High Current Adjustable Power Supply Circuit using LM317

Variable power supplies are available in different ranges but they provide very low current of one or two amperes, they can’t drive high power devices like motor or high power lamps.So to get through such problems I introduce a high current variable voltage power supply circuit which can drive 10A of load, the current handling capacity can be increased simply by connecting 2N3055 power transistors in parallel.

I happened to design variable power supply schematic diagram for my inverter battery for safe and quick charging. My 30Ah lead acid inverter battery takes more time for constant voltage charging when I use a LM317 variable voltage regulator. This is due to its lower output current.

So I connected a power transistor which helps to increase the current without varying the voltage. 2N3055 can handle 15A, so by connecting parallel make add the individual current.

This can be used as a regulated 48V DC power supply or 24 V DC power supply. The maximum input to this DC variable power supply is 37 volts.

High Current Variable Power Supply Circuit Diagram

Variable power supply circuit

Components Required

  1. IC LM317
  2. Transistor 2N3055 X 3
  3. Diode 6A4
  4. Resistor (220 ohm, 4.7K(POT))


  • As you know LM317 is a variable voltage regulator IC. Output voltage of this IC can be varied between 1.25V and 37V.! Yes, this is the best choice of variable power supply circuits.
  • Output of LM317 variable voltage regulator in connected to the base of 2N3055 power transistor. We have already discussed LM 317 Variable DC Power supply, but it was low current.
  • Here the first section of circuit is same as before but in order to boost current we are using 2N3055 transistor.
  • 2N3055 power transistor can handle a maximum of 115 watts and the maximum collector current is 15A.
  • In this circuit the output from LM317 directed to parallel combination of 2N3055 power transistors which will increase the output current hence the power also.
  • Collector of 2N3055 connected to Vcc and Emitter of each transistor are looped together to get  output terminal.
  • By varying the resistance of potentiometer output voltage of LM317 IC can be varied. Correspondingly the emitter voltage of power transistor 2N3055 is varied.
  • Proper heat sink must be provided for the power transistors.

I used these to charge my 30Ah lead acid inverter battery. This circuit is tested under i-St@r labs.
High current variable power supply
Buy this Adjustable Power Supply from here.

You can buy this circuit from our Circuits Gallery Store. All the circuits are tested and 100% working. We will be providing 6 Month warranty to all circuit that you purchase from us.


Diode 6A4 pinoutLM 317 pinout
Diode 64A                                    LM 317
2N3055 pinout

42 thoughts on “High Current Adjustable Power Supply Circuit using LM317

  1. There are no required caps in this schematic? How would you go about adding a power indicator led to this?

  2. Whenever i connect load at output the output voltage (at emitter) reduces while the current increases.
    I regulate a 3.3V at output of lm317 but when i connect a load the emitter voltage drops.
    I require a constant voltage of 3.3V and able to draw 3A current for my circuit as m connecting it to a motor.
    its URGENT !

  3. what if I replace the 2N3055 with TIP35, will it make a change in current and when the input is around 15V-20V,what will be the current output?

  4. Dear Sir,
    first of all thnx for this circuit i want a circuit where i have to connect gps and gsm with car battery and output current should be 2A of regulator IC so tell me can this circuit can i use and can i replace lm317 with 7805 ic

  5. At 1v and 10 amps out with 30v in, you would need to dissipate 290 watts. A 2n3055 on a decent size heatsink can only dissipate 30 watts, so yoare going to need 10 of them. The emitters of the 3055s should have a 0.1 ohm resistors in series with each of them ie ten resistors to share the current otherwise one of them gets hotter than the rest and passes most of the current..
    The 220 ohm resistor connected to the adj terminal should have it’s other end connected to the output..
    Needs a fuse in the output.

  6. Hi!
    I want to be able to adjust the power for a motor between 2 and 3.3 volts with a potentiometer, I am thinking of using this circuit to drive the motor. What will happen if I use a power source of only 3.3 volts?
    Thanks in advance!

  7. Hi can this circuit power an 5 amp drawing electromagnet . I mean lm 317 has short circuit protection ‘so when we use loafs like electromagnets it will not give output . So my question is with the above modification will it power even electromagnets??? Thank you

  8. I needed better regulation than LM317 could provide for charging 4 x 1.2V NiCd up to 400mA. So I did not need a 2N3055.
    Instead I used a BUL6802 ( which I had available). Voltage output just varied from 6.26V at 100mA to 6.15V at 400 mA.
    Thanks for the idea/

  9. Hello, I need a little help troubleshooting my test circuit.
    Im testing the circuit and the voltage does vary at the output with the potentiometer.. (Emitter of the pwr Transistor).
    This happens without a load. However when i hang a load the voltage drops drastically and does not deliver current to
    the load, so it does not work, at least not enough to light the bulb.

    Here are my specific data for my application.
    Regulator : LM338 K Test Load : Automotive light bulb 12V 1 Amp
    Pwr Tran : 2N3055
    Input Voltage ; non-regul 15 VDC
    The power source im using is a 120 VAC to 12 VAC rectified. If i connect the light bulb directly to the power supply\ it lights up ok.

