Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Feedback loop compensation in SMPS

Status
Not open for further replies.

mateuko

Junior Member level 1
Joined
Mar 22, 2007
Messages
16
Helped
2
Reputation
4
Reaction score
0
Trophy points
1,281
Activity points
1,403
loop conpensation smps

What is going on about these poles and zeroes compensation in feedback loop of switching power supplies? Why to worry about this? What if I won't do such compensation?
 

If you don't two scenarios are possible:

1. the loop will oscillate and your output voltage will have large low-frequency ripple, most likely exceeding the specs.
2. The loop may be overcompensated and in that case very slow to respond to sudden changes in output current demand (poor transient response performance). The output voltage will "dip" below the specified minimum and take a long time to recover.

Neither of these scenarios is acceptable.
 

yes , the point 1 i say that this ripple will be a few KHz......depending on the resonant freqency of the output LC filter or the corner frequency of the C(out) and the Load R.

Though the point 2 i feel worried.........

i would have thought that leaving out compenasation components would probably give oscillation...but not a slow transient response?...on the contrary...the response without compensation would be as fast as it gets...and the SMPS would be reacting to everything straight away........

it would be crash banging the switch on and off.

...without compensation....
....we know with the SMPS, that when the feedback divider tells the output voltage is falling , the controller thinks it has to increase duty cycle....however, due to the delays in the system....the uncompensated controller does not realise that when it increases the duty cycle , the output had possibly ALREADY started going back up...but this was lost on the uncompensated controller and now the output voltage goes up much further than it would otherwise have done.....so we kind of need to filter the output voltage.......kind of "moving average" it.....so that we do not respond too hastily and end up making the (voltage) ups go even higher and the downs go even lower.

Though its worth remebering.........the Power integrations LNK304 etc family....these are offline controllers using ON/OFF control an require no compensation components whatsoever, and give perfect transient response and never go unstable.

the biq question is ....since they are so good , why do we not use them at higher powers too(?)
 

Leaving out compensation components would not give you slow response, but it would likely cause oscillation. It's when you use too low a pole frequency that you overcompensate and make the response slow.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top