DC/DC power supply design questions

[bryce1]

Hi,

for my current project, I want to design a nice way of converting 12/24V to 5V. When I had to do this in the past, I'd just choose an easy to use, well documented switching regulator like the LM2574 and be done with it.

Now I dont know how to do it for the following reasons:
- the circuit itself draws about 200mA
- when connecting (powering) an RPi, the regulator should theoretically provide 1.4A

I don't know how to design for 0.2A and 1.4A at the same time. Should I just take two different regulators?

Thanks for your help!

[TVman]

This maybe?
http://www.ebay.com/itm/10pcs-LM2596S-DC-DC-Buck-Converter-Adjustable-Power-Supply-Step-Down-Module-US-/131143985835?pt=LH_DefaultDomain_0&hash=item1e88ca6aab

[bryce1]

Thx. The LM2596SX-5.0 is an interesting part, albeit rather expensive.
What inductor would I need that covers both 0.2 and 1.4A?

[TVman]

Albeit rather expensive.

And what does "albeit" mean? Shipping?

[vk6zgo]

It's a fancy word for "even though"!

[rexxar]

Albeit rather expensive.

And what does "albeit" mean? Shipping?

Generally it's a synonym for "although". Ask Google

Back on the main topic, where does that 200mA figure come from? From what I see in the datasheet, the regulator itself should only draw about 10mA.

[bryce1]

The circuit I actually designed draws about 200mA (Ethernet-IC, uC, LEDs). I want to have the option to power a RPi as well, which adds (worst case) 1.2A.

Sorry, English is not my first language.

[theatrus]

Design for your peak current and you will be fine. Why is there a need for two regulators?

[bryce1]

OK, thanks. I am just confused by the fact, that higher current peaks require lower inductor values.

E: So do I choose the bigger inductor that the lower current requires or the small indcutor for the peak current?

[bryce1]

Anyone? I cant believe the only suggestion here is ebay.

[mariush]

Well, you can make cheaper circuits, but the time required to build it and buy the parts just for one would make it more expensive. Those eBay dc-dc converters are cheap for lots of reasons, making them in large quantities is one of them, the debatable origin of the chips is another...

For example, here's something TI's Webbench throws up.. does 5v @ 1.5A and costs about 2-2.5$ if you make just one:

[bryce1]

thanks mariush! I also want to make my circuit in "quantities" :-D (not millions, but maybe hundreds).

I want to ask for the third time:
Why do the requirements ask for bigger inductors when planning for lower peak currents? Would the circuit in your example also work well with just 0.3A drawn?

[TerminalJack505]

I'm no SMPS expert but I think the reason bigger inductors are needed for lower currents has to do with the fact that most switch mode controllers have a minimum pulse width (5% to 10%, for example.)  That is, the switch is never allowed to have a 0% pulse width (completely off.)

Because of this, there is a minimum on time for the switch.  During this time a bigger inductor will have a lower di/dt as compared to a smaller inductor.  This allows for a lower current at the load.  This is also why most controllers have a minimum load requirement.  (Some will provide this minimum load internally.)

Having said that, there are controllers that will turn off when certain conditions exist.  Linear Technology call this Burst Mode^tm.  This might be worth looking into in your case.

There might also be a "happy medium" value for the inductor, in your case.  I'd have to look but the inductor value would likely have to satisfy a couple of functions that depend on things such as minimum on time and minimum and maximum current.  It might be a trick though, since a bigger inductor may not be able to keep up with a larger demand in current.

[bryce1]

and THAT is why I need your help 
how does e.g. an ATX powersupply achieve this? any system designers here?