PCB Troubleshooting

Discussion in 'The Projects Forum' started by dledge, Jan 13, 2015.

  1. dledge

    Thread Starter Member

    Feb 11, 2014
    60
    1
    First timer...

    I created a PCB layout (attached) using PCBExpress. I received the boards, soldered in the parts...and I'm less than pleased. Issues that I'm seeing:

    *There appears to be a short between the +DC Pin and ground. I've removed all components that could possibly link this pin to ... anything, and I'm still getting voltage. Additionally, the resistance between this pin and ground is a little less than 1K. Is it common for PCBs to have this kind of short?

    *The HEX IC is getting extremely hot. I'm using a CD4572UBE and connected per the attached diagram. I am wondering if the heat is due to oscilations in the chip due to the issue above bleeding a ~5V signal (connection between R13 and pin 12).

    *The backup battery power supply are two 9V batteries hooked up in series; however, the voltage as seen by the voltage regulator is only about 7V rather than the expected 12V. The voltage seen at the ICs are hovering around the 5V mark.

    Any guidance on how I should trouble shoot this would be most helpful. I have tested the other two boards that cam with the order, and I do not have any 'shorting' issue as described in the first bullet.
     
  2. mcgyvr

    AAC Fanatic!

    Oct 15, 2009
    4,771
    971
    Assuming you paid for electrical testing (E-test) or it was included by default with all orders at PCBExpress you shouldn't have any shorts.
    I order thousands of circuit boards each year and its very rare (maybe 1 out of 100,000 that make it to me) to have a short but it does happen. That usually means your PCB vendor really isn't Etesting them and its time for a new vendor :) .
    An E test will find it.
    but 1k isn't a short either really and could easily be from the parts you already soldered.
    I can't think of any PCB's I've troubleshooted that ever had a manufacturing defect of anything close to 1k..
    Its just a dead short.. Typically just some copper that didn't etch properly or other foreign objects in the base laminate.

    Seeing as you already soldered parts you could have a solder bridge/splash or a bad layout from the start or a bad/damaged component.
    At this point you need to solder up the other boards and go from there.

    As you didn't/forgot to attach your schematic we can't help with the hex ic issue.
     
  3. dledge

    Thread Starter Member

    Feb 11, 2014
    60
    1
    I've added the schematic...sorry about leaving it off.

    I was able to confirm that the resistance between the power and ground layers is 0.410K. Does this number have any significance? I would have expected this number to be extremely high.
     
  4. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    Is it happening with all of the boards? You could try using some large current to try to burn that short and see where it is happening (unless you can find it physically by eye)
     
  5. dledge

    Thread Starter Member

    Feb 11, 2014
    60
    1
    I've just built up the one board. I've etested the other board and everything looks good.

    Does the flux from the solder conduct at all? Should I scrap off the flux from the solder points? None of the solder is bridging, but there is a lot of flux residue.
     
  6. Brevor

    Active Member

    Apr 9, 2011
    279
    54
    Please post your schematic, maybe that will help.
     
  7. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    Flux has very little conductivity, most likely you have some solder short sepecially if you pulled off all the parts, can you post a photo of the board?
     
  8. dledge

    Thread Starter Member

    Feb 11, 2014
    60
    1
    Is it possible the hex chip is toast?
     
  9. dledge

    Thread Starter Member

    Feb 11, 2014
    60
    1
    I attached the board schematic above.
     
  10. dledge

    Thread Starter Member

    Feb 11, 2014
    60
    1
    Last edited: Jan 14, 2015
  11. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,291
    1,255
    Post the circuit schematic, not the board.
     
  12. JohnInTX

    Moderator

    Jun 26, 2012
    2,347
    1,029
    What kind of voltage regulator is that? Its pinned incorrectly if its a standard 7812 / LM340T-12. It looks like its installed one position over - GND should be the center pin.
     
  13. dledge

    Thread Starter Member

    Feb 11, 2014
    60
    1
    Circuit diagram attached.
     
  14. dledge

    Thread Starter Member

    Feb 11, 2014
    60
    1
    Its a LM1085.
     
  15. JohnInTX

    Moderator

    Jun 26, 2012
    2,347
    1,029
    Swing and a miss!
    Good luck.
     
  16. JohnInTX

    Moderator

    Jun 26, 2012
    2,347
    1,029
    You are powering U4 from 12V but hitting the input on pin 12 with 18V. It won't like that. Max input volts is Vdd +.5V
     
  17. dledge

    Thread Starter Member

    Feb 11, 2014
    60
    1
    I didn't realize that. Would this cause the ic to fry?
     
  18. JohnInTX

    Moderator

    Jun 26, 2012
    2,347
    1,029
    Yep. It will cause latchup of the CMOS input, turning it first into an SCR then silicon slag. On the photo, there is a solder bubble at pin 12 - hot..
    The datasheet Features indicates an input current spec at 18V but that assumes Vdd = 18V at least.

    EDIT:
    It looks like the input is to sense when wall power is on so.. split the 1.8K to make a voltage divider to pick off the input at a safe level - use some high value R across the LED to pull the CMOS input down when wall power is removed.
     
    Last edited: Jan 13, 2015
  19. dledge

    Thread Starter Member

    Feb 11, 2014
    60
    1
    So...that would explain what I think is a short circuit. Would it be safe to assume this type of fault would have cooked other components?
     
  20. JohnInTX

    Moderator

    Jun 26, 2012
    2,347
    1,029
    It might have cooked U2 if it applied >12.5V to pin 12. Before fixing it, fire it up and measure that voltage. Also when you pull U4 manually jumper u2-12 low and high to see if the output changes (assuming pin 13 is low - OR gate..).
    The rest should be OK but you don't know until you get U4 gone.

    The voltage reg might not be happy now that I think about it. It won't mind short circuits / overloads but won't like being back-driven by a seriously failed U2. Just have to see.

    See the edit above for a suggestion to fix the circuit until you rev the board.

    Have fun.
     
    Last edited: Jan 13, 2015
Loading...