Which voltage multplier circuit is better-Cockcroft Walton, or diy stepup powe

Discussion in 'Electric Bicycles' started by MoRo1212, Aug 13, 2009.

  1. MoRo1212

    MoRo1212 New Member

    Hello forum members,

    I am deciding whether a diode/capacitor CW circuit,or a diy hand-wound step-up transformer is easier,and more practical;I am only needing at most a voltage tripler for my electric pmdc powered bike trailer with dual curry brushed motors-I have magnet wire,T200 toroid,fiberglass tape,some electrolytic caps,hi amp diodes so either way it would be practical to make a decent working voltage multiplier!

    I basically want to know which "VM" circuits that the members have used with good results,and what circuit is more efficient/practical??:confused:

    btw,I have uploaded a few examples of voltage multiplier circuits similar to what I will like to try, although the transformer core I will be using is a toroid design,and not a split-bobbin/E core design-I am leaning toward using the custom hand wound toroid step-up transformer since it should take the "amperes" needed by the Curry PMDC motors !!

    Thanks in advance for a reply!

    Attached Files:

  2. safe

    safe Active Member

    You might fill in a few more details about "why" you need to double or triple the voltage for a Currie motor.

    My guess is that you are starting with some rather large SLA battery that has great capacity, but is limited to 12 volts. You then want to make use of a voltage multiplier to get the performance you want from the motor.

    Efficiency is a major concern... you lose a lot stepping up the voltage and yet since SLA's have a severe Peukerts Effect it's possible that the loss in efficiency caused by the voltage multiplier will be more or less equalized by the reduction in the Peukerts Effect of the bigger battery. With SLA's the bigger the better.

    As for your choice...

    The transformer seems the most straight forward... using the capacitor approach is more complicated, but I think it's actually possible to get better efficiency with the capacitor approach.

    What would be ideal (and I have thought about this before) is to make your controller integrated into the voltage multiplier. That way you can step up the voltage and run your MOSFETS at the same time and possibly do it in a way where efficiency is rather good. (so this makes it a "controller upgrade")

    The "easy" way is to do the transformer and that's probably what your "Version 1.0" ought to be. But if that works well you might think of a "Version 2.0" that integrates the controller with the voltage multiplier.


    Inverters can step up a DC voltage and then convert it to AC which is something that I was just looking at for the Induction motor I have laying around. Inverter efficiency can be up in the 90% range.


    The biggest problem is that DC cannot directly be Stepped Up in voltage wihtout going through an AC intermediate step. Whether you go with capacitors or transformers you still need to introduce some chopping action in the middle of the process. This is why integration with the controller could work the best because the controller is already chopping the current up into (PWM) Pulse Width Modulation chunks. If you could intercept the chunks that are going through the MOSFETS and do the voltage multiplication there you could attain the goal and reduce the overall losses and redundant processes. However... this is some complicated stuff that most regular people are not going to be able to figure out in less than six months.
    Last edited: Aug 14, 2009
  3. heathyoung

    heathyoung Member

    Forget CW - you will incurr massive losses in the diodes for starters - only really suited for low powered devices.

    You need to have a look at the design of power supplies for car amplifiers - these will give you an idea of your power requirements and topology - you could get away with a non-isolated boost converter, but you require a very bulky inductor (to avoid saturating the coil) - well designed, you could see 85%+ efficiency.