Flyback Oscillator Operating Frequency

Sounds like a can of worms...

If it's self oscillating, then presumably there is no intentional feed back path?

Obviously the load is feeding back somehow and effecting the oscillator (could just be DC levels)

This is a pretty speciallised area to be tinkering with, and rather beyond the scope of a quick comment. A circuit diagram would help... this isn't my specific field of expertise.

I'd start by looking at DC levels around the ocillator at your various load conditions.

Why are you worried by the frequency shift?

You will need to examine the transformer characteristics very carefully...

The primary inductance will change with the load on the secondary, the same as if the winding was passing a dc current, as the self oscillation is dependent on the inductance you need to know where on the magnetic B-H curve you are operating...

Maybe then you can select a more appropriate core material...

John.

That's not fair John...

You have the unfair advantage of knowing what you are talking about!

Del

Do I? ahhhh I think one of my ferrets walked across the keyboard and posted that last message!!

By the way do you know that ferrets luuuuurrve cats *evil grin*

John.

There are two main methods to regulate a switching power supply. You can vary the frequency or you can vary the duty cycle. Some supplies use both methods depending on the load. There really is no easy way to modify the design of the supply as it would not perform as intended were you to do so.

It sounds like your supply is duty cycle modulating but with a fixed off time. As more energy is required it turns on the current to the inductor/transformer for longer effectively lowering the switching frequency. If you need a supply that runs at a certain frequency you might have to contact a manufacturer to get detailed specs on operation, special modifications to meet your specification or design your own.

Yours was more of a useful answer, most likely freq depended due too loading, tho a schematic would be nice to tell why.

You don't need a schematic to tell why, though it would tell you how.

The reason why is that in a switching power conversion process, you must alter the behavior of the circuit in order to pump more power in to get more power out. This is called regulation. If there is greater loading the circuit detects a drop in output and compensates by increasing the time that power is stored in the storage element (usually the core of an inductor/transformer). It can increase that energy store by either increasing the duration that the energy is routed into the storage element or the frequency at which a fixed quantity is stored. The particulars are design choices made by the designer in consideration of the needs and purpose of the design. There of course has to be an intervening period during which the stored energy is transfered to the output and hence the multiple possibilities of relating the two phases.

Thanks for all the answers. I am beginning to understand a wee bit. The circuit schematic is given below. My aim is to extract 250W from the circuit at a frequency of about 30kHz. At present it is 12kHz. What changes do I have to make in the transformer?

Your schematic does not appear in my browser. Generally speaking, you would have to achieve a higher flux density to get the same power through. That probably means a larger core or fewer turns to prevent saturation. But the fact that your frequency shifts with load means you would have to make other changes to inhibit that feature.

Generally I would not advise you attempt to modify your supply. If you don't know what you're doing (and that's why you're asking, right?) then the most likely outcome will be that at some point you will make a minor error and the results will be catastrophic failure. There isn't necessarily a lot of room for experimentation and if you saturate the core the current in the power transistor(s) will likely destroy them. You can't assume that the supply design has adequate margins to just be changed to significantly alter the operating parameters. 30 kHz is a long way from 12 kHz.

If your project is a one of, just go buy what you need. If it is for a production project then you need to find someone who can design to your specs.

The only way to get a useful answer is to have a look at your schematic - there are so many ways to design the feedback that everything else would be and stay guesswork.

Sounds to me like you need the services of a SMPS design engineer. Try this web-site first: http://www.smps.us/ It has very basic design tutorials and links to resources you need for design and production. SMPS design has a lot of subtle choices that five or more years of experience will help to clarify. There is not enough information in your query to allow anyone, regardless of experience, to offer any legitimate definitive answers.