Connect with us

Capacitor size

Discussion in 'Electronic Basics' started by Jan Nielsen, Aug 28, 2007.

Scroll to continue with content
  1. Jan Nielsen

    Jan Nielsen Guest


    I have a simple circuit with a PICAXE that controls a relay board, when
    the relay board is empty it works fine.
    But when I put a simple 20W lamp on one of the relays (230Vac) the chip
    resets very often when that relay starts/stops.

    I have a 0.1uf on the power legs of the IC, how do I know what other
    size is needed to prevent it from resetting.
    I didnt make the relay board myself, but from what I can see there is no
    caps on the board.

    The relays use 12V where the IC is using 5V.
    The 12V is from the powerinput, the 5V is from a 7805.

    Does the reset pin need a cap?
    What about the extra 0V pin the chip has (Picaxe 28X1, theres a pin next
    to serial connection with 0V, although both 0V and VCC is on the other
    side too).

  2. anirudh

    anirudh Guest

    It would be a really good idea to provide a schematic of your
    circuit.The description is'nt really crystal clear :)
    Are you sure that the PICKAXE can supply enough current to the relay
    board when the lamp is connected?Measuring the current throught the
    lamp should help.
    It is generally a good idea when working with microcontrollers to
    connect the two pins(Vcc and ground) that are provided on the 2 sides
    of the chip to each other.This helps the microcontroller to source or
    sink high current(high as in the maximum values specified in the
  3. Jan Nielsen

    Jan Nielsen Guest

    anirudh skrev:
    Hmm, the picaxe output goes to a transistor (on the relay board) that
    sends the 12V unregulated voltage to the relay.

    The current to the lamp is not running in the circuit, just connected to
    the screwterminals of the relay board (230Vac) (NC and common of the relay).

    I didnt make a schematic before making the board as it was fairly
    simple, but I can draw one up later.

  4. anirudh

    anirudh Guest

    Have you used diodes between the collector of the transistor and the
    relay terminal(cathode towards collector)?
    Otherwise the back current when the relay changes from on to off may
    be causing the problem.
  5. Jan Nielsen

    Jan Nielsen Guest

    anirudh skrev:
    Yes, I brought the relay board finished, and it includes a diode.

    And the chip only resets when theres something on the relay, not when
    turning it on/off with no load.
  6. Jamie

    Jamie Guest

    well then, the device that is being controlled via the relay must
    be getting it's power from the same source as the chip?
    hence, rail dipping..
  7. Eeyore

    Eeyore Guest

    Relays produce EMI (electro-magnetic interference) particularly when switching
    off the load.

    This EMI can 'couple' into your circuitry and cause random resets and the like.
    Changing the value of the supply decoupling capacitor is fairly unlikely to do
    much to affect it. Long pcb traces that form large 'loops' are one of the worst
    offenders for EMI pickup btw.

    How to design for reasonable immunity to EMI is a huge subject in its own right.
    You may however get some inprovement by lowering the impedance of the reset

  8. Chris

    Chris Guest

    I'd agree with Graham. The relay contact is almost certainly arcing,
    causing EMI that's being picked up by the microcontroller (more likely
    through one of the I/O pins, not the power supply), which causes the

    I'm hoping you're using a relay rated for 240V service -- if, not, try
    replacing the it. Relays that aren't made for switching higher
    voltages have armatures that don't move the relay contacts far enough
    apart. That will draw nice, juicy arcing which will cause your Picaxe
    all kinds of problems (and also destroy the relay in short order).

    If the relay's OK, you'll have to look elsewhere to solve your
    problem. Ideally, you'd want to put an inductor in series with the
    light bulb to limit surge current. But a line rated inductor that can
    handle that much current can be pretty expensive.

    Incandescent loads can be tough, because the turn-on surge current can
    easily be eight or more times the rated current. Your best bet to
    start with might be to get a MOV rated for your line voltage, and put
    that across the relay contacts.

    What might also help is adding some series resistance. If you happen
    to have a 10 watt, 100 ohm wirewound resistor in your junkbox, you
    might want to put that in series with the lightbulb, too. That will
    limit the surge current, and also lower the voltage across the relay
    contact with the arc on make.

    Another thing which might help is just putting some distance between
    the relay and your microcontroller. "The
    dilution". Let the inverse square law work for you, and get that arc
    as far away as you can. If your Picaxe and your relay board are a
    couple of inches apart, you might accomplish your purpose just by
    moving things around a bit.

    Good luck
  9. Jan Nielsen

    Jan Nielsen Guest

    Chris skrev:
    I have solved some of the problem with capacitors, its not perfect, but
    alot better than before.

    I added
    1000uf before the 7805
    0.47uf on the picaxe power pins (that already have 0.1uf)
    0.1uf on another mcu component power pins
    0.33uf on 7805 input
    0.1uf on 7805 power output

    It can now switch just about everything I have tried, but a rare few
    times it still resets.

    The boards is fairly close, as they are contained in a medium sized box
    rated for 230V equipment.

    The board is made on vero, the one with 3holes connected, so theres lots
    of unused traces to pickup emi, although they arent directly connected
    to the picaxe.

    I have 2 free ports on the picaxe, both inputs, would grounding them
    help anything ?

    The relays are rated for 230V 10A each and the system is limited to 5A
    total by fuse.

    Maybe picaxe is just very sentisive, because I did the same system with
    a SBC65EC board from modtronix and it didnt have any problems with it.

  10. w_tom

    w_tom Guest

    I would be guessing what your schematic is. Below are some
    techniques and a concept that others have not discussed.

    The relay contact is not isolated from relay coil. Electrically
    there is still a connection (high impedance) between both. You must
    consider that leakage in your design.

    One technique means somehow gets circuits on both sides of the relay
    to share a common ground. Yes, I know that may cause safety problems
    which is why the common grounding must be designed carefully.

    Second, how does the transistor drive that relay coil? For example,
    if coil is connected to transistor emitter, then failure is an
    option. Better designs connect emitter to the power source or ground;
    collector to relay coil. Now the high voltage base to collector helps
    isolate the controller from relay wiper. Again, appreciate that the
    230 volts on that wiper still connects to relay coil and controller.
    That circuit is not obvious - often denied by those who assume relay
    coil and wiper are isolated.

    Because of how relays work, we often would make dry connections with
    other relays, then only power the connection with a special relay that
    make a non-bounce connection. Sometimes the connection is made via a
    thyristor - that can turn on DC power but cannot make the disconnect.
    Disconnect performed mechanically (this assumed DC power). Or even
    better, use thyristors that hold off their turn on until voltage is
    zero crossing (this assumes AC power).

    Another trick is to put a snubber across energizing relay wiper
    contacts. A 'softer' make or break means less energy travels through
    that circuit that connect relay wiper to coil.

    And finally, the driver transistor is driven by a buffer transistor
    so that any currents from emitter or collector of relay coil driver
    would be connected to power or ground via the buffer transistors - not
    enter controller via base (or gate) of buffer transistor. Sometimes
    that buffer transistor is an IGFET so that reverse (isolation) voltage
    gate to drain and gate to source is more than 20 volts. And or we add
    limiting resistors to the controller to base circuit to further
    isolate that controller from the relay coil.

    Only then are we ready to add capacitors (or inductors) as high
    frequency filters to control circuits between controller and relay.
    All this because the relay wiper - especially at 230 volts - connects
    current to relay coil - especially higher frequency currents.

    Just a few solutions that I have used previously off the top of my
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