National switching regulator LM2678

Mar 24, 2006 5 Replies

I'm using the LM2678-ADJ and have the output set to 4.4V. This is a somewhat arbitrary value during my course of testing a circuit including this regulator. I'm connecting a litte more than a 2ohm resistive load across the output, so it should draw less than 2.5 amps; well within the regulators 5 amp rating. I'm using a 100uH toroid inductor, rated for >8 amps, a 1000uF cap at the input, and 470uF at the output. Voltage is set with a 1K and 2.7K resistance.


Here's the problem: when I raise the input voltage to more than about



17V, the regulator's output starts to reduce. At 25V input, the output drops to about 3V. The output is still pretty clean. This seems like a feature, but it's a bad one in my design, and I can't find anything in the datasheet regarding it.

Any ideas?



May be, that 17V is the point of change from continuous to discontinuous mode.

Try with smaller values in the inductor, for example 33uH.

I finally figured it out. I needed more filtering on Vin. I put a

0.47 uF right across the power pins.

** A 0.47 uF what ???

The data sheet showed 3 x 15 uF electros.

Got them too ?

........ Phil

Read my original post: "a 1000uF cap at the input".

Switchers Made Simple 6.24 specifies a 1000uF cap for nearly any voltage/current combination, and this will eventually go into a variable power supply. The 0.47uF is just for "up close and personal" bypassing.

Phil Allis>

** Don't top post.

Bad usenet ettiquette.

Looks hostile and offensive.

** Does not answer my question.

** For what ?

100 / 120 Hz ripple filtering ?

** Again - a 0.47 uF what ??

Film, tantalum, electro ???

If you had a 1000 uF electro fitted close to the device, a 0.47 uF would do SFA - whatever type.

....... Phil

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required