LM25118-Q1: Large output ringing when switching to boost control

Jackson Thomas

Part Number: LM25118-Q1

We are using the LM25118-Q1 as a voltage pre-regulator into an LED driver. This is running off an SLA battery with a 10 V output, so we used the boost/buck to support low input voltage swings. When the input voltage is >14V, the part operates in buck only mode and the output is fine. When adjusting the input voltage below 14V the boost mode starts to turn on and we see large ringing spikes on the output voltage. This is causing issues with our LED driver control circuit. Are there any changes you suggest to the implementation? We have lots of capacitance on the output and a snubber on the recirc diode. We have also tried increasing the series resistance on the FET gates and adding some smaller caps in place of C10, the smaller output caps. Any help is appreciated.

Regards,

Jackson

over 6 years ago

0 Garrett Roecker over 6 years ago

TI__Mastermind 21720 points

Hi Jackson,

What are each of the waveforms in the screen captures?

Please provided the application specifications (VIN, VOUT, IOUT...) so I can review the schematic.

Thanks,

Garrett

0 Jackson Thomas over 6 years ago in reply to Garrett Roecker

TI__Genius 14415 points

Hi Garrett,

Vin is 8-18V, 12V nominal

Vout is 10V

Iout is up to 6A

In the waveforms, channel 3 (purple) is Vout, zoomed in and out. The other signals are relevant to the LED driver, it is LED current and current enable signals. You can see the ringing in the zoomed in picture is at about 125MHz, and when we have ringing out LED driver malfunctions. The Vout signal is offset by 10V to show the ringing better.

I also used these parameters and plugged it into a webench simulation to see what comes up, and I see the same behavior. As soon as Vin drops below 14V, this ringing appears on the output.

The simulated circuit didn't have the same output capacitance or the diode snubber, but it seems to perform just the same. Is this a typical response of this part?

Thanks,

Jackson

0 Garrett Roecker over 6 years ago in reply to Jackson Thomas

TI__Mastermind 21720 points

Hey Jackson,

It looks like this might be caused by the loop stability. Please change the following values to help stabilize the control loop.

C11 = 420pF (Will increase the slope compensation
R6 = 18.2kOhm
C6 = 47nF
C7 = 1nF

Thanks,

Garrett

0 Jackson Thomas over 6 years ago in reply to Garrett Roecker

TI__Genius 14415 points

Thanks Garrett,

I added these compensation values to the simulation on Webench, and it seems the ringing on Vout is reduced, but is still around 2Vpp when boosting. Is this expected behavior? Are there other ways to reduce this ringing?

--Jackson

0 Garrett Roecker over 6 years ago in reply to Jackson Thomas

TI__Mastermind 21720 points

Jackson,

Did you test the compensation components on the actual board? If not please do so.

Also what kind of diode is D4? I cannot find the part on the web.

Thanks,

Garrett

0 Jackson Thomas over 6 years ago in reply to Garrett Roecker

TI__Genius 14415 points

We did try these modifications and we still see the same behavior. It may have adjusted the voltage where the boost function kicks in but only by about 0.3V.

The diode should be this one. https://www.onsemi.com/pub/Collateral/NTS1545EMFS-D.PDF

--Jackson

Power management

Power management forum

LM25118-Q1: Large output ringing when switching to boost control