Flyback Converter

[whatty]

Recently while working in flyback converter i noticed that AC-DC converters using switching frequency in range of 10 - 100KHZ Ac-DC converters rated for universal supply input ( may be not all but most)
But DC-DC converters working in voltage range 70-30V DC uses switching frequency more than 100KHZ why...?
Does that anything do with impedance of inductor XL=2*phi*F*L or something else because in most application notes the switching frequency is simply assumed there were no reason mentioned for selecting that frequency.

[moffy]

Generalisations are often misleading or false, please provide some detail to support your statement that will make it easier to answer.

[T3sl4co1l]

Generalize what? It's the FCC Part 15 lower limit of 150kHz, operating below lets you skip the fundamental so you only need to worry about harmonics.  Easy couple-dB improvement in EMI.

Lower voltages are also easier to switch faster, give or take.  Not "lowest", mind; 30-200V is probably kind of a sweet spot for Si devices, hence the ratings of most RF amplifier transistors being around here.  (There are RF transistors up to 1200V rating, but they have lower Fmax than others.  Typical commercial radio installations use a ton of 48V modules, is my understanding.  So, transistors in the 100-200V (Vds(max)*) range are very common in that service.)

*RF parts are often rated in DC supply ("design center ratings" system), not peak voltage ratings, so these might be rated 50-100V, but actually break down over twice that.  Go figure.

Probably the most common exceptions are resonant supplies in the 300kHz+ range, or with GaN devices (in any topology) which can push low ~MHz.

Tim

[moffy]

Generalize what? It's the FCC Part 15 lower limit of 150kHz, operating below lets you skip the fundamental so you only need to worry about harmonics.  Easy couple-dB improvement in EMI.

Lower voltages are also easier to switch faster, give or take.  Not "lowest", mind; 30-200V is probably kind of a sweet spot for Si devices, hence the ratings of most RF amplifier transistors being around here.  (There are RF transistors up to 1200V rating, but they have lower Fmax than others.  Typical commercial radio installations use a ton of 48V modules, is my understanding.  So, transistors in the 100-200V (Vds(max)*) range are very common in that service.)

*RF parts are often rated in DC supply ("design center ratings" system), not peak voltage ratings, so these might be rated 50-100V, but actually break down over twice that.  Go figure.

Probably the most common exceptions are resonant supplies in the 300kHz+ range, or with GaN devices (in any topology) which can push low ~MHz.

Tim

Good points, I was focused on the flyback topology, missed the obvious.

[whatty]

so you are saying that lower voltage ranges 20 to 200V are easy to switch more than 100khz than AC - DC supply of 300v hence most fly-back AC-DC converters operate at maximum of 100khz am i right

[T3sl4co1l]

There's more than that, still; high frequency is still a bit more expensive, but it depends.  Mainly when you need high power density, so smaller capacitors and inductors can be used -- but higher quality parts are required, too.

So, cheap commodity units might be made at low frequencies, with old controllers (TL494 is still around..), and not be particularly compact, but they don't care.  Or... they skip the EMI filter, because... they don't care.

The ultimate capability of Si MOSFETs are more of a limiting factor, one among many.  You can use them up to some MHz with high efficiency, 100s of MHz even, but it's increasingly difficult (limited topologies, supply/load ranges, expensive RF types?).  You're probably not going to go that high for a general purpose power supply, let alone one that needs to sell competitively in large quantity.

Engineering is an often multidisciplinary art, that aims to find solutions to problems that contain many, often poorly defined, constraints.  Cost, size, electrical ratings, environmental and performance ratings, etc.

The constraints listed above are merely among them; they might not even be relevant when other constraints dominate the problem.

Tim

[Someone]

so you are saying that lower voltage ranges 20 to 200V are easy to switch more than 100khz than AC - DC supply of 300v hence most fly-back AC-DC converters operate at maximum of 100khz am i right

Go and calculate a figure of merit against a range of MOSFETs for different voltage inputs.... then come back to us with the results.