Best boost converter inductor

Candle Power Forums

Help Support Candle Power:

Burnt_Retinas

Enlightened
Joined
May 8, 2003
Messages
396
City & State/Province
Melbourne, Australia
Does anyone have a recommendation for the best boost converter inductor? I have found that the inductor really does play a large part in the effectiveness and efficiency of boost converters, but I'm yet to find one I'm happy with.

Came across this Vishay info today. Anyone tried them?:

http://www.vishay.com/company/press/releases/2002/021108inductors/

ps links to the data are on this page.

Chris
 
Its really more than just "the best one", its proper matching of the inductance, Coil-Q, coil-Resistance and other factors

Its like fine-tuning an engine for peak performance, you add/subtract turns, change the wire-size and others

Also depends on your type of convertor mode: voltage, current, R^2, Ripple, or Cuk'
 
As Inretech said, but the inductor plays quite a large part if it is not matched properly. And anyway, for hobbyists, I think we can simplify it a little.

When I started making my own inductor coils, I had a little Neosid ferrite core. What I didn't realise was that Neosid had specified this core for low frequency applications. The flyback inductor I built was high frequency. Needless to say, there were huge losess due to hysterisis (NOT hysterics /ubbthreads/images/graemlins/smile.gif ). I was disappointed with the results. Only when I switched to one designed for high frequency...

Hysteresis losess are due to the fact that ferrite (or iron) can be magnetised if you apply a direct current to a coil wound around it. But in a transformer or inductor or similar device, the current is AC. So for one moment, the core is magnetised in one direction. The next moment, the core is magnetised in the other direction. It takes energy to counter the original magnetic direction before the core can be magnetised in the new direction.

Also, an inductor delivers its power as the magnetic field it was induced with collapses. What would happen if the magnetic field in your core didn't collapse completely? You wouldn't get as much power as you could, since it is that magnetic energy that is "fueling" the electrical pulse.

These are simple explainations for hysteresis. Also, with regard to design, there are also considerations for eddy currents, core losess, imperfect coupling, etc.

Generally, for power supply designs, you want a ferrite core designed for high frequencies to reduce hysteresis losess, closed loop (toroidal) and without chips or cracks to reduce coupling losess.

Beyond that, you need those complicated maths to work out the size of wire, size of core, type of core material, number of turns or wire in which direction, and even where to place those turns. And that, IMHO, is beyond the realm of hobbyists like me. /ubbthreads/images/graemlins/tongue.gif

I hope that explaination came out correctly and clearly. /ubbthreads/images/graemlins/smile.gif
 
There is more to it than meets the eye as pointed out by others. I have the calculations necessary but finding the core data then doing the math seemed labourious. I was hoping for a tried and tested solution - flyback boost coverter fyi. I was hoping for a quick fix ie "here - try this, it works". I'd prefer to call it "efficiency", not laziness, but looks like there's some maths to do, or tinkering.

Bummer.

Chris
 
I was lucky to have a very nice Hall-Effect probe for my digital scope while working at Intel; it allowed me to see the inductor current and adjust the inductor size/wire/etc

The ESR of the output cap is also a part of the equation, it will affect your eff, response time (not really a problem for LEDs), feedback noise and other factors
 
Back
Top