I recently bought an ZVS driver from AliExpress.com It's about this model.L1 and L2 Choke are each 93μH,C1 and C2 are 0.33 μF each.I want to determine the operating frequency of the device.On the output A _ B i connected the primary winding of my transformer which is 132μH,by mathematics the frequency should be approximately 18Khz,but when i measure Frequency at the output A _ B with a multimeter( UT61E )it shows me 11KHz approximately.Also when the primary is not connected to output A _ B The frequency remains the same 11KHz.My question!Whether the frequency depends on the L1 and L2 chokes and the C1 and C2?Whether they form resonant tank circuit?How to approach calculation?What determines the frequency?

enter image description here enter image description here

  • 1
    \$\begingroup\$ Post the full schematic for that board, or I doubt anyone can answer. \$\endgroup\$
    – user16324
    Commented Jul 28, 2017 at 22:10
  • 1
    \$\begingroup\$ Seems like it would be easy (but time consuming) to reverse engineer the schematic from the board and simulate it in spice. \$\endgroup\$
    – Voltage Spike
    Commented Jul 28, 2017 at 23:00
  • \$\begingroup\$ I will try to obtain a scheme from the supplier. \$\endgroup\$
    – Nano
    Commented Jul 29, 2017 at 8:27
  • 1
    \$\begingroup\$ Small thing, Nano. "K" is kelvin; "k" is kilo. "H" is henry; "h" is usually hour. \$\endgroup\$
    – Transistor
    Commented Jul 29, 2017 at 21:16
  • \$\begingroup\$ Please clear up the formatting. Sentences starts with capital letters. Comma and period are followed by space. k is for kilo and Hz is for Hertz. You mention transformer in the text, but the schematic shows none. \$\endgroup\$
    – winny
    Commented Jul 30, 2017 at 10:13

3 Answers 3


Since its a Royer derived circuit, the major component determining the frequency is LC tank itself (C1 & L3 in your schematic). Choke is just used to block AC through it & provide necessary DC current/power. I'll try to simulate it & update this answer.

Edit 1: Here is the simulation of the provided circuit schematic on LTSpice:

LTSpice circuit LTSpice circuit simulation

We can see that the frequency of the circuit is coming around 22 KHz in the simulation.

Edit 2: Here is the simulation of the provided circuit schematic on LTSpice with 200uH chokes:

LTSpice circuit with 200uH chokes LTSpice circuit with 200uH chokes simulation

We can see that the frequency of the circuit is coming around 19-20 KHz in the simulation. So, there won't be much of a change for different chokes.

NOTE 1: Some components are replaced with generic components due to lack of LTSpice library support.

NOTE 2: Simulations are not 100% accurate as they mostly use & give ideal solutions and don't give real life results. Also its very difficult to simulate Royer Oscillator which I personally faced while designing a Royer based wireless power circuit.

  • \$\begingroup\$ Remove the primary And check then frequency on C1. When I calculate LC Gives me the frequency of 17-18Khz When I simulate Frequency is 22-23Khz That's 5 Khz higher It's normal ? \$\endgroup\$
    – Nano
    Commented Aug 1, 2017 at 21:21
  • \$\begingroup\$ Example Change L1 and L2 to 150-200μH And look at the frequency change With primary or not. \$\endgroup\$
    – Nano
    Commented Aug 1, 2017 at 21:26
  • \$\begingroup\$ @Nano More than 10 kHz difference is intolerable. 5-8 kHz difference is okay as per my experience. I have updated the answer as per your comment. Check out the notes too... \$\endgroup\$
    – Nihal
    Commented Aug 2, 2017 at 7:18
  • \$\begingroup\$ Thank you confirmed my simulations I get the same results,all in all I was worried about these 5-8khz differences. It was strange to me I'm learning :) \$\endgroup\$
    – Nano
    Commented Aug 2, 2017 at 15:15
  • \$\begingroup\$ @Nano we can't completely rely on simulation for Royer based circuits. Royer oscillator itself being patented has very less documentation available. When I was working on a Royer based wireless power transfer circuit, I got around 20-25 kHz higher output than the simulated output. It all depends on the tolerance of the components used in the circuit at the end. Good luck. \$\endgroup\$
    – Nihal
    Commented Aug 2, 2017 at 15:49

I know this question is a bit old, but it seems no one hinted at what I will say so, here it goes.

Your version of the ZVS driver is self-oscillating. So if you had a 12VDC input and no inductive load on the output, you will measure the same 12VDC on the output.

The way it becomes self-oscillating is via the inductor you place on the output. The frequency will then depend on the henry value of the output inductor.

Make a few inductors of different gauges and number of turns and place one in parallel with an open lead on each output leg. Make sure they are easily detachable.

Then connect your scope to the open leads with an inductor connected in parallel and you SHOULD, in theory, see different output frequencies.

Just be sure to calculate the natural resistance of each inductor as too low an input voltage and thinner wire with more turns can lead to no practical output via too much resistance.

If you are still working with this driver, try these tests and let me know if they worked for you. Your model looks identical to mine and it worked for me.


This is the first ZVS driver without a tapped-inductor I've ever seen --I don't know the schematic is correct or not, though.

Anyway, most of the ZVS drivers are based on push-pull oscillator (a.k.a. Royer oscillator). The same approach is used also in electronic fluorescent ballasts.

Resonance frequency mostly depends on the LC load yet the supply voltage as well is effective on the frequency. L1 and L2 does not affect (IMO) the frequency. They help ZVS action of MOSFETs, so the output will be pure (or nearly) sine.


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.