2
\$\begingroup\$

I have no particular interest in power electronics but I am always impressed by such clever circuit solutions as the Greinacher voltage doubler. The idea behind it is so clear and intuitive that it can be explained in a very simple way as I have done in my answer to a similar question.

Greinacher voltage doubler visualized

There are many sources dedicated to this circuit but they (including my answer) consider its operation in steady state (during one cycle of the mains AC voltage). But how does the circuit get to this state after startup? This is the question that concerns me now...

\$\endgroup\$
2
  • \$\begingroup\$ Just use a simulator. \$\endgroup\$
    – Andy aka
    Commented Feb 24, 2020 at 20:51
  • 2
    \$\begingroup\$ I prefer to use my mind as a simulator... \$\endgroup\$ Commented Feb 24, 2020 at 21:05

2 Answers 2

3
\$\begingroup\$

The doubler bootstraps its bias voltages across the capacitors until it reaches a steady-state where the output is double the half-wave amplitude of the AC input.

You can visualize the doubler's behavior using a simulation, right here using the CircuitLab tool.

Here's an example:

schematic

simulate this circuit – Schematic created using CircuitLab

And here's a quadrupler, showing how the stages stack up:

schematic

simulate this circuit

A detail you need to consider is that the diodes have Vf drop, which will reduce the voltage the doubler can reach. I've shown Schottky types (Vf = 0.3v) to minimize this effect.

\$\endgroup\$
11
  • 1
    \$\begingroup\$ Thanks for the simulation; I am happy to see that it confirms my ideas. Would you clarify what you mean by saying, "The doubler bootstraps its bias voltages until it reaches a steady-state"? \$\endgroup\$ Commented Feb 24, 2020 at 21:48
  • 2
    \$\begingroup\$ As the circuit starts up, current from the successive AC cycles is steered by the diodes to produce a DC bias charge on the capacitors. This is 'bootstrapping': the circuit self-starts without outside intervention. The capacitor DC biases 'stack' on each other, pushed by the AC pulses, to produce the higher DC voltage. \$\endgroup\$ Commented Feb 24, 2020 at 21:56
  • 2
    \$\begingroup\$ Not a philosophy, but a style. It’s mostly about word choices: choosing familiar, mechanical, concrete words to represent concepts. ‘Push’ describes how voltage causes current flow (Kirchhoff current law), ‘stack’ for how the voltages add up (Kirchhoff voltage law), ‘steer’ for how diodes rectify, etc. \$\endgroup\$ Commented Feb 25, 2020 at 16:04
  • 2
    \$\begingroup\$ This guy from Czech Republic does an amazing job of explaining electronics. youtu.be/cX4q0e124C4 \$\endgroup\$ Commented Feb 25, 2020 at 16:11
  • 1
    \$\begingroup\$ There is such a paradox that often complicated things are easy to explain... but it is difficult to explain the simple thing... and very difficult to explain the ingeniously simple thing... but it is well worth the effort. Yes, such explanations are detailed and they take time... but that is the price of understanding. The other is easy and fast but it is not understanding... it is knowledge. My goal is to explain phenomena so that they can be really understood. I do it by revealing, step-by-step, the circuit evolution and showing the meaning of each step... \$\endgroup\$ Commented Feb 25, 2020 at 19:23
1
\$\begingroup\$

How to investigate the circuit operation during the startup

The processes in this circuit after the startup develop rapidly over several periods of mains voltage. In order for our slow-thinking mind to understand what happens during this time, we must somehow slow down the circuit operation. For this purpose, I have replaced the AC source with a DC source - a battery that we alternatively reverse (think of it as of a square wave oscillator with ultra low frequency). Thus, we can follow slowly, in a step-by-step manner, the circuit operation.

The figures below represent the circuit state at the end of the half wave when the capacitors (with the same capacitance) are charged and no current flows. Note that here "V" means the voltage of the voltage source (not "volts"!).

The basic idea behind the circuit operation

The basic idea here is a charge redistribution between two capacitors in parallel.

50% charging. In this arrangement (Fig. 1), first we fully charge the capacitor C1 by connecting it in parallel to the voltage source V (a). The source is connected with its positive terminal to ground in accordance with the next figures and is drawn in a more unusual way - below the zero voltage line (ground) and mirrored since its voltage is negative. The capacitor C1 is also drawn below the ground since it is charged to negative voltage.

Fig. 1

Fig. 1. The capacitor C1 is fully charged by the voltage source V (a); then it is 50% discharged through the capacitor C2 (b)... the result is 0.5V voltage across the capacitors

Then we discharge C1 by connecting it in parallel to another but empty capacitor C2 (b). Since the same current flows through both capacitors, they simultaneously change their charges at the same rate but in opposite directions (C1 discharges 50% while C2 charges 50%). As a result, half the charge is transferred from C1 to C2 and half the voltage V is applied across them. We can best imagine this with the help of the well-known hydraulic analogy of communicating vessels. If we connect a vessel full of water to another but empty vessel, half the water will be transferred from the first to second vessel, and the same water level equal to half of the initial level will be established in both vessels.

75% charging. To further increase the C2 charge, we have to repeat this procedure (Fig. 2). First we fully charge the capacitor C1 by connecting to the voltage source V (a). Then we discharge C1 by connecting it in parallel to the 50% charged capacitor C2 (b). Since again the same current flows through both capacitors, C1 discharges with 25% while C2 charges with 25%. As a result, 1/4 of the initial charge is transferred from C1 to C2 and 3/4 of the initial voltage V is applied across them. In the hydraulic analogy, 1/4 of the water will be transferred from the first to second vessel, and the same water level of 3/4 of the initial level will be established in both vessels.

