Need to drop 12-14v to 8.5 at 6 amps

[xplod1236]

I need to make an 8.5v power supply which will provide ~6 amps. The input will be 12-14 volts from a car. What would be the best solution?

I was thinking about using a 7809 regulator, and feeding its output to a buffer (maybe OPA549). I know I said 8.5v and now I want to use 9, but i'm pretty sure that 9v wouldn't cause any problems. Is there a better/simpler way to do this?

[Bazukaz]

Hi,
There is another Chip , LM338 , which provides min. 5A and has adjustable output voltage.

[cliffforrest]

That is still over 30watts constant disippation. ie HOT!

Why not use a switcher?

LM2678 switcher

[xplod1236]

Why not use a switcher?

I want to keep this as simple as possible.

[N-Channel]

Xplod-

That's why it is called the LM2678 SimpleSwitcher IC. In its simplest form, all that is required (besides the chip itself) for a complete cuircuit are a diode, coil and two caps. For the adjustable version (LM2678-ADJ), just add two resistors for setting the output voltage.

I have used the LM2678 for the past several years in a solar charge controller, and it works very well, with no RF hash, either. In testing my circuit, from no-load to full-load, the only time the chip got warm was when I touched it and it picked up the heat from my hand. Not too bad! The chip's relatively high switching frequency (260kHz) allows the use of small capacitors and magnetics to reduce the size of the overall circuit.

The datasheet is very comprehensive, and even shows how to select the inductor, caps and diode. Nuts-n-bolts procedures, and not too much abstract theory. Go to the link cliff suggests and download the datasheet, and you'll be suprised how easy it really is.

Hope this helps,

Steve

[clem_o]

Originally posted by xplod1236


I want to keep this as simple as possible.

Keeping 30W of dissipation "cool" is not simple, go with the switcher.

Cheers!

[xplod1236]

Okay, I'll give it a try. Thanks for the suggestions.

[rpapps]

From the link given above, the maximum output current is 5 Amps.
Try this one instead.

http://www.national.com/pf//LM/LM2742.html

Cheers
Rob

[N-Channel]

Xplod-

Braincramp on my part! I totally forgot about the 6A load you have, and the LM2678 is good for only 5A.

Rob,

Thanks for pointing out the 5A limitation. I have looked at the link provided. Question: Does this come in a PDIP pkg, or just a SMT pkg? SMT might make things a little more complicated for him.

Xplod-

One PWM chip that stands out is the TL494. You could use it to drive a BIG PNP transistor in a PWM fashion, and get good to excellent results, typically in the 70-90% efficiency range. Go to this link:

TL 494: http://www.onsemi.com/pub/Collateral/TL494-D.PDF

Check out Fig. 21 on p.11 of the TL494 Datasheet. Making the following changes to the circuit shown should yield your desired output:

1) Change the TIP32 (3A) to a bigger PNP unit, like a TIP36 (25A), thus meeting your current requirements;

2) Change the MR850 to a higher current schottky diode, like a MBR1645 or equivalent, thus also meeting your current requirements. Using a Schottky over a regular fast-recovery diode will boost efficienct by a couple of percentasge points;

3) Add a 6.2k W from pin 1 of the TL494 to
ground (this will change the output voltage from 5.0V to ~9.1V);

4) Change the 0.1W power resistor to 0.01W. This will keep the curren-limit feature from kicking in until ~9-10A, which should meet your current requirements, and;

5) Change the inductor's windings from #22 AWG to 3 strands of #20 enameled magner wire, wound tri-filar. Using several smaller wires will enable you to more easily wind the coil. Or, you can just find a 1.0mH coil rated at 10A.

That's it. This schematic is basically the same one used in most cellphone cigarette-lighter adapters, only scaled a little bit up.

Thoughts?

Steve