Connect with us

Fastest AC Electronic Fuse

Discussion in 'Electronic Design' started by D from BC, May 14, 2007.

Scroll to continue with content
  1. MooseFET

    MooseFET Guest

    I worry a bit about the forward current on the Zener. I guess that
    with the values put in, the current will be small enough.

    Also if you don't mind a little leakage:

    Modified version:
    It relies on the device capacitances to hold the NPN on through the
    zero crossing.
  2. Fred Bartoli

    Fred Bartoli Guest

    MooseFET a écrit :
    Under normal working conditions (no tripping/limiting) the supply has no
    current to provide (just biasing the gate), so the average (or even
    peak) current can be made 1mA without any pb.

    Ok, but it doesn't trip, so you'll need some very serious heat sinking
    under fault conditions.
  3. D from BC

    D from BC Guest

    I'll put the goggles on.. :)
    D from BC
  4. D from BC

    D from BC Guest

    That's fast... Fast enough to be a fuse protector.. :)

    D from BC
  5. D from BC

    D from BC Guest

    Ooops..I didn't know this was an auto reset breaker..
    I was looking for negative feedback..I thought this was a
    limiter(current regulator).. I didn't recognize the SCRish transistor
    arrangement..with it's own positive feedback loop.
    Biasing the NPN base??
    That's from when I mistakened this as a linear current limiting

    This circuit is probably going to keep me busy for hours..
    It's all in the bridge :)

    What's the history on this circuit? Textbook? Well known? Own
    invention? Magazine? Old? New?

    D from BC
  6. Fred Bartoli

    Fred Bartoli Guest

    D from BC a écrit :
    Uh, do we really need to copy such circuits from somewhere?
    Just sketched it in 2 minutes.
  7. D from BC

    D from BC Guest

    2 minutes!!

    My imagination failed me :(
    It's a mental block or something..Or I have too much DC in the
    It's the lead.... :p

    It's like shorting a bridge out...that's wack!! :)
    Obviously...I have only used bridges in basic applications..

    D from BC
  8. qrk

    qrk Guest

    Bussmann TPS fuses clear in less than 5ms. TPS series are 170VDC rated
    fuses. In our test setup, we shorted out a 2.5F cap bank charged up to
    100VDC. Made a tiny nick in the copper shorting block. Peak current
    was around 2k Amps (don't quite remember the details). Fuse worked
  9. whit3rd

    whit3rd Guest

    It should be noted that microsecond turnon is usually a problem for
    either a photoSCR or an integrated SCR/Zener combination (SIDActor
    will perform best with rapid turnon requirement.

    Either a fuse, OR a positive-tempco polyfuse should work well; if
    you use the polyfuse, a few seconds of power-off will reset the
  10. MooseFET

    MooseFET Guest


    I've added some "*"s
    No, it does trip. Note the resistor marked with the "*"s. When there
    is a drop on the MOSFET, it biases the NPN on thereby providing the
    positive feedback needed for the latching action.
  11. Fred Bartoli

    Fred Bartoli Guest

    MooseFET a écrit :
    Ah, yes missed that. Then it's perfect.
  12. Marra

    Marra Guest

  13. D from BC

    D from BC Guest

    Of course.
    I'm getting better at the noise reduction layer in electronics

    What might be a kickass combo would be the bridge based breaker
    circuit(as posted in this thread) followed by a triac shunt.
    But..probably overboard...

    I'm currently playing with mutations of the bridge breaker circuit..
    Such as:
    RL current sensing
    Turn slew control
    Low current consumption control electronics
    Single IC solutions
    D from BC
  14. joseph2k

    joseph2k Guest

    Have you tried telephoning or emailing any fuse manufacturers? Every time i
    have asked one they were very helpful.
  15. D from BC

    D from BC Guest


    Hopefully, I won't need any fuses if I cook up an AC limiter/breaker
    based on posts in this thread.

    I've cooked up a 2 op amp latchup circuit for inside the bridge.
    It has massive positive feedback for latch up.
    This allows a cool 0.1ohm current sense resistor.
    Thermal stability and precision current trip point are also benefits.
    Faster speed can be accomplished with high speed comparators.
    I've come up with a single IC circuit with no discrete small signal

    D from BC
  16. Ian

    Ian Guest

    Don't knock the discrete transistor approach.
    To turn the Mosfet off fast you need to pull the gate charge out, which
    relatively high peak currents. The discretes give you this capability.

