Connect with us

7805 failure mode

Discussion in 'Electronic Design' started by Winfield Hill, Jul 23, 2007.

Scroll to continue with content
  1. My co-author Paul described an interesting
    failure mode for a 7805 regulator IC, which
    was in a circuit with a modest load current,
    and with about 1uF of load capacitance.

    The failure symptom is that when 12V power
    is first applied, the output voltage first goes to
    12V, before dropping back to 5V and operating
    properly. When the power is turned off, the
    output voltage again spikes high.

    One possible clue: when the 7805 was first
    used in this circuit, a hand-wired prototype,
    the ground pin was left floating. But when
    the failure mode was observed this pin was
    properly grounded.
     
  2. Just wondering if all 7805's are the same circuit?


    martin
     
  3. Rich Grise

    Rich Grise Guest

    Dumb question, but after observing this, did you try the same experiment
    with a new 7805, but wired correctly before applying power?

    Thanks,
    Rich
     
  4. mpm

    mpm Guest

    What are you measuring with (O'scope?, voltmeter?)
    What is the duration of the spike?

    Either way, this sounds like a busted 7805.
    Or, you have something else hung on the output, intentional or
    otherwise.....

    -mpm
     
  5. Winfield

    Winfield Guest

    Yes, we can wonder. This was a Fairchild part,
    and Paul has given me the rest of his stock,
    two pieces.
     
  6. Winfield

    Winfield Guest

    I have not, but I have the two remaining parts
    from the manufacturer. I also have the bad part.
     
  7. Winfield

    Winfield Guest

    It certainly appears busted, all right.
     
  8. Jim Thompson

    Jim Thompson Guest

    The probable failure mechanism was...

    Turn on without a ground. While it had no reference, it should have
    current-limited without damage.

    But the failure occurred when it was unpowered and the output
    capacitor dumped thru the device to the input.

    ...Jim Thompson
     
  9. Some clarification needed: by 'the ground pin' you are referring to the
    7805 common pin, correct?

    When the pin was properly grounded, did the 7805 resume normal
    operation? If so, I wouldn't call what you experienced a 'failure'. It
    would be 'unspecified behaviour'.
     
  10. Joerg

    Joerg Guest

    AFAIK they have no reverse discharge diode protection. Might have been
    killed on the first turn-off. Anyhow, I have pretty much completely
    adopted the LM317 for almost everything positive. It can take at least
    modest spikes in reverse.
     
  11. Jim Thompson

    Jim Thompson Guest

    It's the things that don't show on the schematic... isolation
    junctions.

    ...Jim Thompson
     
  12. That's my thought (the 11V or so on the output would be enough to
    break down E-B junctions and maybe there would be enough current to
    cause damage if the input had a heavy load or short to ground),
    although I don't see from the schematic what could make it act as
    described.

    http://focus.ti.com/lit/ds/symlink/ua7805.pdf


    Best regards,
    Spehro Pefhany
     
  13. Paul Burke

    Paul Burke Guest

    We had a funny failure with 7908s a few (15?) years ago. One batch of
    ST- made devices gave an output of -10 volts, within a few mV,
    absolutely consistent. We thought we had been given mismarked units
    until we found that ST didn't make 10V versions, so we tested further,
    and found that the exact -10V was pure coincidence, and that above a
    certain forward voltage, the output tracked exactly the input, just a
    few volts down. As the PSU was a kludge (a 12V supply made to give 8V by
    heating up the world), the input was over this threshold. Problem solved
    by using someone else's devices, and we never bothered investigating
    further.

    Paul Burke
     
  14. Joerg

    Joerg Guest

    Or a too small area for the reverse jolt from the output cap upon
    turn-off. IIRC that's next to nothing for these regulators.
     
  15. Winfield

    Winfield Guest

    It operates normally, except for the short turn-on
    voltage spike every time the power is applied.
     
  16. Winfield

    Winfield Guest

    Yes, it appears the current-limit finction should work
    properly with the common floating, but the substrate
    would be pulled well above the output-pin's voltage
    by the startup and over-temp shutdown circuit.
    I'll check with Paul, but I don't believe such a
    shutdown-current path exists in this case. Also,
    doesn't that kind of mistreatment destroy the part?
    Paul says this one works normally once it's running.
    But it makes a voltage spike each time it starts up.
     

  17. Some data sheets show a diode from the output, back to the input to
    prevent damage.


    --
    Service to my country? Been there, Done that, and I've got my DD214 to
    prove it.
    Member of DAV #85.

    Michael A. Terrell
    Central Florida
     
  18. Jim Thompson

    Jim Thompson Guest

    Win, Have you examined Paul's circuit to make sure it doesn't have
    some design defect... perhaps loaded under the minimum?

    ...Jim Thompson
     
  19. I wouldn't call that normal if the device specifications state that it
    should maintain voltage within a certain envelope following input supply
    transients. Unless the device spec actually addresses operation with a
    floating ground pin (doubtful).

    I also wouldn't call it a failure if, once the pin is hooked up
    properly, the device commences operation within spec. On the other hand,
    it would be a device failure if, once the pin is properly connected,
    this anomalous behavior continues.

    Sorry if I'm a bit of a terminology Nazi here. But I've done quite a few
    FMEAs for avionics and other stuff. The FAA (and others) really pick
    nits about which is a failure and what its effects (subsequent failures
    vs operation out if spec) are.
     
  20. There's no minimum load for the 7805 (unlike the 317)... there's an
    internal 5K or so to ground (the feedback divider chain) so it should
    even be able to sink a bit of current, at least at room temperature.

    Of course if an input filter e-cap was connected from input to output
    rather than from input to ground you'd get the spike on turn-on but it
    would spike negative (or not at all) on turn off.

    It really seems a bit odd that the output should spike right to 12V
    (the input voltage) on turn on from some internal fault... usually
    those things have a drop of a volt and change even at zero output
    current, but perhaps Paul was speaking loosely.



    Best regards,
    Spehro Pefhany
     
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

-