How do I cut power to my circuit in the event of meltdown?

Discussion in 'The Projects Forum' started by jellytot, Jun 8, 2016.

  1. jellytot

    Thread Starter Member

    May 20, 2014
    I have a 2 Amp (max rated) wall adapter that powers my project.
    My project draws about 1 Amp (not including the solenoid).
    The solenoid draws 2.5 Amps, but it is only for a fraction of a second (milliseconds), several times a minute. I've been told this won't overload my power supply, since the pulse is so quick.
    However. Twice, my project locked up and the solenoid was stuck in the "on" position, but I unplugged it within seconds. This happened a long time ago, and has not occurred since. I have added measures to keep this from happening again (e.g. flyback diode on the solenoid, caps on the MCU, etc).

    I want to leave my project on, overnight. But I'm worried about the tiny chance it will lock again, and draw 3.5 Amps continuously. What will be the consequences if this happens? Fire? Can someone recommend a fail-safe, for example, is there something I can build into my circuit that will cut power if it detects 3.5 Amps being drawn for too long (or something like that)?

    Last edited: Jun 8, 2016
  2. ISB123

    Well-Known Member

    May 21, 2014
    #12 and jellytot like this.
  3. jellytot

    Thread Starter Member

    May 20, 2014
    Ah! Thank you. Never knew about polyfuses! I have some follow-up questions if you don't mind:
    1. I can easily get polyfuses with Trip Current between 2.2A - 2.7A. Since the power adapter is rated 2A, but normally the project draws 1A (but there are regular but extremely short spikes of 3.5A), I'm thinking a 2.2A polyfuse should be good?
    2. I should place the polyfuse on the positive wire, in series, between the power adapter and the project?
    3. If it trips, should I worry about the polyfuse overheating?
    4. Some of the polyfuses I have access to have a Trip Time of 9 seconds. Is that too long, that components will be damaged?
    5. Should I look for a higher Holding Current rating, because of my amperage pulses?
    6. Once the polyfuse trips, does it completely cut off power? That would be desirable.
    7. I read that a tripped polyfuse will "stabilize at a significantly higher resistance (up to 4 times initial value)". Does this mean that the fuse's Holding and Trip Current stats will decrease?
  4. GopherT

    AAC Fanatic!

    Nov 23, 2012
    Other things that keep a project safe for testing are
    - set project on a non-combustible surface (ceramic tile, concrete, ...)
    - use an extension cord to get the wall adapter off the wall (onto the non-combustible surface)
    - place a battery-powered smoke detector near (above) the project
    - ...
  5. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    You may have already done the 9 second exposure test the last time your device locked up. If everything worked after...

    I don't like a poly fuse here unless the device resets and releases the solenoid before it resets itself and turns power back on. A standard fuse may be better or safer, though will need replacement every time the problem occurs.

    A slow blow fuse would be ideal here.
  6. jellytot

    Thread Starter Member

    May 20, 2014
    The last time it locked up, I unplugged it after a few seconds. I guess it would have survived after 9 seconds?
    But yes, I'm also leaning towards using a standard fuse (slow blow), with replacements. Thinking 2.2A Trip should be good. Thanks!
  7. bertus


    Apr 5, 2008

    If you want to know more about fuses and polyfuses have a look ath the attached PDF's

    jellytot likes this.
  8. ian field

    Distinguished Member

    Oct 27, 2012
    The first thing you could try is a big electrolytic to cover the deficit when the solenoid draws its peak current.

    A slight improvement on that is a big electrolytic on the wall-wart output to handle the solenoid draw, feed the project via a diode and give it a reservoir electrolytic all to itself - then it won't see the momentary drop caused by the solenoid. A Shottky barrier diode loses less Vf than regular silicon.

    Don't forget the back emf diode on the solenoid coil.
  9. mcgyvr

    AAC Fanatic!

    Oct 15, 2009
    Or just buy a power supply capable of not going into meltdown when an "issue" occurs..
    Probably very little difference in price.. maybe a bit larger if that..
  10. DNA Robotics


    Jun 13, 2014
    Maybe a thermal fuse placed near the parts that would overheat.
  11. jellytot

    Thread Starter Member

    May 20, 2014
    Thanks! Wow, there's a lot of info in there. :)

    They make those?! What's the feature called?
    Last edited: Jun 9, 2016
  12. Dodgydave

    Distinguished Member

    Jun 22, 2012
    You could put a resistor say a third of the coil value in series with the solenoid coil, with a 470uF capacitor in parallel with the resistor, then the solenoid current will slowly be limited when the capacitor charges up.
  13. mcgyvr

    AAC Fanatic!

    Oct 15, 2009
    I meant just buy one thats rated to 5 Amps or so..
    Then even if it gets stuck on its not an issue anymore..
  14. jellytot

    Thread Starter Member

    May 20, 2014
    Lol. OK, I see.

    Good advice. Thanks!
  15. grahamed


    Jul 23, 2012

    If the solenoid is only ever activated for a few miilli-seconds how about feeding the base of the solenoid driver through a series cap?
    AnalogKid likes this.
  16. AnalogKid

    Distinguished Member

    Aug 1, 2013
    That is an *excellent* idea. You could put a 100,000 uF cap in series with the solenoid, but your way does have a certain advantage. OK, a huge advantage.

  17. #12


    Nov 30, 2010
    General Motors uses a polyfuse on their door lock solenoids. I had to fix one. The polyfuse protected the car from burning down.
    There is also the LM723 foldback circuit. Try to use too much current and it pretty much shuts down.