Failure mode of 7805?

Discussion in 'General Electronics Chat' started by coldpenguin, Jun 17, 2010.

  1. coldpenguin

    Thread Starter Active Member

    Apr 18, 2010
    165
    9
    I have a small 7805 circuit, a 7805 and two capacitors, which I have been using as a regulated power supply for a project.
    However, one day, after no modifications that I know about, the circuit is reading ~2V across my voltage test points. The circuit seems to work when I use the ICD to power it, so I guess it is the 7805. Or would it be the caps?
    I was surprised that the supply provides 5V unloaded, but only shows 2V when attached to my final circuit. I would have expected 0V output!
    Is this normal?
     
  2. dsp_redux

    Active Member

    Apr 11, 2009
    182
    5
    Maybe a shorted cap? Can you post the schematic you are using? What kind of load are you using?
     
  3. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    I'll bet that you don't have a heat-sink on the 7805, it overheated, and shut itself down to protect itself. Let it cool, an the +5V will back just long enough for it to overheat again.
     
  4. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    Actually, if you stress them enough they will shut down and not restart, or have their current capacity drastically reduced. Early Commodore 64's had this problem, they tried using epoxy as thermal conductive material. It worked, until you buried the box under something (like newspapers). I remember taking a screwdriver and using it as a chisel using a hammer to get to the device, then adding a metal heatsink to the new regulator to fix them. It was a really crappy design, and they made huge numbers of them.
     
  5. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    If the input and/or output caps are missing or intermittent, they can oscillate at high frequencies (MHz range). If you're using a meter to measure the output, you'll see anywhere from 1v to 3v on the output.
     
  6. coldpenguin

    Thread Starter Active Member

    Apr 18, 2010
    165
    9
    It really doesn't warrant a schematic! This was a 'power supply' I had been using and swapping between projects. The current draw was at most 0.3A at the 12V rail (I had been checking as I was trying to reduce the current draw). The heat sink, I think is decent enough (see photo IMG_004, small amount of heatsink gunge used).
    Think I drew the schematic wrong, the -ve, white bar on the capacitors are to the ground rail. These are 47 uf, 25V.

    There is no physical signs of damage, which is why I am suspecting the regulator itself, I suppose I am just surprised that it isn't a 0V output rather than 2V
     
  7. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    What is the input voltage? The power dissipation in the regulator is (Vin - 5)*0.3.
    If Vin =24V, for example, the power dissipation would be 5.7W, which would make it get quite hot even with the pictured heat-sink.
     
  8. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Your 47uF electrolytic caps are large and quite a ways from the regulator.

    The specifications call for an 0.33uF cap between the IN and GND pins, and an 0.1uF cap between OUT and GND. They should be as close as possible to those pins, and have very short leads.

    Electrolytic caps can fail internally with no obvious external signs. Just the other day, I replaced a 220v cap in a TV power supply that had opened internally.

    The 0.33uF and 0.1uF caps are minimum requirements. You should use metal poly or ceramic caps. You can add more if you like, but those are required.

    Using a large cap on the output can cause the regulator to go into crowbar (shutdown) mode
     
  9. coldpenguin

    Thread Starter Active Member

    Apr 18, 2010
    165
    9
    Sorry, Vin is currently 12V regulated hobby supply, later will be 12-14V ish (gel battery, later intended to be solar charged).

    I'll try replacing and moving the caps first then before throwing it in the bin.
     
Loading...