Connect with us

Reverse and Overvoltage protection

Discussion in 'Electronic Design' started by Richard Hosking, Jan 11, 2004.

Scroll to continue with content
  1. Dear all
    I want to design a power supply for a low power rig with protection for
    reverse and overvoltage. The requirement would be about 1A at 12-14V.
    What is the best way of achieving this? I guess a diode would give
    reverse voltage protection but the 0.6V drop is a problem.

  2. If the voltage drop from a diode is a problem, perhaps consider using a
    larger (voltage) transformer with a readily available 12-v voltage
    regulator like a 7812 chip. The regulator chip comes in several
    varieties, and I'm not sure, but I think the 7812 that radio slap sells
    can handle an amp an a half. This will keep the voltage at 12v pretty
    steady, but you'll need a 15-volts-when-loaded transformer to feed it

    They also have other voltage regulator modules that would provide a high
    enough voltage that a few .6 volt drops would be acceptible. May not look
    very pretty, but once it's in the case, nobody will know.

    You can also do what I just did... I thought I fried my 7812 regulator,
    and I found that it hadn't been regulating at all (because the supply
    voltage was at 12v when under load, not enough for the chip to regulate),
    so in my soldering, I got it backward once, and it got really hot. I went
    and bought an LM317 from Radio Stank and have been using it. Now I can
    vary my voltage from my new 26 volt transformer from 1.2 up to 35v (35
    under no load). Only bad thing here is my box is too small to fit the
    larger transformer and 3-inch-square heatsink I need ;-/ But with a large
    enough box, it should do quite nicely.

    By the way, at the RS, the 7812 regulator is $1.49, and the LM317T is
    $1.99. Both are TO-220 cases.

    Just my thoughts...
  3. scada

    scada Guest

    If you use a "Crowbar circuit" for overvoltage protection, that should take
    care of your reverse voltage problem as well. A crowbar is usualy an scr
    connected in paralell to the output and is trigered by a comparator circuit
    that will output to the scr gate when a reference voltage is exceeded. That
    can be as simple as a zener diode! The supply output is shorted resulting in
    a tripped breaker or blown fuse! Now if you were to encounter a backfeed in
    excess of that crowbar voltage, the supply would trip. If the supply itself
    were to exceed the reference it would also trip the supply!
  4. Mac

    Mac Guest

    I saw a really cool reverse polarity protection device in a Bob Pease
    streaming video from National's website. I think it was just an N-channel
    MOSFET hooked up in series with the ground return, but with the drain
    grounded, and the source connected to the negative side of the load. The
    gate was connected to the positive terminal of the supply. Like this: (use
    courier or similar font)

    V+ -----------------------+-------------------|
    | |
    | |
    ---- +--+
    drain | |source | | load
    V- --------------------+ +-----------+ | |
    | | |
    | +--+
    | |

    So when the polarity is correct, the MOSFET is on, and the source-drain
    diode is forward-biased. But when the polarity is reversed, the MOSFET is
    off and the source-drain diode is reverse biased, so no current flows.

    I believe I have it right, although it seems strange to have the current
    flowing from source to drain under normal operation. But if you do it the
    other way, it won't work because the diode will be forward-biased under
    reverse voltage. And that's not what we want.

    By the way, this technique is apparently patented.

  5. Well since you need both : Use a Zener diode.
    ASCII schematic ( use fixed width font like a courier to see correctly )

    in ----[=====]-----+-------- out
    Fuse |
    / \
    gnd ---------------+-------- gnd

    Reverse polarity : The zener works like a regular diode. Blows the fuse to
    smithereens : load protected

    Overvoltage : the zener will conduct creating a short := fuse blows to
    smithereens : load protected.

    Get a zener diode that can handle 3 to 5 watts. If you need to protect a
    load of 12 volts use a 12.1 volts zener. Your load will never seem more
    then about 12.3 volts ( there is some tolerance before the zener really
    goes 'HARD' into on state. )

    Other solution
    --->|---------| |-------------
    diode | 7812 |

    This'll get you in the ballpart. The 7812 will create 12 volts at its
    output. The diode at the input protects against reverse polarity
    drawback you'll need around 15 volts at the input in order for it to work
    if you cool the 7812 you can drain 1.5 amps out of it ( provided it doesn't
    need to sink too much voltage : check max power dissipation )

  6. JeffM

    JeffM Guest

    I want for reverse and overvoltage.
    A crowbar-type clamp gives reverse protection with zero insertion loss.

    +V +--------+
    ---|| ||--+------
    +--------+ |
    / \
    /___\ D1
  7. JeffM

    JeffM Guest

    I saw a really cool reverse polarity protection device
    Still has a 1-diode drop. FET is more expensive than a diode.
  8. I think the point is that when the supply is correctly connected, the
    voltage drop across the FET is across Rds, which should be pretty low
    for a decent sized FET (say 0.1 ohms). The drop across this would only
    be 100mV at 1A.

  9. Mac

    Mac Guest

    I shouldn't have even mentioned the diode in the forward direction case.
    In fact, it might not be forward biased because the drop from source to
    drain should be very low. The resistance may potentially be only a few
    milliohms for some FET's.

    Also, please be careful when "quoting" people. You completely re-drew the
    circuit. And though you used indents indicating quoting, you didn't
    actually attribute my words to anyone.

    best regards,
  10. The Mosfet may be destroyed by the first transient that comes along.
    Abs max Vgs is usually only +/-20V.

    Protection may be as simple as a large value resistor in series with
    the gate. (It forms an LPF with the large Cgs.) Use a zener if
    paranoia persists.

    Also, the OP's "low power rig" probably has a negative earth. A
    P-channel FET in the positive lead may be more appropriate than an
    N-ch in the negative lead. The adventurous may try an N-ch in the
    positive lead with a suitably protected high side driver.

  11. GPG

    GPG Guest

    Relay contact

    Vin ----------------o-o/ o- Vout
    Relay |_/_|-

  12. JeffM

    JeffM Guest

    and the source-drain diode is reverse biased, so no current flows.
    Look again.
  13. In addition to the other suggestions: consider using a Schottky diode.
    Lower forward drop (around 0.4v or a bit less). Disadvantage is relatively
    low reverse voltage tolerance, perhaps 40v depending on which one you
    choose. Higher reverse voltage means higher forward drop, and vice versa.
  14. Mac

    Mac Guest

    I don't see why looking again should change anything. You kept my
    signature, but not there is no attribution at the top.

    Even if keeping the signature constitutes attribution, you still changed
    my diagram a little bit without making it clear that you did so.

    It's no big deal.

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