LM25118: Overshoot at the frequency buck switching

Hi Romain,

thanks you for using the E2E forum.

Can you add a picture of your measurement setup, esp. how the probe of the output voltage has been connected to you board.

Best regards,

 Stefan

Hi Stefan,

I attach a schematic with the two measured points.

/resized-image/__size/320x240/__key/communityserver-discussions-components-files/196/12V_5F00_TI_5F00_probe.png

These points are shown on

We try to add a serie resistor in the Q401 gate drive. It work's but 10ohm is not enough to slow the drive, we put a 50 ohm resistor. Overvoltage is reduce from 3Vp-p to 1Vp-p but the maximum output current is reduced.

It seems to be not a good solution.

Sincerely,

Romain

Hi Romain,

unfortunately the resolution of the attached file is very low - so I can not identify the details

Can you add a picture of your measurement setup, esp. how the probe of the output voltage has been connected to you board.

=> a Foto would be good

Best regards,

 Stefan

Hi Stefan,

When you open the link in the precedent post, do you have the schematic in good definition? Is that right for you?

I had a picture of the mesurement setup.

In this last setup we had a 50ohm gate resistor.

For a load of 2.3A@Vin=14V we obtain this measurement.

In Yellow : Vout

In Pink : Q401 gate drive

In Blue : D400 Vout

Sincerely,

Romain

Hi Romain,

with your scope probe (esp. the Ground wire) your are build a big loop with will pick up a lot of noise.

You need to make this much smaller and use the so call tip and barrel method for measuring this signals.

See an example here: https://www.ti.com/seclit/ml/slup385/slup385.pdf (page 28).

Best regards,

 Stefan

Hi Stefan,

The picture was made rapidely to answer you.

In fact, we use very small ground loop to measure the output overvoltage. Please see the picture attached.

As you can see, the overvoltage is still there.

On the LM25118 datasheet there is no point about the need of use a gate drive resistor to reduce the overvoltage spike.

If we remove this resistor, we had 3Vp-p overvoltage for an 12V output.

Do you have any idea about what we have overvoltage at CI oscillator frequency?

Sincerely,

Romain

Hi Romain,

thanks for this details - this ensures that the measurement is done in a right way and we are not working in the wrong way.

Using a Gate resistor can reduce the ripple and EMI issues but with a value greater then 10 Ohm you get into a very long switch transition time and this will bring down the efficiency and heat up the FET (often to much).

You mentioned that you have added a snubber on the FET. This is not shown in the schematic. How have you dimensioned that?
Here is a good guidance on how to do that: 

https://e2e.ti.com/blogs_/b/powerhouse/posts/calculate-an-r-c-snubber-in-seven-steps

I might be also good to review the layout - maybe you can share the SMPS part.

Best regards,

 Stefan