LM2841 Duty Cycle

JerryLin

We design a Buck converter for LM2841X. The input voltage is 16V ~32V, output voltage is 3V , load current is 0.15A. The schematic is the same as that of Webench, L is 120uH, C is 22uF*2, Cboot is 150nF. We probe duty cycle by voltage of SW pin.

We found that the duty cycle of the circuit is not constant. Why? The Buck should be in continuous conduction mode for the working condition, and the duty cycle should be constant. And for higher input voltage, it is more obvious.

The following is the captured oscillator image. 3652.duty.tif

over 10 years ago

0 Frank De Stasi over 10 years ago

TI__Guru 57555 points

This could be simple instability or noise. Did Webench select the 120uH inductor ?

That seems rather large ? You could try a smaller L.

Be sure to follow all the appropriate guidlines in the data sheet regarding layout and input bypass.

0 JerryLin over 10 years ago in reply to Frank De Stasi

Prodigy 180 points

Yes, Webench select 120uH. For 32V input, Webench even select 150uH inductor.

How to determine whether it is instable or noise?

When the probe is connected to SW pin, the output ripple increase very much.

0 Frank De Stasi over 10 years ago in reply to JerryLin

TI__Guru 57555 points

I would try a smaller inductor as a test. If that does not clean up the

duty cycle I would look at you layout. You can send us your layout and we

can take a look at it.

0 JerryLin over 10 years ago in reply to Frank De Stasi

Prodigy 180 points

Yes, a smaller inductor can improve the problem. But why? The Operating values in Webench show that the higher input voltage, the larger phase margin (more stable). For large inductor, the circuit is stable for low input voltage. But for higher input voltage, the duty cycle is not constant. This is not the same as results of Webench. Why?

For this working condition, what is the recommended value for inductor?

Power management

Power management forum

LM2841 Duty Cycle