MP2481 dimensioning not very evident from datasheet

DC-DC Power Converters
1

I inherited product ownership for a design from an ex-colleague which incorporates a MP2481 LED driver. Input voltage is 12VDC. Connected LED chains consumes approx 260mA at -18.6V.

When I follow the datasheet it states to choose my ripple current to be approx 30% of my max load current. Since my max load current is pretty steady at 260mA that would be 78mA.

The datasheet a states as well how to calculate the inductor and when I follow these formulas I get to an inductance value of 66uH. That is considerable higher than the 10uH the datasheet mentions
"If the calculation of above equations results in an inductance greater than 10uH, a 10uH
inductor is still preferred."

When I, after some browsing, repeat this calculation using the values as used in MPS Publication “AN009, Wide Input Voltage Range White LED Driver Solutions with MP2481” I get the correct results. In this example the load current is 250mA. However, these calculations use a ripple current of 0.6A. That’s such a large ripple that one can even discuss if “ripple” is the correct term to use.

But in the end my issue is that the used inductor (10uH, DCR=400mOhm, Isat=1.4A) is getting quite hot. Replacement with a 4.7uH, (DCR=200mOhm, Isat =1.4A) does not impove things. I definitely could improve things in my physical layout but the inability to determine a correct coil value still gives me the feeling I’m far from being on top of things.

Anyone who can shed a light on how to deal with this ripple current selection?

2

Coilcraft on their website has a really nifty inductor selection tool (DC-DC optimizer.) In this case you are building a buck-boost supply with Vin of 12V and Vout of -18.6 at 260mA. Chuck in your desired inputs and outputs and it spits out a list of inductors and their associated losses and sizes. It is super cool.

I would suggest getting a demo board if you can from MPS.

3

for what it is worth it suggests 15uH.

4

Hello JShannon,
Thanks for putting my attention to this tool. Exvcellent to say the least. It also reminded me I could have started up Wurths Red Expert which serves a similar goal.

With respect to my heat issue this tool does confirm what bench testing showed as well; Use a physically bigger inductor (meaning, among other things, a lower RDC) which has enough cooling capability on its pads.

All above serves as a solution but I’m still interessted in what MPS has to say about their very unclear calculation.

5

well good luck with that. They focus on the big guys and give good support to them, the whining of small “Freds in sheds” just adds up to a low background hum.