LM5017: FB Ripple

user6533883

Part Number: LM5017

I am using two LM5017on my board. One for 5V and the other for 24v. When before the design i used to webench.ti. By the way my input voltage is 70v DC.

The picture is recommended design by Ti.

And I used this design.

I copied this design by datasheet. And then observed 5v output i have to following graphics

I have 640 mV ripple as can be seen in the picture. But I simulate on wibench.ti i should have to this graphics.

And i designed the 24V buck converter from wibench.ti and this graphs are also the graphs of the 24V line.

As can be seen in the pictures the graphs are not what hey should be. Ripples are bigger than it should be.I added capacitor and inductor to the output voltage line with values greater than the values found in the schematic. But i could not fix it.

How can i fix this problem?

over 4 years ago

0 iAlejandro over 4 years ago

TI__Expert 8130 points

Hello,

Can you please share the images in which you are referring to? They are not demonstrated on the post.

Thanks,
Alejandro

0 user6533883 over 4 years ago in reply to iAlejandro

Prodigy 40 points

Hi, I relink the pictures.

0 iAlejandro over 4 years ago in reply to user6533883

TI__Expert 8130 points

Thanks for the images. Before we continue any further, lets make sure we are using a good measuring technique; improper measuring technique can exaggerate the output noise. How are you probing the scope-probes? Clip-on method? Please take a look at the following:

How to measure ripple for better design outcomes

Testing switching power supplies includes many different tests, one of them being the output-voltage peak-to-peak ripple. Output-voltage ripple is the alternating current (AC) component of the direct current…

By George Lakkas over 9 years ago in Technical articles > Power management

Thanks,
Alejandro

0 user6533883 over 4 years ago in reply to iAlejandro

Prodigy 40 points

Thanks Alejandro,

I used Clip-On Method. Even so, 640mV ripples are too big on 5V lines. Also I have 3V ripple on 24V lines. I do not think the measurement will have such a high effect.

0 iAlejandro over 4 years ago in reply to user6533883

TI__Expert 8130 points

I understand, we just have to make sure we are getting a realistic idea on the amount of noise first before we can focus on what filters/changes might be needed to help with the unwanted noise.
Can you please provide the new results?

Thanks,
Alejandro.

0 user6533883 over 4 years ago in reply to iAlejandro

Prodigy 40 points

This time I used the pigtail method. But the result is still not good. you can see the pictures below.

0 iAlejandro over 4 years ago in reply to user6533883

TI__Expert 8130 points

Thanks for the waveforms!

I noticed you do not have your circuit schematic as WEBENCH recommended. From the schematic you provided, the circuitry is configured as Type 1. For the minimal ripple configuration, a capacitor (C_ac) should be placed from output of inductor to FB node. Please see Ripple Configuration Section in the datasheet. Setting your device as Type 3 (minimal ripple configuration) would give you best results.

Regards,
Alejandro

Power management

Power management forum

LM5017: FB Ripple