LM5170-Q1: LM5170 boost mode, how to reduce the vout overshoot during power-on by enable.

Dear Youhao:

Appreciated your reply so much.

I will try it tomorrow and feedback here.

In fact. When I tested LM5170 power up by manaul, the vout is rising by VREF rising level. If the VREF is slowly, It is real effective to reduce overshoot . So your idea shall be good to reduce the overshoot.

After the LM5170 is controlled by MCU (the VREF is set on target and power up by enable signal), the vout overshoot is always there.

About your suggestion, That means no chance to reduce the vout overshoot by modifing SS cap or compensation value. Is it right?

Thank you again.

Aska

Dear Youhao:
You are right that the vout overshoot can be reduced by VREF rising time.
In fact, the VREF rising time and level are controlled by microcontroller. So we can easy to find the best value for Vin=45V.
But it doesn't workable at Vin=32V while the VREF rising time is similar as 48Vin. The rising time of VREF is too long to meet LM5170 power up quickly.
For VREF control solution, it seems not prefect result at Vin=32V.
May I reduce the value of ramp cap or increase switching frequency? Now they were 1000PF and 150KHz.
Could you please provide some direction or idea?

BTW, I found the Ramp voltage is charge from VINX(VIN,from HV). I don't know whether any effect from Vin=32V or 48V for Ramp voltage?

Thank you so much.
Aska

Dear Youhao:

Added the overshoot waveforms of Vout.

Because the voltage of feedback divider resistor is different to Vin48 and Vin32V, we set 1.21V for 32Vin and 1.87V for 45Vin. The target level of VREF is 2.47V. So the rising time of VREF at 32Vin is around 4msec (long than 45Vin/2msec).

Do you know why it doesn't work on 32Vin by VREF control? (It is ok at Vin48.)

BTW, we tried to modify the Cramp but no bigger difference.

Could you please provide some directions?

Thank you so much.

Aska

Hi Aska,

I reviewed your waveforms but the labels confused me.  Could you clarify?

Thanks,

Youhao

Dear Youhao:
Sorry for confuse.

Ch1 ISEA (pin 45 of LM5170), it comes from voltage loop.

Ch2 EN (pin39 and pin43 of LM5170)

Ch3 VREF, it seems C92, but it is controlled by microcontroller. We modify the rising time and level by microcontroller to achived your suggestion.

Ch4 Vout, it is the output voltage of boost converter. It seems HV of LM5170 EVM.
Thanks.
Aska

Hi Aska,

Thank you for the clarification.  Please understand that the actual soft-start operation is to control the duty cycle growth at a slow rate to avoid overshoot.  Since larger duty cycle is needed for lower input voltage, the soft start process would take longer time than a lower input voltage.  What I recommend is, to find a minimal Vref slew rate for your min input voltage, then always apply the same slew rate to other higher input voltages then you will always avoid getting an Vout overshoot.

Thanks,

Youhao