No Script, No Fear, All Opinion
RSS icon Home icon
  • EEVBlog #473 – Microcontroller Voltage Doubler

    Posted on May 25th, 2013 EEVblog 17 comments


    Dave explains the Dickson Doubler building block circuit. a.k.a diode charge pump.
    Use a spare microcontroller pin, some diodes, and capacitors to create a useful voltage doubler or voltage multiplier.

    Be Sociable, Share!
    • Jay

      Very well done, Dave!

      Except … you cannot get to the Moon on 1 mA. Get real. :D

      • huh

        You can get much more than 1 mA, just use bigger capacitors and drive them through a complementary transistor/mosfet pair.
        The technique is old: I just saw something similar in a 1979 Elektor magazine that referried to a RCA app note.
        This approach however does not scale well; beyond a few mA the switching approach becomes the best choice.

    • someone

      With just a reconfig of the components we can also have a voltage inverter, witch is very nice to get negative voltages.

    • tchicago

      I think it is worth mentioning that this solution is RF-noisy as nothing explicitly limits the cap charge current at the moments when the controller pin switches state. The inductor-based boost regulators are better in this regard. The high current charge also reduce the efficiency due to losses in mosfets ON resistance.

    • EP

      Another fantastic post Dave. Just the sort of pro tips us young players need!

    • http://wardyprojects.blogspot.com wardy

      This is really interesting, I was planning to mess around with my BeagleBone Black this afternoon, but I think I’ll try building one of these circuits instead.

      I love the way you wrote a chess game for your uWatch – that’s just brilliant :D

    • V_King

      great tutorial.

      it might be useful to make a section of fundamental fridays with all the useful tips.

      Or even better – consider writing a book, I would want one!

    • Ingo

      hahaha !!!
      You have to walk on the catwalk.

      Today, You generated a big, healthy smile in my face – thank you.

    • Unixon

      Nice tutorial :)

      A single thing is missing here is a tip for optimal choice of components for driving a certain type of load. It would be nice hear a bit more about the theory of these circuits in terms of analysis of how their performance (capability of driving loads at certain voltage and current, amount of losses and ripple values, etc) depends on the parameters of a circuit (number of stages, capacitance as a function of stage number, voltage drop on diodes, switching frequency) with some graphical representation of these dependencies.

    • http://area0x33.es Miguel

      I like so much fundamentals friday, Big thumbs up for you ! thanks you so much !

    • Lefuneste

      Great one Dave, fundamentals Fridays are great indeed, but WHAT HAPPENED TO YOUR CEILING LED LIGHTS!! The ambiant light on your whiteboard part is absolute crap!! You look GREY!!

    • septer012

      @5:52 magic happens.

    • Ertew

      Great explanation.

      I have one idea about driving LCD. Original HD44780 should work on 3V. New ‘HD44780 compatible’ LCD IC’s always work on 3V and less. The problem is voltage, which you need to drive liquid crystal inside display’s glass.

      Decision to drive uC from 3V and LCD from 5V can make voltage conflict. Also all chip’s require more current when it’s driven from higher voltage.

      In my opinion best solution is drive LCD from 3V. Next, ‘voltage doubler’ should be used for change +3V to -3V, that goes to pin 3 of HD44780 connector (named VS, VLCD, V0 or contrast), because voltage for liquid crystal is voltage between V0 and VCC

      In these circuit, You also have limited control on LCD contrast by change frequency.
      I think (if uC can measure supply voltage and itself temperature) You can make nice feature by automatically set optimal frequency -> voltage -> contrast of LCD.

    • Janez D.

      Dave,

      I’m right in the process of doing some really low voltage single-AA uC experiments and it fascinates me how wall-clock mechanisms work with only single AA battery. It does have a quartz crystal but what is it that really makes it “tick”? 1.5V and then lower is something that blows my mind. Can you explain that?

      Best regards from Slovenia, EU.
      Janez (John)

      • Ertew

        Janez,
        Typical quartz have its typical circuit and typical voltage, right?
        So just imagine, that 2^15 hz is strange, untypical, low voltage crystal, that still can work below 1V and for work require low power. Also, you can physicaly destruct clock quartz if You connect it to typical hi voltage circuit, by to high amplitude of oscillations.
        Just look at any schematics of ancient VCR, that have two typical quartz circuits inside main processor. For RTC quartz you should found extra two resistors, that reduce energy and voltage applied to crystal. So You can reduce voltage and quartz oscillator must still work.

        Ertew, Poland, theoreticaly UE.

    • Albert

      Went to checkout the uWatch but my browser says there is malicious software on the site? :S

    • Pingback: Video tutorial on using voltage multiplier - Do It Easy With ScienceProg