DC voltage regulator

[zacsophonist]

A project I'm building takes 12V DC, and gives both 12V and 5V outputs. The 12V are barrel connectors x4, and the 5V are USB A connectors x4 (no data pins).
I want to offer 2A per barrel and 3A per USB.
From what I've read, to regulate the 12V circuit I need a bulk capacitor, smoothing capacitor, 0.1uF capacitor, switching regulator, and another 0.1uF capacitor before the output. Somewhere in there would be an LC filter for noise.
I would then use these 12V-5V 4A bucks (the extra amp is for safety), one for each USB port.
Am I correct that the total watts I need to regulate are 12V*2A*4 + 5V*4A*4 = 176W (translating to ~15A at 12V)?
I'm struggling to pick an adequate switching regulator.

Please do let me know if I'm going about this all wrong, I've been going in circles with my reading.
I've learned that a linear regulator is not a MOSFET, and that it wouldn't work for my 12V-12V application, since it naturally acts as a buck.

Thank you,
Zac

 

[betwixt]

Your calculation is almost correct but you also have to take into account that the regulators are not 100% efficient and will lose some power as heat. This means you have to budget a little higher to allow for the losses. For switching regulators, which should be fairly efficient, add at least 10% to your power requirement.

However, something doesn't quite seem right, you say "takes 12V" so are you just filtering the existing input or are you trying to regulate the 12V outputs within your circuits? If you need to internally regulate the 12V and you are starting with a nominally 12V source you need a buck-boost design that can increase as well as reduce the input..

Also check the ratings of your USB connectors, some are not rated to carry 3A.

Brian.

 

[FvM]

What's the input voltage range, e.g. 10 - 14V? For 12V output you need buck-boost topology, best performance is achieved with 4-switch converter like this one https://www.ti.com/lit/gpn/tps63070.

As for your input power calculation, consider converter losses with an efficiency factor, e.g. 0.85 - 0.9.

 

[zacsophonist]

However, something doesn't quite seem right, you say "takes 12V" so are you just filtering the existing input or are you trying to regulate the 12V outputs within your circuits? If you need to internally regulate the 12V and you are starting with a nominally 12V source you need a buck-boost design that can increase as well as reduce the input..

Sorry for the confusion, my terminology isn't yet up to par. The idea is a 12V 50Ah battery will power my project via a barrel connector. From what I understand, I have to filter this input and regulate it so that everything downstream gets a steady 12V.
How does this LM5175 look as a buck-boost option? It is rated for 20A output (I need just under 15A).
Another option would be this LM5155. It's ~1/5th the price, though only rated for 15A, so I could just run two circuits in parallel?

What's the input voltage range, e.g. 10 - 14V? For 12V output you need buck-boost topology, best performance is achieved with 4-switch converter like this one https://www.ti.com/lit/gpn/tps63070.

As for your input power calculation, consider converter losses with an efficiency factor, e.g. 0.85 - 0.9.

Yes, around 10-14V; it's a 12V battery connected to a solar setup.
Looking into the 4-switch converters, I will need more amps than the tps63070, so I'm leaning towards the LM5175 posted in reply to betwixt.

Zac

 

[stenzer]

Hi,

you could also spilt the voltag lines i.e. one buck-boost converter for your required 12 V • 2 A • 4, and one buck converter for your 5 V • 4 A • 4. I assume this will give you a large selction range.

As you already considering switching regulators from TI, have look at their Webbench power designer [1], which might be helpful.

[1] https://webench.ti.com/power-designer/switching-regulator

BR

 

[zacsophonist]

you could also spilt the voltag lines i.e. one buck-boost converter for your required 12 V • 2 A • 4, and one buck converter for your 5 V • 4 A • 4. I assume this will give you a large selction range.

I'm considering this option of splitting. Though I read somewhere that it's best to use a separate buck for each USB 5V output. Is this accurate?
If that's the case, I could split the input into two 12V lines, then buck to each 5V from one of the 12V lines.

 

[stenzer]

Using individual 5 V buck-converters may be favourable due to several issues i.e. short current scenario. If you are using four buck converters each one with a adjustable current limit prevents an interference of the supply voltage of the other 3 USB outputs, if one fails short. You may also disable an individual buck-converter, if no load is connected.

In genereal it's a question with respect to your budget and available PCB space .

BR

 

[zacsophonist]

If you are using four buck converters each one with a adjustable current limit prevents an interference of the supply voltage of the other 3 USB outputs, if one fails short. You may also disable an individual buck-converter, if no load is connected.

That's perfect! Thank you.
Going forward with this idea of four buck converters for the 5V outputs, and in terms of regulations,
Must I make and filter two 12V lines from the input?
Or is it okay to make one 12V line, as well as four 5V lines from the input?

 

[zacsophonist]

^^ My concern was with parallel converters, which apparently is not good (without management) if they're sharing the same load.
Since I'll have different loads pulling from each converter, I believe this to be safe so long as there are adequate capacitors before each converter. Will continue digging to double check this assumption.