Shop OBEX P1 Docs P2 Docs Learn Events
Wall wart through a bridge rectifier? — Parallax Forums

Wall wart through a bridge rectifier?

ArchiverArchiver Posts: 46,084
edited 2003-07-16 21:00 in General Discussion
I was asking on usenet about combining wall wart power with a 9V battery,
using diodes to protect the battery and detect when the wart was supplying
power (can't use the jack's auto-battery ground disconnect feature.

I figured I could run the wall wart + through a diode (1N4148) to the 7805.
The battery + would go through another diode to the 7805 as well. When the
wart is plugged in, the battery's diode would be reverse biased and shut off
the battery.

I could then run a trace from between the wall wart and its diode to the MCU
and detect when power is coming from the wall wart (for power saving
features when not present). Still have to figure out how to protect the MCU
from the wart voltages.

Anyway, a suggestion came back to use a bridge rectifier for the wart
instead of a single diode. The idea being that then it wouldn't matter what
polarity the jack is or even if it was AC - everything would work.

Anyone know of any gotchas with this or is it a good idea?

Comments

  • ArchiverArchiver Posts: 46,084
    edited 2003-07-16 17:20
    Yeah, I think the bridge rectifier would work... That's what the AVR STK500
    dev. kit uses on its input. It seems to work really well.

    - Robert



    Original Message
    From: "pm" <pmeloy@s...>
    To: <basicstamps@yahoogroups.com>
    Sent: Wednesday, July 16, 2003 9:48 AM
    Subject: [noparse][[/noparse]basicstamps] Wall wart through a bridge rectifier?


    > I was asking on usenet about combining wall wart power with a 9V battery,
    > using diodes to protect the battery and detect when the wart was supplying
    > power (can't use the jack's auto-battery ground disconnect feature.
    >
    > I figured I could run the wall wart + through a diode (1N4148) to the
    7805.
    > The battery + would go through another diode to the 7805 as well. When the
    > wart is plugged in, the battery's diode would be reverse biased and shut
    off
    > the battery.
    >
    > I could then run a trace from between the wall wart and its diode to the
    MCU
    > and detect when power is coming from the wall wart (for power saving
    > features when not present). Still have to figure out how to protect the
    MCU
    > from the wart voltages.
    >
    > Anyway, a suggestion came back to use a bridge rectifier for the wart
    > instead of a single diode. The idea being that then it wouldn't matter
    what
    > polarity the jack is or even if it was AC - everything would work.
    >
    > Anyone know of any gotchas with this or is it a good idea?
    >
    >
    > To UNSUBSCRIBE, just send mail to:
    > basicstamps-unsubscribe@yahoogroups.com
    > from the same email address that you subscribed. Text in the Subject and
    Body of the message will be ignored.
    >
    >
    > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2003-07-16 17:24
    One Gotcha is that the Rectified AC has a LOT of ripple on it.
    (120 HZ). The 'dual diode' approach would prevent the 9V
    from providing power ONLY when the wall-wart output was
    above 9.4 volts. This works fine for a DC wall-wart. The
    ripple from an AC wall-wart would let the 9V conduct many
    times a second.

    --- In basicstamps@yahoogroups.com, pm <pmeloy@s...> wrote:
    > I was asking on usenet about combining wall wart power with a 9V
    battery,
    > using diodes to protect the battery and detect when the wart was
    supplying
    > power (can't use the jack's auto-battery ground disconnect feature.
    >
    > I figured I could run the wall wart + through a diode (1N4148) to
    the 7805.
    > The battery + would go through another diode to the 7805 as well.
    When the
    > wart is plugged in, the battery's diode would be reverse biased and
    shut off
    > the battery.
    >
    > I could then run a trace from between the wall wart and its diode
    to the MCU
    > and detect when power is coming from the wall wart (for power saving
    > features when not present). Still have to figure out how to protect
    the MCU
    > from the wart voltages.
    >
    > Anyway, a suggestion came back to use a bridge rectifier for the
    wart
    > instead of a single diode. The idea being that then it wouldn't
    matter what
    > polarity the jack is or even if it was AC - everything would work.
    >
    > Anyone know of any gotchas with this or is it a good idea?
  • ArchiverArchiver Posts: 46,084
    edited 2003-07-16 18:36
    At 08:48 AM 7/16/03 -0700, pm wrote:

    >I figured I could run the wall wart + through a diode (1N4148) to the 7805.

    Use a bigger diode - any of the 1n400x family is good. The 1n4148 is only
    good for about 75 mA and will not withstand large surge currents (such as
    when the capacitors charge up).

    We generally use 1n4005 but any of that family will work.

    >The battery + would go through another diode to the 7805 as well. When the
    >wart is plugged in, the battery's diode would be reverse biased and shut off
    >the battery.
    >
    >I could then run a trace from between the wall wart and its diode to the MCU
    >and detect when power is coming from the wall wart (for power saving
    >features when not present). Still have to figure out how to protect the MCU
    >from the wart voltages.

    Use a resistive voltage divider. The resistor chain should be low enough
    value (total resistance) to ensure that the input voltage collapses all the
    way to ground. Figure on 10K total.

    The actual resistors in the divider depend upon the wall wart voltage which
    you have not specified. You want the voltage to the stamp pin to be well
    above the 1.4V threshold but below the 5V supply. For example, with a 12V
    nominal input voltage, 7k5 series resistor and 2k4 shunt resistor will work
    well (about 4:1 divider ratio). Input reads a logic 0 below about 6V and
    will withstand 20V before foward biasing the pin ESD protect diodes on the
    stamp.

    >Anyway, a suggestion came back to use a bridge rectifier for the wart
    >instead of a single diode. The idea being that then it wouldn't matter what
    >polarity the jack is or even if it was AC - everything would work.

    Good idea. Only 'gotcha' is the extra diode drop which is not usually a
    problem. You will also need 2 more diodes for your wall wart detect
    circuit (feeding the resistive divider) - 1n4148 is OK here since the
    currents are low. Or the bridge could feed the 7805 via another diode
    (wall wart sample point is junction of bridge and output diode) - but now
    your total drop is 3 diode drops (~2.1V).

    dwayne

    --
    Dwayne Reid <dwayner@p...>
    Trinity Electronics Systems Ltd Edmonton, AB, CANADA
    (780) 489-3199 voice (780) 487-6397 fax

    Celebrating 19 years of Engineering Innovation (1984 - 2003)
    .-. .-. .-. .-. .-. .-. .-. .-. .-. .-
    `-' `-' `-' `-' `-' `-' `-' `-' `-'
    Do NOT send unsolicited commercial email to this email address.
    This message neither grants consent to receive unsolicited
    commercial email nor is intended to solicit commercial email.
  • ArchiverArchiver Posts: 46,084
    edited 2003-07-16 21:00
    Using Schotkey rectifier diodes can reduce you total drop to ~.45
    volts per diode. 1N5819 is a 1A Schotkey diode and a 1N5822 is a 3A
    Schotkey diode. There leakage current is about 500 uA so keep that in
    mind.

    Jason



    "but now your total drop is 3 diode drops (~2.1V)."
Sign In or Register to comment.