    Slightly off topic: when I were but a slip 'o a lad, I worked for a summer
    in a test lab. One fun bit was helping a Real Engineer test a circuit
    breaker - a 1A model was the target. First we checked that it tripped
    properly at the nomial current, then came the fun bit.

    A circuit breaker has to break the circuit, especially under fault
    conditions. That means you have to check that it does so under
    not just reasonable conditions, but also unreasonable ones.

    The Real Engineer's test area was surrounded by a metal cage,
    and in back a door led into a vault which had double walls, and
    was filled with a couple of tons of heavy duty batteries, switchgear
    and massive bussbars through the walls into the test area.

    We spent a morning making sure the batteries were all fully charged,
    and connecting up the poor little unsuspecting Consumer Unit
    1A circuit breaker in a safety chamber. We then retired to a "shack"
    within the test area, and threw the switch. ISTR the current was
    a "pass" as it broke the circuit ;-)

  17. MooseFET

    MooseFET Guest

    I suggest you use the comparitor. Dollar for dollar, they are usually
    much faster than op-amps. IIRC, you can get a comparitor and
    reference in one package from Linear.

    IIRC: The LTC1440

    It seems to me that if you find the right comparitor, one comparitor
    will do. You only need to find some way to do the positive feedback.
    Many comparitors have inverted and non-inverted outputs.

    If you use the trick of using the drain voltage of the MOSFET as the
    place to take the positive feedback from, you may be able to make the
    circuit reset on each zero crossing.

    BTW: I think you also need a stout MOV across the bridge. On an
    inductive load, the turn off could be a bit exciting.
  18. D from BC

    D from BC Guest

    True...That MOSFET needs to be pulled down hard for fast switch off..
    But...I've come up with another solution for that bridge based AC
    Using a Mosfet driver chip!
    Some of the mosfet driver chips have all the features I need to make
    an AC breaker/limiter such as:

    hard switch off
    cool resistor current sensing
    overcurrent latched shutdown..
    internal reference
    error output
    Power rail ripple rejection
    Maybe glitch rejection
    Maybe low idle current??

    This reason why I gravitate toward IC solutions is:
    * Less parts for faster PCB developement
    * IC's are often temp compensated
    * Smaller PCB
    * Often less complicated math
    * Possible faster developement time

    About that high peak amp breaker story..
    Why did it the breaker circuit blow up but saved the unit?
    No fuse pop?
    Slow breaker?
    Shunt type?
    Goofy design?
    D from BC
  19. D from BC

    D from BC Guest

    Breaking an inductive load = zap!
    So.yeah...MOV for sure..

    I might drop the comparator idea and use a mosfet driver IC with
    current sensing.

    But let's say I keep going with comparators...
    I believe you're right that only one comparator with a positive
    feedback arrangement can turn off the mosfet.
    I might use a comparator with differential outputs.

    The 2 op amp latching circuit I mentioned is a little goofy..
    Op amp 1 works in linear mode and limits the MOSFET current..
    There is a lot of negative feedback..
    Op amp 2 senses the Mosfet gate in linear mode and with positive
    feedback (gain^2) shuts down op amp 1 to shut off the mosfet.
    Manual reset required.. (But that's ok)

    The general idea is that a temporary linear mode may clip the current
    spikes but may not be enough to trip the latch for breaking. It's a
    way of getting overcurrent protection with fewer interruptions.
    Note: This is employed in some Mosfet drive IC's.

    D from BC
  20. Ian

    Ian Guest

    Mosfet driver chip sounds good.

    On the circuit breaker test, the breaker has an electrical path,
    a current sense for that path (usually magnetic), and a switch.

    The issue is whether the switch opens at high overcurrent,
    or if the contacts get welded shut. You have to test a number
    of breakers at steadily increasing currents to make sure
    there is no intermediate fault current that produces the
    welding action. The final test blew the breaker to pieces,
    but the circuit was broken so the result was a "pass".

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