Undershoot causes PMU to shut down!

[coco]

I met a problem while using Samsung's LCD driver S6B33B2B, the input of the 1'st booster is the main power, i.e. 2.9 volt, when I change the setting of the booster from 1.5x to 1.0x, there is always an overshoot on the main power supply, which is around 300mv, following a undershoot which sometimes exceeds 200mv, and the undervoltage is detected by the power manegement ASIC as a short ciruit event ,and shut down.(an undervoltage more than 10us and - 174mv can be detected by the PMU, and shut down itself)All the component should be surface mounted and the space is quite tight , so large filter circuit or an extra regulator is not applicable,

Do you have any idea to solve such a problem. Do I have to add some capcitors to filter it out, or should I minimize the load and then change the setting of the 1'st booster..., what can I do to minimize this effect? Thx in advance.
(attached is the waveform of the issue)

 

[bobcat1]

HI

If this is a working system check & replace the output capacitor's

Add protecting diodes {in revers to the output voltage)

add small cap e.g 0.01 ~ 0.0001 uf


All the best

Bobi

 

[coco]

Thanks for the reply, I have tried to replace the capacitor of the Booster's output C16 (1uF) with a smaller one 1nF, and the overshoot on 2.9V disappears, but as you can see, the waveform from the output of the 1st booster becomes so bad, what are the capcitor values that you recommend, can anybody give me some help, thx

 

[coco]

As you can see, there is no apparent over/undershoot of the waveform below at the falling edge after replacing the C16 with a small capacitor 1nF, can you please explain this effect for me. By the way, is there some kind of trade-off between load regulation and transient response of the LDO? How can I minimize this effect (undershoot below 50mV is acceptable) thanks a million.

 

[bobcat1]

Hi

Don't replace the cap add one more because the smaller cap can handle high frequency and short pulses

and the bigger cap handle low frequency and wid puls like changes on the ps line


All the best

Bobi

 

[coco]

I have tried 1nF+1uF for C16 also, however, it does not seem to have much improvement, it behaves just like the 1uF it self, I wonder is there anything wrong with the capacitor values, what values do you recommend? I also upload a booster circuit inside the driver, I think the problem occurs when the "switch" switches, the load current decrease suddenly, which cause the LDO to increase its output voltage(overshoot), is this correct? Does adding some capacitors at the output of the booster solve the problem?

Many thanks.

 

[coco]

and the 1nF+1uF looks like this...

 

[bobcat1]

Hi

You can add limiting diode accross the cap{ i think}, this will limit the pulse to 0 volt


All the best



Bobi