),4(output),5(+VCC) to GND for high resistance.
    and between pin 4 and 3 , and pin 4 and 5 for high resistance.
    Then check all output terminals to GND for high resistance.
    LM1875-pins.jpg
     
    Last edited: Feb 20, 2016
  20. 73's de Edd

    73's de Edd

    2,922
    1,235
    Aug 21, 2015
    dataworx . . . . . . .


    Look at your very first photo in post 15 . . . . . .

    I would initially take a look at those glass clip in fuses at the end of the differential power supply capacitor banks of 3 +3 capacitors and , if they are not being blown,
    I would then primarily be concerned with the Switch Mode Power supply area, if your initial 4 FWB diodes were undamaged.

    Timely Technical Tip . . . . . . . # 11,203

    Relevant to:

    "I need to get the PC board out of the chassis so I can get"

    If you have not tested those diodes out of circuit . . . . . or even initially in circuit . . . . as a dead short on one of them would be glaringly evident .

    To test diodes from the boards topside . . . . . . .
    Rosin flux up the most soldering iron accessible lead of each separate diode, move to the first diode involved and use the tips of some needle nose pliers at the diodes epoxy case junction / to / adjunct diode wire lead to grip the diode to carry off sustained generated heat from the diodes lead .
    Apply enough solder down at the diode lead / PCB junction to form a BB size blob and hold the iron in place while pulling up on the needle nose, then that very efficient transfer of heat should have the original lower foil to diode connection also melting and the lead will then pull up and be out of circuit for its final out of circuit testing.

    Thasssit . . . . .


    73's de Edd
     
    Last edited by a moderator: Feb 20, 2016
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-