about compensation of LDO!

[qutang]

ldo frequency compensation chava

there are many IEEE papers about LDO's stability,but many of them don't talk about thecompensation in OPAMP.
I want to know the OPAMP's poles&zero effect the LDO's stability?
why these papers don't talk about the OPAMP's pole&zero influence the other pole&zero?

 

[sisching]

ldo compensation scheme

In my opinion, the componsation of the opmap is not important.

Since one of the input (opmap) should be the bandgap reference, it should be no
ac signal.

 

[bigpiyo]

frequency compensation for ldo

I think that still have some papers to discuss about stability for pole/zero ...

 

[pelo]

The most important thing is the compensation of the system

The Opamp make you hign gain and GBW

For the Opamp, offset is too important to analysis

 

[qutang]

My idea is the opamp's pole & zero will be include in the system 's pole&zero???.

Added after 4 minutes:

I don't know can i add the opamp's pole&zero in the system's frequency response,if a opamp has at lest 45 degree phase margin ?

 

[ericliu]

I treat the LDO as a non-inverting amplifier. It has a large capacitor load combined with a variable resister. Base on these loadings to do the opamp's compensation.

 

[gszczesz]

You are compensating a closed-loop system, not jus tthe op-amp. DO NOT look just at the op-amps stability, look only at the whole system stability.

Poles and zeroes anywhere in the closed loop path will affect frequency response, whether they are located inside the op-amp circuit or elsewhere. For maximum gain-bandwidth product, you want to compensate the system at the dominant pole. If you try to first stabilize the op-amp by itself, then you will be introducing a larger pole that may NOT be the dominant pole in the LDO.

In any case, there are LDO circuits that try to make sure the dominant pole is the output capacitor, so that once you exceed a minimum sized cap, it will be stable for all loads.

 

[jutek]

there are many IEEE papers about LDO's stability,but many of them don't talk about thecompensation in OPAMP.
I want to know the OPAMP's poles&zero effect the LDO's stability?
why these papers don't talk about the OPAMP's pole&zero influence the other pole&zero?

maybe they don't treat opamp's compensation because it's an one stage one.

try this way of compensation. it really works

A robust frequency compensation scheme for LDO regulators

 

[jefferson]

gszczesz is right. poles induced by opamp is non-dominant pole,it's located at high frenquce point.waht we care about is the dominant pole whicth is critical to the stability of the whole systerm.

 

[leebluer]

I agree with gszczesz. You should compensate the whole LDO close-loop, but not the opamp. Because there are not only opamp in the close loop, but also other MOS transistors. You should also consider the Pole/Zelos from other MOS transistors. In other words, we compensated to stable only for close loop

If there is any other sub-close loop in the whole loop, you should firstly compensate it stabiliy

 

[qutang]

I also think gszczesz is right. but another question is if we design a different loading regualtor , we should modify the whole stability, including modify something inside opamp?

 

[jutek]

I also think gszczesz is right. but another question is if we design a different loading regualtor , we should modify the whole stability, including modify something inside opamp?

hi

not exactly modifying something, but you'll have to change biasing and transistor sizes. If assume more load 100mA instead of 10mA for example, it means bigger pass device so phase marging decreases. So you need more curren or additional compensation circuit.

regards

 

[qutang]

to jutek:
that means we should calculate the pole/zero of the close loop including the inner pole/zero in opamp by hand, then we adjust it under working conditions, isn't it?

 

[zenisle]

using common technology is ok

 

[johnsmith101]

Read this document starting on page 10. It talks about the various factors in LDO stability.