dc-dc converter design for LED driver

[Meri96]

Hello everyone, I'm working on led driver design for the endoscopy system. Circuit sizes should be as small as possible. I will work at a switching frequency of about 2 MHz. My LED specs are 2.82-3.15 V forward voltage at 1050 mA current, maximum power 10 watts. my input voltage is not clear yet. I am researching a dc-dc converter circuit topology where I can operate at high efficiency, high frequency and small dimensions. This will be my first application in power electronics, Your advice is very precious to me.

 

[KlausST]

Hi,

there are many IC manufacturers providing solutions for what you want. Just do an internet search to find manufacturers, go to the manufacturers internet site and use the interactive selection guide.
Very important: Read the datasheet very thoroughly - especially the design considerations. Many mnufacturers provide application notes and/or design examples. Read through them, keep to them as close as possible.

Detailed recommendations may follow after you did your first design trial.

Klaus

 

[zenerbjt]

linear.com buck led driver.
(now analog.com)
What is your available input voltage?
If choice, then use say 12V and use a sync buck led driver.

--- Updated Dec 5, 2020 ---

wat about LTM8040

--- Updated Dec 5, 2020 ---

LT3474

--- Updated Dec 5, 2020 ---

LT8613 is synchronous and very efficient for you.

 

[Meri96]

Thanks for your suggestions, LT8613 seems suitable for my project. I decided to use batteries for my work. If I use a 12 volt lithium-ion battery, the buck converter and the appropriate key driver will do the trick.

So what is your opinion on using the resonance converter in LED drivers? I thought using LLC converter might be good at providing soft switching, but as I said, I have no experience in this area.

 

[zenerbjt]

for 12v you would need four Li ion batteries.
For a 3W load , and such low voltages, you would never use a resonant converter. Best efficiency would be from a synchronous buck converter.

Resonant converters increase conduction losses. Switching losses are minimal at such low voltage that you are using.

Do you want to try and drive the led off a single li-ion cell instead.?...just to make it cheaper?

 

[Meri96]

It doesn't make sense to use resonant converters then, got it, thank you.

Considering the size, using a single li-ion cell would be better than using 4 batteries.

 

[zenerbjt]

ok ill look for a led driver that can go down to 3v input...so it'll be a sepic or buckboost now.

--- Updated Dec 5, 2020 ---

..try the LTC3454...or the LT1618 as a SEPIC

--- Updated Dec 5, 2020 ---

Here i give LTspice simulation of one cell led drivers for you...one with LTC3454 and one with LT1618.
Also is LTspice quik start guide for you

--- Updated Dec 5, 2020 ---

and anyway...resonant converters are not all theyre cracked up to be......they do nothing to reduce conduction losses...they only reduce switching losses, and often only at a sweet spot of electrical conditions.

 

[Meri96]

Thanks for your suggestions and posts, I will consider them all

 

[crutschow]

Note that, for best efficiency, you want the converter to operate in a constant-current, not a constant-voltage mode.

 

[zenerbjt]

maybe monolithic power systems or ti.com have some cheaper chips...worth a look.
Diodes.com also do some cheap low power led drivers.

 

[Meri96]

actually I'm going to do this as a project at university and my teacher wants me to use LLC converter, Unfortunately he didn't accept my other suggestions

 

[zenerbjt]

LLC converter is no good for you....your spec is vin = 3-4.2v, vout = 3.2v and iout = 1A.
No way is LLC any good for you...too many components.
Resonant converters ONLY reduce switching losses...and you dont have much switching loss.
Not only that, but LLC converters dont do well as current sources , and a led driver needs a current source....but admitteldy it could have components added to make it workable..its just not worth it.

LLC also needs a transformer.......and that'll make your circuit too big..........unless you design your own core which will cost a fortune......or use an integrated transformer, but that would be a custom build and cost a bomb

For 3w of led light from 3.5V input, no way an LLC converter.

Does your tutor know what an LLC converter is?

Ask him this question........if an half bridge LLC is operated at the upper resonant frequency with a 1:1 transformer...and a vin of 100V, then what is vout?...the answer, without needing calcs, is 50V (minus any diode drops)...ask him this...if he doesnt know the answer, then bin him.

Seriously, LLC?.......What sort of output do you want...split coil and two diodes, or full bridge rectifier?....ridiculous.............if you must have resonant, then why not a resonant non isolated converter.

BTW, do you need isolation......?

 

[Meri96]

Actually, all we need is a small volume high efficiency circuit operating at a high frequency such as 2MHz. I think he wants this topology because his previous work was on LLC converter.

 

[BradtheRad]

This reply doesn't have better advice than the replies so far.
Anyway here is a simplified concept of an LLC converter. The op amp automatically switches at the resonant frequency, by detecting zero crossings at the low-ohm resistor. In hardware the op amp drives a half-bridge rated for 2 Amps, or use a power op-amp.

In the MHz range you can get by with small component values. However 1 or 2 Amperes could require, for instance a gang of capacitors.

 

[Easy peasy]

also look up resonant forward converter ....

 

[zenerbjt]

So isolation is not needed, and your vin ~ 3.5V, VOUT=3.2V, IOUT=1A.

Definetely no transformer isolated topology as you say small size is wanted.
So a simple buckboost LED driver is all you need.

Your supervisor is trying to trick you into doing an isolated resonant converter...when you have designed it, they will pull the bluff on you and tell you how mad it is, and how you would have had just as good efficiency with a simple buckboost LED driver.

Truthfully, transformer isolated converter for 3W? when isolation is not needed...you are being tricked....it is a cruel trick designed to make you look silly......sorry but i must tell truth,

Resonant converters increase conduction losses....due to their designed in circulating currents

 

[Meri96]

I will meet with my teacher again, thank you for your advice