regulation help

Discussion in 'The Projects Forum' started by LuisMVG1, Dec 15, 2012.

  1. LuisMVG1

    Thread Starter New Member

    Dec 15, 2012
    3
    0
    Hey guys , I am planning to build a regulated power supply but cannot get my hands on a voltage regulator, but i do have a few npn transistors can i use them like in the image to regulate the voltage?
     
  2. panic mode

    Senior Member

    Oct 10, 2011
    1,320
    304
    not like that,
    try this

    http://www.rason.org/Projects/discreg/discreg.htm

    Q2 and Q3 are acting as one transistor with high gain (Darlington).
    R2 and Q4 are current limit or protection (can be removed if you like)
    if you need different limit change value of R2 (R2 = 0.7V/current)
    R3,R4,R5 is voltage divider to check if output voltage is really where it should be
    (if not, transistors make an adjustment and make sure the output is correct and stable).
    R4 is used to adjust desired output voltage.
    Q1,D1 are part of the correction circuit.
     
    Last edited: Dec 15, 2012
  3. #12

    Expert

    Nov 30, 2010
    16,298
    6,809
    or like this...
     
  4. panic mode

    Senior Member

    Oct 10, 2011
    1,320
    304
    this too will not regulate voltage (which is his goal). output voltage would vary when load changes.
    and this happens all the time: you turn on an LED or whatever, load changes, in turn voltage across load changes unless 'regulator' is voltage regulator.
     
  5. #12

    Expert

    Nov 30, 2010
    16,298
    6,809
    It will regulate a lot better than the one Luis posted and it is simple enough that he might understand it.
     
  6. panic mode

    Senior Member

    Oct 10, 2011
    1,320
    304
    this is an example of a basic voltage regulator (open loop)
    one in earlier link is closed loop and much more stable
     
    • reg.png
      reg.png
      File size:
      3.1 KB
      Views:
      16
    LuisMVG1 likes this.
  7. #12

    Expert

    Nov 30, 2010
    16,298
    6,809
    Yes. That design is the logical next step of improvement on the one I posted.
     
  8. panic mode

    Senior Member

    Oct 10, 2011
    1,320
    304
    i disagree. the key to voltage regulation is having voltage reference. the one you posted has none. it is equivalent of using a rheostat in series with load (no transistors required). it can limit current but it cannot adjust voltage to a changing load (and all practical loads are changing) and it cannot adjust for changes in input voltage.

    an example (easily tested in simulator or on breadboard):

    using regulator without voltage reference where supply is 9V one adjusts it to get 5V out.
    then you disconnect source (9V) and connect another one which is 24V. what will output voltage be and will the load survive?

    if instead of circuit without voltage reference we used real voltage regulator (such as 7805) in both cases output would still be 5V. the 7805 would get warmer and probably need a heatsink but it would do it's job - provide regulated output.
     
    Last edited: Dec 16, 2012
  9. LuisMVG1

    Thread Starter New Member

    Dec 15, 2012
    3
    0
    I will go with this one, very simple thankyou.
    I have another question, can i put two transistors in parallel so they could handle more current?
     
  10. panic mode

    Senior Member

    Oct 10, 2011
    1,320
    304
    yes, you can, you can put several of them.

    but they will not share the load equally (one may supply 80% of current and other 20%). to help balance the loading, one adds small resistors in emitter circuit of each transistor (say 0.1 Ohm for example).

    usually power transistor have poor gain (unless you use Darlington) so you may need to add one more transistor to increase gain. this way you are not drawing big current from reference circuit (you don't need large power zener, potentiomer).
     
    • reg.png
      reg.png
      File size:
      5.3 KB
      Views:
      13
    LuisMVG1 likes this.
  11. LuisMVG1

    Thread Starter New Member

    Dec 15, 2012
    3
    0
    OK thanks ,i really appreciate the help :D
     
  12. #12

    Expert

    Nov 30, 2010
    16,298
    6,809
    Premise: regulation is always a matter of "how much".

    Assuming that a FWB is providing +10VDC,

    in the "regulator" presented by Luis, the transistor is always saturated until it smokes because there is no limiting restance in series with the base. There is no adjustability and no regulation.

    by changing the configuration to an emitter follower, the circuit becomes adjustable. Regulation is only improved by the ratio that the output voltage is reduced compared to the input voltage.

    by adding a zener diode, the circuit is still adjustable and has some further rejection of the supply voltage changes (including ripple rejection).

    adding active feedback where the voltage difference between the base and emitter of Q2 is without ripple voltage, some ripple rejection exists and the circuit is adjustable. The major weakness in this circuit is the thermal drift (Vbe) in Q2.

    in the fifth regulator circuit, the delta Vbe is reduced by using a differential pair. It is adjustable and has rather good ripple rejection.
     
    Last edited: Dec 16, 2012
  13. #12

    Expert

    Nov 30, 2010
    16,298
    6,809
    This is basically a bump. I did it because I had enough time today to provide a proper response.
     
Loading...