Regulator as part of a transistor switch

Discussion in 'General Electronics Chat' started by SPQR, Feb 10, 2013.

  1. SPQR

    Thread Starter Member

    Nov 4, 2011
    379
    48
    Howdy,

    I'm building a little gizmo to turn a tiny DC motor off and on with a microcontroller. The motor is just a "signal" to a person, and goes on for about 500msec then is off for a few seconds, then comes on again. PWM is not used/needed, just off/on for a short period of time.

    Here is the circuit:
    [​IMG]
    Motor - 3V 70mA, used as a vibrator in cell phones
    2N3904
    AMS1117 3.3V regulator

    I added the regulator to the circuit so that the motor always sees 3.3V, no matter what Vcc I choose at some time in the future.
    With a voltage divider I'd have to change the value of the resistors.

    Questions:
    1. Is it bad practice to use a regulator as part of a switch like this?
    2. Do you have a different suggestion that allows the motor to always see 3.3V but allow the overall circuit to see more volts?

    Thanks again.
     
  2. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    I can't think of a better solution, unless you were to use a small switching regulator. What is the range of supply voltages you might apply in the future?
     
  3. SPQR

    Thread Starter Member

    Nov 4, 2011
    379
    48
    Nothing too fancy. Maybe 5-12V

    So it's not crazy to use a regulator like that? Good to know.

    I hesitated to use the regulator like that because they are mostly fired up when the power is on,
    and I've never seen them used "off" and "on".
    But I can't argue with success - I breadboarded it and it works.

    Thanks!
     
  4. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Just keep in mind that you might not want to touch it if you use a 12V supply. The power dissipation will be (12-3.3)*0.07=609mW. The part will survive, but you might get a blister.
     
  5. takao21203

    Distinguished Member

    Apr 28, 2012
    3,577
    463
    It's worth considering to build a small buck converter (you need a PNP for that)- using a pulse train from the controller designed for that motor load. So you don't need feedback actually.

    You can save the voltage regulator, and you can do away with the heating as well.

    Just a matter of experimentation. Something between 220uH and 2000uH may work as inductor, and maybe a BC327. The 2n3906 is not so much capable, may work on it's margin, but I do not know. You need a small cap (100u) and a flyback diode as well.

    20 KHz or something would be nice otherwise you will hear a noise.
     
  6. Sensacell

    Well-Known Member

    Jun 19, 2012
    1,129
    266
    Do also consider the quiescent current in the regulator- flows at all times in this circuit, wasting power.

    Another idea is to find a regulator with a SHUTDOWN input and do away with the transistor switch- AND reduce the standby power consumption.
     
  7. SPQR

    Thread Starter Member

    Nov 4, 2011
    379
    48
    Ah yes! The infamous "touch test" - I have more that one scar related to that advanced electronic technique!:D

    VERY interesting idea - an analog of PWM, but it lowers voltage - never thought of that.
    I'll experiment a bit.

    This is actually an important issue because I imagine the entire device being attached to a person and run on a battery. I find from the datasheet that it is 5-11 mA, so not trivial.

    "Shutdown" exists on an LT1129 - very nice.

    -------------------
    Again, excellent ideas.
    I thank you all.
     
  8. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,009
    1,530
    Would you even need a regulator? DC motors are usually very forgiving to over voltages. Other than turning faster. For the length of time that the motor would run vs the time it is off, you would probably generate less heat in the motor than in the regulator.
     
  9. SPQR

    Thread Starter Member

    Nov 4, 2011
    379
    48
    Certainly a possibility. I ran it at 5V with no problems, and since it will be on for such a short time it may not damage the motor.
     
  10. ramancini8

    Member

    Jul 18, 2012
    442
    118
    A resistor would do as well as the regulator because it is a single use. Change resistor for supply voltage change.
     
  11. crutschow

    Expert

    Mar 14, 2008
    13,006
    3,232
    Another thought: You could connect the motor from the transistor emitter to ground. If the microprocessor output is 5V, then the emitter voltage across the motor will be about 4.3V. You could add a rectifier or two in series with the motor to further reduce the voltage (each diode will drop about 0.7V). Note that the transistor (and diode) will now be dissipating the power.
     
    SPQR likes this.
  12. SPQR

    Thread Starter Member

    Nov 4, 2011
    379
    48
    Yes, that voltage divider might be useful.

    This is one of those example of using "internal resistance" to control voltage/current outside of the transistor.

    I've seen that concept on the forum for LEDs, where you may not need a current limiting resistor in particular situations.

    The diode idea is also interesting. I read about that in the context of rectification of power supplies.

    Very nice. Thanks!

    PS - Do you always consider the words "rectifier" and "diode" as equal?
     
  13. crutschow

    Expert

    Mar 14, 2008
    13,006
    3,232
    Technically a diode refers to a electronic device that only conducts current in one direction whereas rectifier refers to a device that converts (rectifies) alternating current to direct current (it could even be mechanical such as occurs in a DC generator commutator). But commonly rectifier refers to a semiconductor diode that carries a higher current rating than a small-signal diode.
     
    SPQR likes this.
Loading...