A two-diode voltage quadrupler

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Very often, one needs a purely static bias voltage: the voltage just needs to be present, but there is no, or negligible current drain.

Think of condenser mics, tube electrodes, varicaps, APD's, etc. for example.

Many solutions exist, but among the preferred ones are voltage-multipliers.
It is well-known that the number of diodes dictates the multiplication ratio.
A quadrupler for example would require four diodes.
But is it really the case?

Here is a challenge to the common wisdom:
This quadrupler ~quadruples the input voltage, but it does so with only two diodes, plus a magic component.

Here is the circuit without the magic component: it is just a plain vanilla doubler, with the input cap split into C1 and C2, which changes nothing to the way it operates.
A 1kHz, 15Vpp squarewave is converted to a ~15VDC voltage

attachment.php


Input and output voltages are perfectly coherent.


Now, add the magic component (red question mark): the output voltage is effectively quadrupled.

attachment.php


Can you guess what this miracle component is?
Hint: it is nothing complex or fancy, certainly not a resonator or even an inductor, simply one of most basic dipole possible...

I will give the answer in my next post, since the forum doesn't seem to offer the spoiling option..
 

Attachments

  • Vquadr1.png
    Vquadr1.png
    55.6 KB · Views: 446
  • Vquadr2.png
    Vquadr2.png
    59.6 KB · Views: 436
PRR got it right of course: it is just a humble resistor

attachment.php


Of course, an inductor works too, as would more complex reactive circuits, which is why I excluded them upfront

When properly differentiated, the 15Vpp becomes 30Vpp.

I found it interesting and paradoxal that adding a parallel damping component in the path of a signal manages to "increase" it.

Of course, you cannot extract lots of power from this oddity, but it could still be useful and spare a pair of diodes
 

Attachments

  • Vquadr3.png
    Vquadr3.png
    77.3 KB · Views: 360
It was an interesting quest.
One thing I do not understand is why it is called a voltage quadrupler when it only doubles the voltage swing of the AC signal. For many years the appended voltage doubler design has been the standard and seems to do the same with two diodes and two capacitors, operating with high efficiency.
 

Attachments

  • VoltageDoublerCircuit.jpg
    VoltageDoublerCircuit.jpg
    11.4 KB · Views: 33
Last edited:
Historically, voltage multipliers were used with pure AC sources (mainly sinewaves, but not always).
Anyway, the average voltage was always zero (transformers, alternators).

This means that the significant parameter, when averaged over a significant number of stages, is the peak to peak voltage halved and the reference voltage (the "foot" of the multiplier) is zero.

Nowadays, when the AC source is a logic gate, we can chose to reference the diode string to Vdd or Vcc, but this option didn't exist in the early 20th century, when these topologies were devised.

It might also seem that asymetrical waveforms could be an exception, but this exception vanishes as soon as the threshold of 2 (a simple doubler) is reached.

Note that if you reference the quadrupler to the +15V supply, you will get 45V minus the losses... a voltage tripler?

This way of thinking doesn't work if you need a negative voltage: according to the usual classification, the quadrupler will work equally well for positive or negative voltages wrt. ground, but the additional help of the +rail won't be present for "enhanced" multipliers having a negative ground.

The traditional way of classifying voltage multipliers is certainly not perfect, but at least it is not too much context-dependent
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.