Understanding switched mode power supplies

[ShubhamSharma]

Hello everyone
Can someone please help me in understanding the right half plane zero problem usually found in switched mode power supplies(SMPS)? Please tell me the problem created by the right half plane zero in
SMPS and solution for the same.

 

[CataM]

RHP zero gives a phase inversion, so the output will take the opposite path before returning back to normality. Compensator should not see that, so it must have reduced BW.

 

[E-design]

It is usually only a problem that comes up with indirect transfer converters, like boost, buck-boost and flyback. Operating deep into CCM will make the problem worse, and as stated the crossover must be reduced very low to avoid the RHP zero effects.

 

[asdf44]

In plain language a right hand plane zero means that the initial response of the system to a “request” for more voltage is that voltage briefly dips.

It’s like some cars where if you floor the gas the engine will lose power at first before picking up speed.

This is a inherent property of sometopologies but not others.

The bottom line is that while sources feel a need to mention it, it rarely matters unless you’re really trying to push the performance of your system. If loop bandwidth 1/10th or lower of the switching frequency, which is typical, it doesn’t matter.

 

[ShubhamSharma]

How can I lower the crossover?

 

[E-design]

You will select the proper feedback value capacitor on the error amplifier opamp to reduce the total gain to 0 dB at the desired crossover frequency.

 

[ShubhamSharma]

Does reducing the crossover frequency reduce the bandwidth also??

 

[asdf44]

Does reducing the crossover frequency reduce the bandwidth also??

Yes that's the same thing.

 

[ShubhamSharma]

What is the effect on the stability due to the reduced bandwidth? If there is a trade off between stability and bandwidth, then how can it be solved?

 

[asdf44]

No, in power supplies lower bandwidth almost always makes stability easier.

So much time is devoted to studying advanced compensation schemes but actually standard converters can all be stabilized with a 1-pole integrator.

So the point of 'fancier' compensation schemes isn't stability - it's stability AND performance (high bandwidth).

 

[treez]

RHPZ is to be heeded, as you know.
Your crossover frequency shopuld be no more than a third of your rhpz frequency.
The book by Basso gives the rhpz frequency for various converter topologies.
Discontinuous mode converters dont have RHPZ's.
Also, Buck converters dont have RHPZ's.

- - - Updated - - -

Also remember that constant off time control means you dont get an RHPZ

 

[ShubhamSharma]

What are other ways, besides reducing the crossover frequency, to nullify the effect of RHPZ in switched mode power supplies?

 

[mtwieg]

Only two options:
1. Use another topology which doesn't have the RHPZ (buck-derived topologies).
2. Don't operate in continuous conduction mode.

If you use a buckboost or boost-derived topology in CCM, then the RHPZ will be present, and its effects cannot be "nullified." You can raise the RHPZ frequency somewhat by decreasing your inductance, or just cut your control bandwidth, but obviously those two options have drawbacks.

 

[ShubhamSharma]

can you tell me the value of voltage sources used in the circuit so that we can perform the simulation?