Fig. 2

Fig. 2. The capacitor C1 is fully charged by the voltage source V (a); then it is 25% discharged through the capacitor C2 (b)... the result is 0.75V voltage across the capacitors

...and so on so forth...

So, the voltage across two arbitrary charged capacitors of equal value connected in parallel is equal to the arithmetic mean of their initial voltages.

This conclusion will help us to reveal what happens at the startup of the voltage doubler. For this purpose, we will investigate the first three cycles of the mains voltage; the rest are analogous.

Investigating the circuit operation

Cycle 1 (Fig. 3). To imitate the first negative half wave, we connect the positive battery terminal to ground and C1 in parallel to battery to fully charge it (a).

Fig. 3

Fig. 3. The capacitor C1 is fully charged by the voltage source V (a); then it is 100% discharged through the capacitor C2 (b)... the result is voltage V across C2

Then, to imitate the positive half wave of the first cycle, we disconnect C1 and connect it in series to the source (b). I have shown in my previous answer that we also have to reverse the voltage source so that both voltages V and VC1 to be summed. So, we first connect the negative battery terminal to ground; then we connect the whole "battery" (consisting of the source V and the fully charged C1 in series) to the empty C2. Since the same current flows through both capacitors, they simultaneously change their charges at the same rate but in opposite directions. But here C1 is "lifted" with V; so it fully discharges to zero while C2 charges up to V. As a result, the entire charge is transferred from C1 to C2 and the whole initial voltage V is applied across C2. In the hydraulic analogy, all the water from the first vessel is moved to the second one... and the water level of the second vessel is equal to the initial level.

Cycle 2 (Fig. 4). To imitate the second negative half wave, we again connect the positive battery terminal to ground and C1 in parallel to battery to restore its charge (a).

Fig. 4

Fig. 4. The capacitor C1 is fully charged by the voltage source V (a); then it is 50% discharged through the capacitor C2 (b)... which voltage becomes 1.5V

Then, to imitate the positive half wave of the second cycle, we reverse the source V, connect C1 in series to it and the whole "battery" (V and the fully charged C1) to the fully charged C2 (b). C1 discharges with 50% and C2 increases its charge with 50%. As a result, 1/2 of the initial charge is transferred from C1 to C2 and 3/2 of the initial voltage V is applied across C2. In the hydraulic analogy, 1/4 of the water is moved from the first to second vessel, and the water level of the second vessel becomes 3/2 of the initial level.

Cycle 3 (Fig. 5). To imitate the third negative half wave, as usual, we connect the positive battery terminal to ground and C1 in parallel to battery to restore its charge (a).

Fig. 5

Fig. 5. The capacitor C1 is fully charged by the voltage source V (a); then it is 12.5% discharged through the capacitor C2 (b)... which voltage becomes 1.75V

Then, to imitate the positive half wave of the third cycle, we reverse the source V, connect C1 in series to it and the whole "battery" (V and the fully charged C1) to the 3/2 charged C2 (b). C1 discharges with 1/8 of the initial charge and C2 increases its charge with 1/8. As a result, 1/8 of the initial charge is transferred from C1 to C2 and 7/4 of the initial voltage V is applied across C2. In the hydraulic analogy, 1/8 of the water is moved from the first to second vessel, and the water level of the second vessel becomes 7/4 of the initial level...

...and so on so forth...

For the curious...

The figures above represent the circuit state at the end of the square half wave when the capacitors are charged and no current flows. But it is interesting what happens at the beginning of each square half wave when two capacitors with different voltages are connected in parallel. Is it possible? Will there be no conflict?

Yes, it is possible since there is always some resistance in the loop. For example, in Fig. 6 below, I have illustrated the case when the voltage source has some internal resistance R. The figure represents the circuit state at the beginning of the half wave when the capacitors are not charged to the desired level and maximum current flows. As you can see, the voltage diagram is more complicated... but I hope you will enjoy it!

Fig. 6

Fig. 6. At the beginning of the square half wave, the internal source resistance R limits the current; the voltage diagram is more complicated...

(Fortunately, voltage doublers are supplied by sinusoidal AC voltage; so there is no such a problem there.)

\$\endgroup\$
6
  • 2
    \$\begingroup\$ Depending ESR of C , Vbat and power rating of diode thus ESR, there is no single source of current limit. \$\endgroup\$
    – D.A.S.
    Commented Feb 24, 2020 at 20:00
  • \$\begingroup\$ Agree, Tony... I have considered the case with just one resistance for simplicity. It would be interesting to see a simulation of the operation during the startup... \$\endgroup\$ Commented Feb 24, 2020 at 20:02
  • 2
    \$\begingroup\$ its easy to draw and add voltage and/or current scopes in falstad \$\endgroup\$
    – D.A.S.
    Commented Feb 24, 2020 at 20:21
  • 1
    \$\begingroup\$ "the voltage across two arbitrary charged capacitors connected in parallel is equal to the arithmetic mean of their initial voltages" is not necessarily true. This is exactly how students get confused about how circuits work. Please be more careful in the future. \$\endgroup\$ Commented Feb 25, 2020 at 14:52
  • 1
    \$\begingroup\$ And for the same reason your "water level" analogy is incorrect. The hydraulic analogy only makes sense if you have a closed system with no air in it, and you don't rely on gravity as the only means of storing energy. \$\endgroup\$ Commented Feb 25, 2020 at 14:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.