So today I had to do what nobody should ever have to do... Clear out the workshop!! I've always found it very hard to let go of things because you never know, you might need it one day!!!

Anyway, among the bags which are going to the tip I've got a ton of High Frequency transformers that I've taken out of computer SMPS. I'm not really sure why I kept them but you know what it's like...

My question is, is there anything I can do with them? Maybe make a circuit to drive them? Do they have any other use? I'd rather have a tinker with them or do something fun with them before I throw them out?! At least give me an excuse to keep them!!?

 

make an inverter or flyback recovery psu,

 

Start by thinking of, "normal" transformers. Big iron cores are necessary to contain the energy of a magnetic field at 50 Hz. You put in the right frequency and watch the transformer output a different voltage at the same frequency.

Now, a high frequency transformer. Small ferrite core because the frequency is so high that they don't need pounds of iron to contain the energy of the magnetic field for 1/50th of a second. More like 1/50,000th of a second. You put in the right frequency and watch the transformer output a different voltage at the same frequency. Ta-daa! Are you prepared to build some high frequency drivers for them?

There are also ways to use inductors for frequency filters or, "inductive kick" high voltage generators (boost converters). You can do that with whatever kind of inductor, but the high frequency transformers have such a limited ability to contain energy that you simply must use them in the right frequency range. It's all same, just in a different frequency range.

 

Aghh I see, very interesting stuff. Just to be clear, inside a computer psu, there are a few different sized hf transformers... do they all work at the same frequency or do they vary due to their size? How do you find out/ know the frequency that they are suppose to operate at?
And yeah, I'm absolutely up for making some high frequency drivers!

 

make an inverter or flyback recovery psu,

There a various uses for SMPSU transformers used back to front - some of them not permitted on this forum.

 

Aghh I see, very interesting stuff. Just to be clear, inside a computer psu, there are a few different sized hf transformers... do they all work at the same frequency or do they vary due to their size? How do you find out/ know the frequency that they are suppose to operate at?
And yeah, I'm absolutely up for making some high frequency drivers!

Everybody chooses their own frequency. The sizes mostly change depending on the amount of power needed to transfer. Measure the frequency of the power supply before you take the transformer out or you're in for a re-design that takes you all the way down to measuring inductance and permeability.

 

Aghh I see, very interesting stuff. Just to be clear, inside a computer psu, there are a few different sized hf transformers... do they all work at the same frequency or do they vary due to their size? How do you find out/ know the frequency that they are suppose to operate at?
And yeah, I'm absolutely up for making some high frequency drivers!

A typical ATX PSU has a main half-bridge transformer and a completely independant transformer for the AUX+5V rail, they're not synchronised and run at whatever frequency the designer decided was best. There's usually a driver transformer so the control chip can drive the half-bridge transistors - this obviously runs the same frequency as the main transformer.

A tiny minority of PSUs use a feedback transformer instead of the usual opto-coupler.

 

Measure the frequency of the power supply before you take the transformer out or you're in for a re-design that takes you all the way down to measuring inductance and permeability.

Opps.... I'd better get googling on how to measure the inductance and permeability then!!

 

A typical ATX PSU has a main half-bridge transformer and a completely independant transformer for the AUX+5V rail, they're not synchronised and run at whatever frequency the designer decided was best. There's usually a driver transformer so the control chip can drive the half-bridge transistors - this obviously runs the same frequency as the main transformer.

A tiny minority of PSUs use a feedback transformer instead of the usual opto-coupler.

That's good to know. I'll have to take another one apart and try and make sense of it and see what drives what.
With regards to the frequency they run at... surely they have a optimal frequency so its transforms efficiently? Or is it quite a wide range once you get into the Khz?

 

surely they have a optimal frequency so its transforms efficiently? Or is it quite a wide range once you get into the Khz?

Two people just got done telling you that the frequency is unpredictable for any random power supply.
It can be anywhere from just above audio range to a few megacycles per second. It depends on what year the supply was made and who made it.

 

Two people just got done telling you that the frequency is unpredictable for any random power supply.
It can be anywhere from just above audio range to a few megacycles per second. It depends on what year the supply was made and who made it.

Sorry, I understand that. I meant it as " do HF transformers have an optimal frequency in which they transform." But you said in your 1st post " It's all same, just in a different frequency range." So they quite obviously do.


Thanks for all the help.

 

The math for any particular transformer contains the highest amount of current that can be applied without saturating the core (much). When the stimulating voltage (DC) is switched on, it increases current across time, and the amount of current is used to say when to stop. The faster you can switch the current, the faster you can operate and the smaller the core you need.

Now, here is a critical point: Switching speeds have been improving for decades, and good mosfets revolutionized the art. That is why the year of manufacture is relevant. So, how much current, how fast, and what voltage are all interdependent. One transformer can be driven with different voltages at different rates to accomplish different things (within a certain practical range). The, "optimal" frequency range for one core is here to here, and the one next to it has an optimal frequency range from there to there, but you can't tell without measuring, and when you get done measuring, you will have a range of voltages, currents, and frequencies to work with.

 

That's good to know. I'll have to take another one apart and try and make sense of it and see what drives what.
With regards to the frequency they run at... surely they have a optimal frequency so its transforms efficiently? Or is it quite a wide range once you get into the Khz?

For beginning experiments - the AUX+5V transformer is the easiest to get a project actually producing results.

The driver transformer drives a pair of B/E junctions, so the secondaries only have a small number of turns - using it for step up, you either have to use a very low voltage to drive it, or up the frequency to the point where winding capacitance on what used to be the primary becomes a significant loss.

The main transformer is usually wound for half-bridge, but some of the smaller/cheaper/older PSUs were flyback. Generally I've found the flyback variety easier to do something interesting with.