Thread: Power supply design

abbas welcome to the club!

Just a advice always do a search for the topic you want to know or need to post about .. cause there are most topic already discussed here.. !!

so just do a search on "linear power" .... you will know the drawbacks of the above...




Mastero

Actually - see if you can dig up that link. If there are more efficient surface mount components, they might be a good idea to research. Currently we're building PSU's from TO-220 package chips, and getting about 65% efficiency. Perhaps some ingenious surface mount manufacturer has found a better way?

Radio shack linear power convertors can't handle the Amps and voltage range needed to get the power we need, so we're using switch-mode convertors from Maxim-IC. Those present the problem that the overall psu size is somewhat bulky (albeit it's pretty damn small for a PSU... 1.5" x 5" or so)

Jeff Mucha (zootjeff) has designed a PSU that uses surface mount comps, and his is even smaller than the sproggy described above.

hey CL i agree with you but the thing with SMT components are that everyone cant solder them ... its kindoff SMALL ..... and the pins are mostly thighly packed to solder at home..

And as i remember one off the member here had some new psu idea using those SMT chips from .. National .. but i dont know who it was ..

Dont even know what happend to his work ..

Mastero

Thanks for all the replies!

Forgive me but I think I did not make clear that I was not talking about a linear converter but a full high performance switching power supply. I believe a linear power supply will not be able to achieve that kind of efficiency ( > 90%).

I think we should take a look at Linear technology web site especially at the following components:

LT1931 or 1931A (inverting DC/DC converter) for the -5 and -12 v low amperage but that's all is needed if I read all the post correctly.

LTC1628 / LTC1628-PG step down switching regulator for the both the 5v and 3.3V. That is a one tiny chip solution.

Also LTC1625 step down switching regulator.

Included in the PDF files from the manufacturer are sample designs.

The nice thing also is the input voltage could be as low as 5V and as high as 36V. That would probably address the voltage drop on the battery.

According to what I read so far we ought to be able to produce a switching PSU that is very small ( 2"x4"x0.5")

What do you guys think?

Www.linear-tech.com

PS: I have no relation of what so ever to linear tech. I just read the intel design and many other articles.

the LTC1628 is for both the 3.3V and the 5V
3.3V at 4A
5V at 4A

LT1931 one for -12V at 350mA and one for -5V at 350mA

LTC1625 for the 12V at 2A


total around 64 W

Originally posted by abbas
the LTC1628 is for both the 3.3V and the 5V
3.3V at 4A
5V at 4A

LT1931 one for -12V at 350mA and one for -5V at 350mA

LTC1625 for the 12V at 2A


total around 64 W

Total waste of time if you ask me.... How small will this thing be?? If it isn't smaller than 2inches by 2inches and 1inch thick max.. then its just pointless as teh Sproggy supplies are better than that, and I don't think that will be able to power most systems...

Mp3kid - It might be prudent for us to explore these components as a possible solution: Just how long do you think Maxim is going to let us keep getting samples?

Besides - What's the "drawback" of having more options to choose from? So far there's only 3 choices:

Keypower/Arise: $165/$100
ZootJeff: $150
Sproggy: $130 pre-fab, $50? with samples

Hell - I think if you designed a dual layer design like Zootjeff's supply, you probably could go smaller than 2x4". Draw up a quick parts list and let's see where we could pull together parts from. One of the major obsticles of the Sproggy DIY kit is that you need exotic inductors from Coilcraft or some other inductor manufacturer. (well not really exotic, just annoying to get)