    The real purpose would be to regulate the voltage of an automotive fuel pump . I couldn’t find the power
    consumption data of the pump anywhere but i suppose it uses around 8 Amps at 12V. Sine I’m planning
    to basically reduce such a voltage and thus the current for an experiment, this circuit would work okay in terms of power for my application I believe if I get it to work.
    Any help would be appreciated
    Thanks Again.

    • If you are using the circuit as described by the OP, which IMO is not ideal at all, then you can make a quick change that will greatly improve regulation simply by connecting the 220 ohm resistor between the 3055 emitter buss (output) and regulator pin 1 instead of the way it is shown, which does nothing more than hold the voltage at the 317 output constant regardless the load on the 3055’s. In a pass configuration like this those NPN’s actually have to be driven harder as the load increases, and the better and more common high current solution for current assisted 3-tem regulators uses a PNP with the base sinked by the input side of the regulator. There are other changes that would improve things as well, such as removing that unnecessary blocking diode and instead tying the regulator output to the three 3055 bases with three 10 ohm resistors in parallel, but that first change will at least get the output to be regulated.

      • Thanks for the reply. . .
        you say . . “which does nothing more than hold the voltage at the 317 output constant regardless the load on the 3055’s. .”
        –>So this means the circuit will not regulate?

        you say . . “and the better and more common high current solution for current assisted 3-tem regulators uses a PNP with the base sinked by the input side of the regulator.”
        –> don’t quite understand “sinked” could you explain
        –> . . I only using 1 transistor . .do i still need the 10 ohm resistor you propose?
        –> An additional question would be . .what is the diode used for? what was the intention? (you say it is superfluous).
        Thanks, i put this project on hold for lack of info (i went through electronics Academics when younger
        at school graduated as an electronics technician but i never practiced . . i went into the software world)
        . .and i can’t find my notes from school . . i think i still have them somewhere
        Appreciate the help

        • Yes, the circuit as shown will have little to no regulation under variable load. With a static load, for example a light bulb, you would have constant voltage, but add a second bulb and the voltage will drop. Move that resistor to the 3055 emitters instead and the load will be sensed and bases driven harder to compensate.

          The main purpose of the 10 ohm resistor is to limit the current through the 317 and put most of the load on the 3055 drivers, since most of the base drive current essentially goes on to the load as well. The limitation would be if the unregulated source voltage is not high enough for the 317 to supply sufficient drive current through that resistor at the desired output voltage, in which case you would either reduce the value, reduce the regulated voltage or increase the unregulated voltage.

          Best thing to do is go to the TI data sheet on page 16 and look at the high current application example: http://www.ti.com/lit/ds/symlink/lm317.pdf

          The main reason it makes sense to do it that way is that the driver transistors are biased just .7 volts below the source voltage by that PNP pre-driver I told you about (“sinked” by the 317) so you can get the maximum possible regulated output. I also know from experience it just works better and is easier to get working properly.

          • Tom, for my 24V 10A PSU application this circuit initially looked promising in LTSpice simulation. But as you point out the LM317 (substituted with LT317 in my case) is not regulating the emitter bus output. Changes in load result in 20mV-1V output variation. Granted, this is tolerable when suddenly removing a 10A load, but the entire benefit of a voltage regulator is stability, right?

            So, with your comments and advice I found a rock-steady solution. I went back to the high current app example in the TI datasheet and began simulating. As it turns out the datasheet example ripples from startup and easily drops out of regulation. The current amplification circuitry’s resistor values are way too high. By reducing them to values of 1 and 2 Ohms, the performance improves. Those bypass caps also need to be around 0.1uF-1uF to prevent circuit oscillation at startup and recovery.

            Adding another NPN with low value parallel resistors driving their base pins allows the circuit to handle varying load up to at least 20A without output voltage twitching even once.

            Amazing. Thank you for your advice and pointers.

            Can I take advantage of your knowledge and ask why this NPN current amplification works? I’m new to electronics and it seems that the powerful additional current (20A) is being conjured out of nowhere!

  10. I need a 48 volt 15 a power supply. In this thread it is noted that this can handle that voltage but when I click the highlighted statement it redirect me to a dual low voltage circuit .{ regulated 48V DC power supply or 24 V DC power supply}

  11. Design could be greatly improved, IMO. For better regulation you should connect that 220 ohm resistor to the 3055 emitter output buss, as shown there is no direct feedback loop to the regulator and the output voltage will generally be .7v lower under load. It is not important what voltage the regulator is driving the output transistors, only the output itself. Second, instead of that diode you should use three low resistance resistors, like10 ohm, in parallel feeding the bases of the three 3055s. All the current from the regulator driving the output transistors is passed to the load, and depending upon the gin of the 3055s you may be able to increase the value of those resistors (lower the drive current) to allow the regulator to run cooler, it depends upon the ratio between your input and output voltages. A more common version of this circuit uses a 2955 PNP or Darlington connected pair working off the regulator input side. I guess all us old graybeards are dying off these days.

  12. hi
    i want to design 5 volt dual power supply with current raring upto 3 Amp , can anybody tell me how to design it .

Leave a Reply

Your email address will not be published. Required fields are marked *