book on Royer converter recommendation ?

[Akwasi]

Hi,

I'm looking for a book which explains the details of the Royer converter, any recommendation ?

Thx,
Andreas

 

[schmitt trigger]

If I recall properly, A.I Pressman's book "Switching Power Supply Design" did have a discussion regarding Royer oscillators.

 

[Akwasi]

The book from Pressman I already have, I was hoping to find some additional information in another book,
I just bought the "Resonant Power Converters" Marian K. Kazimierczuk (Autor), Dariusz Czarkowski (Autor), let's see if this one gives me some insight..

 

[schmitt trigger]

Something that I do know, is that for the optimum efficiency, Royer oscillators one must use a high-mu, square hysteresis loop core.

I can't recall exactly the material name for the core... but it definitively wasn't ferrite.

It was the same material used in magnetic amplifier cores. But I haven't seen a mag-amp design in ages.

 

[Akwasi]

I think the original Royer was designed to switch between the transistors at the point where the core is going into saturation, in this case such a core would be useful, the current or voltage on the primary side would be a square wave, but when a parallel capacitor is used then it's a sine wave, but I don't know if the core is also in this case going into saturation, that's one of the things I would like to understand

 

[schmitt trigger]

I think the original Royer was designed to switch between the transistors at the point where the core is going into saturation, in this case such a core would be useful, the current or voltage on the primary side would be a square wave, but when a parallel capacitor is used then it's a sine wave, but I don't know if the core is also in this case going into saturation, that's one of the things I would like to understand

To do so, it must be CURRENT FED, which essentially means an inductor in series with the supply.
In this instance, said inductor is the one that connects on the transistor's emitters.

This is similar to what one does for fluorescent ballasts. The inductor prevents cross conduction losses.

 

[Akwasi]

In my application I'm using a 150uF coil at the emitter of the transistors,
the transformer is coiltronics CTX210659 (20uH/1:100), Cpar=220nF,
so this should be then a "current fed" topology im my case
the circuit is based on the Linear App-Note ANN-118, Fig.11, with the difference that I don't use the LT1534 controller

I'm trying to optimize the circuit, started looking at the high supply current of 120mA/5V without load, I hoped to find in the book some information on the dimensioning and the expected waveforms of the circuit,
normally I shouldn't be concerned much about current without load because I anyway shut it off most of the time, but it could be a hint that there might be something wrong

 

[FvM]

If you supply the circuit with constant voltage then the base resistor wiill set the supply current respectively output power.

 

[Akwasi]

If you supply the circuit with constant voltage then the base resistor wiill set the supply current respectively output power.

Thanks for that input, that is something I should try, I anyway would like to be able to tweak the circuit by an microcontroller, switching between different base resistors could be one option

- - - Updated - - -

btw: here is a picture of the circuit,
(instead of the CTX210609 I'm using CTX210659,
same inductance and turn ratio but different power spec)

 

[schmitt trigger]

normally I shouldn't be concerned much about current without load because I anyway shut it off most of the time, but it could be a hint that there might be something wrong

Excellent attitude! You are correct, a high idling current means that something may not be quite right. The investigation will allow you to decide whether a design change is required or not.

Do you have an oscilloscope? If so, take primary waveform readings with the circuit as is. Save those waveforms and then start tweaking your circuit.

 

[Akwasi]

maybe one problem is that I switch on the FET at the bottom for too long time until the coil at the emitters goes into saturation, (I don't know if also the transformer coil could go into saturation),
actually, the MOSFET and the coil seem to be kind of buck or boost ?

the parallel cap is a cheap one, I'm waiting for digikey to get some good PP cap to try out.

the application for my circuit is to charge a 47nF capacitor in some milliseconds and then I switch it off and wait until the capacitor voltage drops 10% lower

yes, I have an oscilloscope and will try to take waveforms asap