LM5005 Buck Converter

[sac1991]

I have designed a buck converter 48V-12V 1A using Webench Design Tool. Every PCB behaves same way. Initial 30s output becomes 12V then it starts to show weird ripple in the output voltage. It would be clear from the scope screenshot I have attached. After reading the datasheet again I have added a ramp resistor between Vcc and Cramp as per datasheet calculation. But still it shows no improvement.

 

[prairiedog]

For 300KHz I think 120uH L is too big. I think the LM5005 times out instead of achieving current-limit?

 

[treez]

120uH does indeed sound too big as PrairieDog says....

Also, try make rcomp smaller value, and ccomp bigger value, say 1k/10n...also, try add resistor between divider point and the ccomp...say 22k or more

- - - Updated - - -

i think webbench sims it for you?...what delta I are you getting on the inductor?

- - - Updated - - -

hang on no....as attached, your delta i at full power would be 250mA which is not bad...so try the above resistor changes.
Do you mean it is unstable in the sim, or on the bench?

 

[sac1991]

120uF seems big, but I did some calculation, there should be no problem with it. I will change the resistances and look at the output. It is getting unstable on the bench, simulation was perfect.

 

[treez]

actually i see you only have 10uF output cap so maybe you need something like 4k7/4n7 in feedback....and remember the series resistor to reduce gain..that often works.

- - - Updated - - -

Maybe if the ramp cap is too big then you hit the peak current limit and it motorboats, as you are seeing...maybe your ramp cap should be smaller in value?....sorry i havent had time to run the calcs on it.

- - - Updated - - -

You are using schottky as buck diode?..

 

[sac1991]

I tried everything then I dropped the design and went with LM2576 which is simple.