Help with unregulated [car] 12v dc —> 5v dc @7.2 amps

[littlepixel]

Hello Electro-tech Forum!

I'm really new to the the whole idea of making something I need with some components and a soldering iron, but lack of an off-the-shelf solution to the specification of a power-supply I need has led me here and I hope I'm a) not asking a REALLY stupid question and b) if someone might be able to help...

What I need to build is a 12v dc —> 5v dc @7.2 amps power regulator. Regulator might be the wrong word so I'll fill in the gaps as to it's use and maybe you can advise on a way ahead for me?

So I have an Apple TV media device - these are usually used as TV set top boxes to stream media, but I have hacked mine to run a full operating system that allows me to basically have a music server controlled headlessly with a smartphone. All that bit is working great; The problem is to run it in my car (Which is the the hope) I'll have to run it with a 12v DC —> 220v AC inverter. This seems incredibly inelegant to me since the unit has it's own built in power supply that then converts back to 5v @ 7.2 amps.

So what I'd like to do is build a new unregulated 12v—> regulated 5v supply that can provide what I understand is a pretty high ampage and that can run in a car trunk or perhaps under a seat and not get dangerously hot.

That's about all the info I can give for now except to say that the ac/dc power-supply built into the unit is joined to the motherboard with a connector that has 4 blue wires and 4 black ones - the dream solution would have a female connector of the same type on its 5v end so that it's super simple to swap out the supply for home/car use.

Is that something that could be done? I can solder and am pretty adept at other things with assembling/disassembling electronics without bricking them, but I've just never built anything from scratch from parts bought from Maplin.

Let me down gently if what I'm proposing is madness. I'm happy to use an inverter for now; was just hoping for something smaller and more efficient.

Thanks for your time reading this. I look forward to any pointers you can offer!

Best regards
Huw (littlepixel)

 

[colin55]

The LM350Tis a 3 amp Regulator adjustable from 1.25V to 33V and costs less than $2.00. You can put 3 in parallel with a 0.1R resistor to allow them the current-share.

 

[OlPhart]

I'll 2nd colin55s' method: simple & complete. Be sure to use lots of heatsink, worst case: ~13.5V - 5V = 8.5V, 8.5V * 7.2A = 59.7W

BTW, what would cause that thing to need 7+ amps? I'd measure it, I must be missing something... <<<)))

 

[littlepixel]

Thanks for the quick replies guys - helpful, though I'll have to have a think about how I'd put it together; and guidance on heat sinks would be useful too!

Re the 7.2 amp rating - this is just the information printed on the internal power supply. In full it says:

"Delta Electronics Inc.
Switching Power Supply
Model ADP:36BF A C.C.: AF
Input: 100-240v~1.0A 50-60Hz
Output: 5v===7.2A"

If that's any help?
Thanks again!

 

[OlPhart]

I'd start by measuring the actual current it uses, under various activities, trying to hit worst case/max load. Then add a margin and build to that power.

As to heatsink: there's virtually no such thing as too much. Just distribute the heat sources and monitor the temperature during initial testing. My rough rule: too hot to hold = not enough, even though that's Way under most semiconductor temp limits.

Does this have an amplifier to drive speakers? I'm still trying to get why it pulls so much current. G.H... <<<)))

 

[littlepixel]

Thanks again; To answer your question - it doesn't power speakers per se - it has audio outs (phonos) but they will go to an amplifier before powering speakers. The box is basically a small no frills computer with innards not unlike a laptop or mini-itx pc - it has an intel processor, fan, wireless and 2.5" hard-drive to power. But maybe still overkill?

I'm wondering now if this supply should be made with switching regulators based on the info on the existing power supply. Also - should I be thinking about adding a failsafe fuse into the mix at the 12v end to deal with any surges before the regulation? (You can probably tell what a newbie I am with this sort of question!

PS I don't have anything that can measure the current / amps so I guess I was being as near to the specs of current rather than dealing in real-world useage stats.

 

[crutschow]

For that high a current I would suggest using a switching regulator as you mentioned. That way you don't waste all that power as heat, which has to be dissipated. A quick google search for 12V to 5V DC converters turned up these units. I'm sure there are other similar ones available.

A 12V fuse is always a good idea.

 

[littlepixel]

They actually look great - if I were to get one of those should I opt for the 10amp or 5amp?

I'm not very clear about amps/current 'pushing or pulling' - ie if it's rated at 10a and the device needs 7.2a then does that mean the device will be force-fed 10a and burn out or just that 10a would be the ceiling of current it could ever conceivably pull and hence just a sort of 2.8a safe buffer zone?
Does that even make sense????

 

[RCinFLA]

Look at these.

https://www.current-logic.com/shop/index.php?main_page=product_info&cPath=4&products_id=8

Don't even think about using a linear regulator for that much voltage drop and current requirement.

 

[OlPhart]

They actually look great - if I were to get one of those should I opt for the 10amp or 5amp?

You'll need the 10, because...

I'm not very clear about amps/current 'pushing or pulling' - ie if it's rated at 10a and the device needs 7.2a then does that mean the device will be force-fed 10a and burn out or just that 10a would be the ceiling of current it could ever conceivably pull and hence just a sort of 2.8a safe buffer zone?
Does that even make sense????

...the latter is correct. But I suspect it doesn't draw anywhere near 7 amps, even at max.

 

[littlepixel]

Thanks so much - getting one of these looks like a great way forward - less soldering for me!

 

[crutschow]

If you measure the current it actually takes, then you could determine if the 5A unit is sufficient. You typically want about a 25% derate margin, so if it drew 3.75A or less you could go with the 5A unit. But for the small difference it price you may just as well go with the 10A unit.

 

[OlPhart]

Thanks for the link RC, 1st I'd heard of them. You can't even heat the iron & gather parts for that price... <<<)))

 

[colin55]

Not only are the 3-amp regulators cheaper, but using 3 of them will allow you to dissipate the heat over a much larger area.

 

[crutschow]

Not only are the 3-amp regulators cheaper, but using 3 of them will allow you to dissipate the heat over a much larger area.

Still a lot of power to dissipate.

 

[colin55]

**broken link removed**Still a lot of power to dissipate.

That's why I suggest using 3 separate regulators.

 

[crutschow]

That's why I suggest using 3 separate regulators.

It still the same amount of total power. It now can be distributed over three heatsinks but the required total heat sink area is essentially unchanged.

 

[colin55]

But the temperature-gradient is totally different.

 

[crutschow]

But the temperature-gradient is totally different.

And how is that important?

 

[colin55]

The junction-temperature on each of the three regulators will be lower than that of a single regulator and there will be much less chance of three regulators failing, than the single high-powered regulator.