Royer Oscilators

[riccardo]

Hi,

I have been playing with a Royer type circuit. (I think its actually a "collector resonance oscillator" since there's not feedback winding like in a Royer) and have noticed a couple of versions on-line and was wondering about the differences.

In the attached image;
The top circuit uses two chokes and resonates the LC load.
The middle circuit uses a single choke to resonate an LC load with a centre tapped coil.
The bottom circuit is like the first but with one choke removed. Or as the second with the choke connecting to one side of the LC load instead of a centre tap.

I basically just tried the third version to see what would happen, and it seems to work just fine. Whichever side the choke goes it seems to just work.

My question is what is the advantage of the two choke version (top) when using just one (bottom) seems to work the same?

 

[Tony Stewart]

Skip these and look at the improvements made by James Baxandal (famous for his audio filters), adapted for use today for CFL's dimmers.

you want a symmetrical switch with some deadtime for crossover and feedback or current feedforward with a loaded Q of 10.

**broken link removed**

 

[riccardo]

Interesting but it does not answer my question.

 

[Tony Stewart]

The difference is what happens when it L approaches or saturates with excess AC current over DC , where I think the middle one is preferred with L1 using different ferrite.

 

[Mosaic]

I can't say I have ever heard of Squegging before. Baxandal uses that term quite casually. What is it?

 

[Tony Stewart]

The sound a bird makes with chirps of oscillation in a relaxation oscillator mode.

 

[dr pepper]

The diodes are your feedback, the circuit wont oecillate without them.

I have made this circuit with just one choke, works fine, 2 chokes probably lessens voltage transients on the transistors or something like that.

 

[Tony Stewart]

The diodes prevent major cross-conduction losses and are critical to oscillation like a flip flop.

WHen one FET clamps to ground it shuts down the gate of the other. There is always some latency so cross-conduction is inevitable but limited by the speed of the device.

 

[MrAl]

Hi,

I have been playing with a Royer type circuit. (I think its actually a "collector resonance oscillator" since there's not feedback winding like in a Royer) and have noticed a couple of versions on-line and was wondering about the differences.

In the attached image;
The top circuit uses two chokes and resonates the LC load.
The middle circuit uses a single choke to resonate an LC load with a centre tapped coil.
The bottom circuit is like the first but with one choke removed. Or as the second with the choke connecting to one side of the LC load instead of a centre tap.

I basically just tried the third version to see what would happen, and it seems to work just fine. Whichever side the choke goes it seems to just work.

My question is what is the advantage of the two choke version (top) when using just one (bottom) seems to work the same?

Hi,

Sometimes an advantage comes simply in the form of a lower parts count. If you use 10 resistors for one design and 9 resistors for the other design and they work both the same in the application, then the 9 resistor circuit is preferred. Not only does the cost go down, but the reliability goes up, even if just a little bit.

I would think there would be other things you could look at, like trying to use the lowest gate voltage MOSFETs you can get because it looks like this is a current saturated switch oscillator. That's where the current level has to reach a certain value for one of the transistors to start to turn off, and while it is turning off the power dissipation is greatest.

Back in the day we used a Royer for the bias supply in a solar power converter. It used two feedback windings. I remember one transistor always got warmer than the other. We had multiple windings on the secondary though, for the H bridge transistor drive supply. Ran on about 100 to 200vdc and put out various low voltages like 10v or so.