Anyone know any useful references for multi-resonant buck converters?
I have US4857822A - Zero-voltage-switched multi-resonant converters including the buck and forward type
- Google Patents as a start.
But there's probably other possibilities (parallel resonance?) and I'd like to learn more.
This is also related to LLC converters but I don't yet see exactly how.
Anyone understand this stuff?
David
I have US4857822A - Zero-voltage-switched multi-resonant converters including the buck and forward type
- Google Patents as a start.
But there's probably other possibilities (parallel resonance?) and I'd like to learn more.
This is also related to LLC converters but I don't yet see exactly how.
Anyone understand this stuff?
David
Thank you, looks like some useful information, will take some time to read.
What is your interest in this technique, for audio/hobby or are you familiar with it professionally?
The industrial uses tend to be for very specific niches, like Point Of Load in a CPU or 48 V rack distribution, not quite what we need for audio power.
So it is not possible to just blindly drop in some kind of "precooked" circuit.
That means I really need to understand this stuff and I haven't been able to find a overview to see how the different approaches relate.
For instance the LLC analyses are heavily frequency domain, the multi-resonant converter stuff is time domain.
Fourier may say they are equivalent in a linear system but doesn't help much to understand the behaviour here.
You know a decent overview?
Best wishes
David
What is your interest in this technique, for audio/hobby or are you familiar with it professionally?
The industrial uses tend to be for very specific niches, like Point Of Load in a CPU or 48 V rack distribution, not quite what we need for audio power.
So it is not possible to just blindly drop in some kind of "precooked" circuit.
That means I really need to understand this stuff and I haven't been able to find a overview to see how the different approaches relate.
For instance the LLC analyses are heavily frequency domain, the multi-resonant converter stuff is time domain.
Fourier may say they are equivalent in a linear system but doesn't help much to understand the behaviour here.
You know a decent overview?
Best wishes
David
I just quickly skimmed through your Google Patent link and the others and one observation I have made is that it makes use of the innate physical properties of both active and inactive devices (capacitance ) as a means of helping to run the circuit .
If that's correct then its ingenious , this presupposes that high quality manufactured components must be used as certain component standards are required.
If that's correct then its ingenious , this presupposes that high quality manufactured components must be used as certain component standards are required.
Well observed, you are correct and it seems clever to me too.
Kind of Judo tactics, deflect your opponent's momentum so that it actually helps you rather than hurts you.
I found one presentations that pointed out that the best circuits will exploit all of the stray capacitances and inductances.
That seems a useful rule of thumb to decide if a circuit is worth a closer look.
Your first link does discuss the sensitivity of circuits to component variation and the costs to be saved thereby.
They did achieve lower sensitivity but not very useful for audio (or at least my application), as far as I can see.
But another useful rule of thumb.
Best wishes
David
Last edited:
