SMPS goes unstable as electrolytic output caps age?

[treez]

Hello,

I am doing a 45W, DCM flyback (isolated) with 90-265VAC mains input.
(Vout = 25V)
The feedback is optocoupler based.

When the electrolytic capacitors in the output get older and have a greater ESR, this will reduce the ESR zero in frequency.
Will this mean that its likely to go unstable?

Or will the reducing ESR zero frequency actually make the SMPS more stable?

 

[mtwieg]

Depending on where the zero is relative to the crossover frequency, it could possibly hurt or help stability. For example, if fc>>fesr, then further decrease in fesr will not significantly change the phase response, but will cause the loop gain to rise, probably thus degreasing gain margin. On the other hand, if fc is near fesr, then decreasing fesr will likely cause gain and phase to increase, possibly causing fc to increase while gain/phase margin remains high.

 

[tpetar]

Selecting Output Capacitors In Power Applications
**broken link removed**

Choices in switch mode power supply capacitors
https://www.electronicproducts.com/...s_in_switch_mode_power_supply_capacitors.aspx

 

[treez]

if fc>>fesr

crossover freq is not likely to be much greater than fesr, unless loads of 'lytics have been used....so it looks like the "increasing esr with age problem" isnt really a problem at all.....it doesnt appear to bring on instability?

 

[mtwieg]

In my experience at least, fesr is never much lower than fc, so in such cases I don't predict that lowering fesr will harm stability. But only simulation or experimentation could confirm for sure.