Retrotechtacular: Understanding A Voltage Doubler

This very stern looking gentleman is about to explain how voltage doubles work in a plodding, yet satisfyingly thorough manner.

We’re not certain when this US Air Training Command video was produced. Obviously it was used to train servicemen who were responsible for keeping electronics running during war time. We’re glad for that, as they really found just the right balance to present a concept that required some knowledge, but is approachable for even the most basic of new electronics hackers.

The demonstration board shown on the right is the voltage circuit highlighted in the lesson. Here the pointing stick is being used to trace out the circuit function during one phase of the input transformer. The capacitor/diode pairs rectify the voltage, with the capacitors discharging in parallel series to double the output voltage. But how does the variable load (RL) affect the output? This is demonstrated under several different conditions using an oscilloscope to illustrate the change.

The discussion of how the diodes work reminded us of a modern tutorial we just ran across this weekend. It’s a bit bizarre, but explains the PN junction in a different way than we’re accustomed to. In this case you will already need to be familiar with how semiconductors work to understand the presentation.

Both clips can be found below the jump.

[Thanks Katja] for the voltage double tip

PN Junction video found [via Reddit]

28 thoughts on “Retrotechtacular: Understanding A Voltage Doubler

  1. hey, i just watched the “modern tutorial” video and that’s exactly how i was taught pn junctions(minus a bucket load of math), so im just curious what is it you are accustomed to?

    1. yeah, I got bored quickly and started skipping through it (she talks in damn near monotone), everything I saw was pretty standard. significantly more in depth than a lot of explanations, sure, but definitely nothing crazy.

  2. There’s a 2003 Microchip PDF which has a bunch of microcontroller tips’n’tricks, number 10 shows how to get a higher voltage by using a PWM pin with 2 diodes and 2 capacitors:
    http://ww1.microchip.com/downloads/en/devicedoc/40040b.pdf

    I discovered this being used on the DIY scientific wristwatch http://www.calcwatch.com (for some reason Google reports it as being an attack site, yet the only external content appears to be Google Adsense…) where he used the technique to power a 5v LCD display from a 3v coincell.

    I tried it myself with a Picaxe chip, PWM output at 10khz with 50% duty cycle, 0.6v drop down diodes with 0.1uf & 10uf capacitors and tested different input voltages to the microcontroller to see what I got on the output (with no load):

    3.3v outputs 6v
    3v outputs 5.4v
    2.5v outputs 4.44v
    2v outputs 3.47v

        1. Well, this is assuming that the doubler modules all have the same capacitors, diodes, timing, etc. In theory, you can in fact stage them to get something like 10KV from a 9V battery. And I don’t see how it’s trolling to realize that you could run your house for a very small fraction of a second or only a 1w lightbulb on that setup? :P

          Normally, these are called divider-multiplier circuits and not just doublers. For example 3 capacitors in series. Switching power supplies can use this idea trivially in a capacitive mode, although scaling up the wattage is not so trivial. Might want to Google “Buck converter”.

    1. Short answer, no.

      What comes out of the transformer is AC. If you want DC you’d still need to put 4 diodes to rectify transformer’s output. So just keep those 2 caps and 2 diodes and suddenly you don’t have to find a transformer with half the ratio :)

      1. well, yes but won’t this generate a lot more noise than a traditional rectifier + capacitor? Dave from eevblog shown a similar circuit some time ago but he if i remember correctly the example was generating the ac from toggling a microcontroler pin then boost it to power an lcd, in this case we already have ac, we already have a transformer, the only reason i can think to go at it this way is if you can’t change the output of transformer nor get another tap. Am i missing something is this is the idea of the movie?

        1. Sometime, it makes sense (and cents) not to pay for a transformer or one
          with multiple windings.

          Back in the older days when your PC used to a 110V/220V switch on the
          PSU. They actually use full wave rectification for 220V and a voltage
          doubler for the 110V. In this case the voltage doubler consists of the 2
          dioded from the bridge rectifier and the switch simply connects the
          neutral wire to the mid point of the caps.

          BTW The rectified DC is used to run the high frequency DC/DC converter
          inside the PSU.

          So would you pay for a huge transformer or redesign the DC/DC stage just
          for the 110V market?

          In terms of noise, the ripple is like that of a half wave rectification. Nothing
          that can’t be filtered out or handled.

      1. You can also use a trick that light painters are familiar with. Record one scene and then wind back the film and record again with the second scene. It’s a cheap additive blending trick that works wonders. Some film makers were even clever enough to use tinted lenses to adjust the degree of blending.

    1. Haha they’re doing it a perfectly correct way IMO. For the longest time I didn’t understand this concept, and I think everyone who works in electronics should make an effort to.

      For those that don’t know, here’s my understanding:

      It makes sense to think of current as the flow of physical electrons (they actually go from – towards + terminals on a battery)…this is your electron current. There is a surplus of electrons on the (-) terminal, so they want to flow to the positive terminal to balance out.

      The modern way is to think of current as a flow from + to – terminals. This is also correct, but in a different way. I view this method as a movement of positive charge +. As electrons move to the + terminal, the overall flow of positive charge is actually towards the negative terminal….the opposite of the former idea of current.

      1. Is it really the “modern” way? I always thought it was a misconception that some how become mainstream. After all, electrons are real thing and positive charge is not. It is kind of like saying “vacuum” is the force sucking dirt into vacuum cleaners.

        1. Current is the movement of charge. Positive charge is moving in a direction opposite to that of the negative charge. Using conventional current makes writing the circuit equations simpler to me. I showed it to a friend of mine who teaches high school students, and it was easier for them than using the method their Physics book taught.

        2. I thought the same. A misconception that they stuck with because so many of the existing books used it. This is back in 1798 or something.

          It is about time it changed. It’s impressive that doing it the wrong way round has worked so well so far, relativity for ya.

          All they need to do, is decide electrons have a positive charge instead of a negative one. It’s an entirely arbitrary decision, like which way up to make a globe. Electricity makes more sense if it flows from + to -, so just fix the mental model to match the physical one.

    2. AFAIK current is still the movement of electrons in a conductor, yea I know electron has been replaced by charge. IMO the time they realized got it wrong they should have started using what the new science was showing them,that’s the way science is supposed to work we are told. Yes additional science had to be taught when semiconductor active devices hit the scene, but the modern way doesn’t explain the operation of vacuum tube. Vacuum tube operate a lot each day in the homes of those who can afford a microwave oven. [shrug]

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.