Identifying and repairing SM Diode PCB

Discussion in 'Technical Repair' started by dan2511, Oct 25, 2018.

  1. dan2511

    Thread Starter New Member

    Oct 25, 2018
    4
    0
    I have a Sonos Play:3, which is past its warranty. I dismantled it to find it has the well documented fault with a blown capacitor. However, there is also damage to one of the surface mount diodes. I have sourced some new capacitors to replace those, however I am not sure what spec this diode is, and wondered whether any of you have had the same issue and if any solution was found? It appears to be D307, and has a code of AN 97 printed on it. As far as I can see, I will have to bridge the track to wherever the diode's anode was connected.

    Any advice would be greatly appreciated, as I cannot find a schematic or circuit diagram anywhere, and the board is quite badly charred!!
    Capture1.JPG
     
  2. ebp

    Well-Known Member

    Feb 8, 2018
    2,332
    811
    I think the marking on the diode is actually L6. The small characters are probably just a data code. Usually the part code is read with the [EDIT - wrong 'ode originally] cathode band to the left.

    Judging by the slots between the pins, there is quite high voltage on the bridge rectifier (assuming that is what it is - sure looks like it) near the burned area. This almost certainly means that the entire burned area would need to be carefully removed, which is a pretty tedious task. The charred remains of the epoxy are electrically conductive. A rotary tool (Dremel type or equiv.) works well, but there will be glass particles sent flying so it is hazardous. Filling the hole and reconstructing the circuitry is another matter. This is something I would do only on something quite valuable or irreplaceable.

    Some photos showing more of the board around that area would be helpful to give an idea of what the circuit might be. I see a fuse, which together with the bridge rectifier suggests a power supply.
     
    Last edited: Oct 26, 2018
  3. dan2511

    Thread Starter New Member

    Oct 25, 2018
    4
    0
    Thank you for the reply! It is a power supply, yes. I'll upload some more photographs. Thank you so much for taking the time to reply. I've completed similar repairs in the past, just this one has stumped me with the small surface mount part being damaged, and the size of the damage on the PCB. Sonos, in their infinite wisdom, won't sell replacement parts or repair, so one's left to fix themselves! Might there be a solution to re-build part of the board somehow, or bridge the broken tracks? Even making a separate board to house the parts / area that was damaged, and connecting it back to the main board could be an option. Your advice, as those at the top of their game, would be greatly appreciated! :) Dan

    IMG_4180.jpg
    IMG_4181.jpg
    IMG_4182.jpg
    IMG_4185.JPG
     
  4. ebp

    Well-Known Member

    Feb 8, 2018
    2,332
    811
    I'm at a loss to even decide what type of converter that is. Usually I can tell "instantly" but in this case I can't. The only reason it matters is because it would provide some clues as to what that diode did.

    The circuit sort of looks like a forward converter. The clue is the inductor with two apparently separate windings on the secondary side. This is typical of an inductor required for filtering in a forward converter and putting multiple windings on the same core improves cross-regulation for multiple outputs. BUT the capacitors right up against the rectifier heatsink suggest it is a flyback converter with just some extra filtering.

    In a forward convert an extra winding is used on the primary side to discharge unwanted stored energy in the transformer. A diode is used in series with the winding. In a flyback, there is energy stored in "leakage inductance" and that is usually discharged with a "snubber" usually consisting of a diode, a resistor of moderate power rating and a ceramic or film capacitor - and I don't see the latter two items. There is an alternative to the RCD snubber, which is basically a transient suppressor diode in series with a normal(ish) diode. None of which helps much in moving toward repair.

    Were there any parts other than the small diode been removed?
    Have you looked around the web to try to find some photos of an intact board? It will be very difficult to come up with a fix without something more for guidance in terms of what was there originally.
    Which capacitor(s) was blown?
    Can you read the markings on D301 (backlit photo)? It may be one of the parts that seems to be "missing."

    It takes something very hot for a long time to completely burn down to the glass cloth in the PC laminate like that.
     
  5. pfofit

    AAC Fanatic!

    Nov 29, 2006
    57
    8
    BZX384-B9V1 has a marking code of L6 but it is a zener.
    https://assets.nexperia.com/documents/data-sheet/BZX384_SERIES.pdf
    Do you see any other component marked on that board with a Zxxx to indicate a zener?
    Was your L6 "diode" labeled with a "D"?

    I believe these had a multilayer track from the bridge to the filter caps that cooks the board from within when the caps go toast.
    Do you have a pic of the backside?
    Did that little round brown fuse survive?

    IMG_4182.jpg
     
  6. dan2511

    Thread Starter New Member

    Oct 25, 2018
    4
    0
    EBP:

    Thanks again for your reply. In answer to your questions:
    • No other parts were removed other than that diode
    • Find attached a photo of the intact board. If the issue with the diode identification is sorted, it will then be a case of working out where the left-hand-side of that diode was routed to, to re-establish the connection with a bridging wire.
    • There was one of the three capacitors blown, as you can see from the gap where one once was. They are all the same spec, voltage and capacitance.
    • I will have a look at D301 to see if I can gleam anything re a code on it. If it is the same as the missing D307 I guess it is then a case of obtaining a replacement, whether the same or similar.
    Pfofit:

    Thank you too for your reply, in answer to your questions:
    • The product in that link does look very similar, however I guess looks don’t necessarily mean anything.
    • I can’t see any other component on the board marked with a Z…
    • L6 diode was indeed labelled D307
    • Backside is not transparent/translucent in this area but will get you a photograph.
    • I haven’t tested the fuse yet, but will do. Other forums suggest that for this issue one should replace that fuse and the three caps so they are balanced.
    • Capture2.JPG
     
  7. dan2511

    Thread Starter New Member

    Oct 25, 2018
    4
    0
    Fuse tested - survived! Hoorah!
    D301 is a much larger component with different markings
    Photograph of back side attached, you can see the orientation from the shape of the scorch and the missing cap legs.
    IMG_4191.JPG
     
  8. pfofit

    AAC Fanatic!

    Nov 29, 2006
    57
    8
    The photo has a dark shadow over the + lead of the bridge.
    Thoughts.
    The "+" needs to get to the+ of the caps and it did not look like there was a trace on the front side. Also needs to probably get to the 6 pin device.
    Looks like a small trace going to TP310 would block it on the back and they have used an internal trace that
    cooked when the cap/6 pin device shorted/whatever and no place for the heat to go.
    However it looks like the burning stopped at the cap and did not burn toward the 6 pin device.
    Is there a fuse on the primary before the bridge, or is the round one it?
    Is there any other diodes with a L6/97 marking.
    The solder on the bridge pins is lacking or looks like they are cooked to, or did you have it out?.
    Have you checked for a short in the bridge?
    Do you have the number of the 6pin device mounted to the heat sink. A data sheet may help to see what the smd stuff is doing.
     
Loading...