Connect with us

Circuit protection ideas?

Discussion in 'Electronic Design' started by [email protected], Jul 2, 2008.

Scroll to continue with content
  1. Guest

    I am putting together a design for an RGB LED controller using PWM to
    control color and intensity on RGB LED's. I want the controller to
    work with various LED's, so I am using a 4 position screw terminal.
    The controller just supplies 12 volts and switches ground on the 3
    other terminals to effect the PWM. Then, on the LED module, I will
    use the appropriate resistors to work with the 12V power supplied from
    the controller.

    I use an NPN transistor with a current rating of 800mA to switch the
    grounds, and the transistor is controlled by a microcontroller. I am
    trying to add some sort of circuit protection such that the transistor
    will not be destroyed, even if the user of the device does something
    they shouldn't - like shorts the PWM-switched grounds right to power
    (which would happen if the LED wires are not inserted well into the
    controller).

    In my prototype design I tried resettable fuses from Littelfuse, but
    the transistor was still destroyed when I tried shorting it to
    ground. I think the fuses are not fast enough, despite being marketed
    as super fast.

    Can anyone offer suggestions? I am sure there are feedback circuit
    designs out there that would turn off the output if the current rose
    above a given value, or perhaps there is an easier solution out
    there? Board space is at a premium so I am trying to keep the circuit
    as physically small as possible.

    Any input is appreciated!

    Thanks
    CJ
     
  2. Frank Buss

    Frank Buss Guest

    Did you use the right one? Tripping current is much higher than hold
    current. E.g. if you use the 2920L030, which has a hold current of 0.3A, it
    needs up to 3 seconds @1.5A tripping current. And the min resistance is 1.2
    Ohm, which means there could be high current spikes, e.g. 100ms with 3A,
    see diagram on page 2:

    http://www.littelfuse.com/data/en/Data_Sheets/Littelfuse_2920L.pdf

    I would suggest to build a constant current source. If space is a problem,
    there are some nice LED drivers from Maxim, with integrated output
    short-circuit protection:

    http://www.maxim-ic.com/quick_view2.cfm/qv_pk/5755

    Digikey doesn't have it, but you can buy it from Maxim:

    https://shop.maxim-ic.com/storefron...vent=PartSearch&menuitem=PriceAndAvailability
     
  3. Jamie

    Jamie Guest

    The transistor may be just fine how ever, you didn't specify in which
    manner the device is destroying it self?

    So allow me to come up with a scenario.

    Assuming you're using a open-collector (CLC), did you put in the
    circuit a resistor in series in the emitter for example to limit
    the current? Selecting a resistor to allow for max current when the
    transistor is in saturation is normally what I do, others may do it
    other ways.
    Also, did you account for wheeling voltages that form from arc and
    inductive loads? A unidirectional TVS diode works very well for this.

    It'll protect the transistor for reverse states and for over voltage
    states. Selecting one that is very close to the normal operating voltage
    of the transistor is normally a good idea.

    That is my contribution to this subject be is useable or not.

    http://webpages.charter.net/jamie_5"
     
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

-