I thinking regulation loop is to slow to response fast enough. Remember, most of active output stage (transitors, mosfet etc.) working as current source (high impedances) to new information comming from regulation loop.
For handle transient step properly, you need infinite bandwith on regulation loop (and still stable) - in practical is not possible and you need output capacitance with low impedanse on high frequency (here > 200 kHz) to feed power under approx 2.5 us to regulation loop gains new setting.
test also load step with reactive and resonant load and see how this works.
- I think resonace circurit load around 400 KHz can oscillate nicely after load step with above settings ;-)
---
remember real worlds loads is never pure resistive and in 95% of case large capacitive on low frequency. - and bad regulation loop design and reactive load can make whole thing to oscillate even if workings stabil on resistive loads situation.
In switched Power supply-design with many filtering step to output and want fast response, make reactive load step very often instability and oscillating in different load situations - and many case, customer find out this 'feature'...
jutek,
Overshoot and undershoot values are with respect to the steady-state value. In your case, both the overshoot and undershoot are with respect to the 1.385 value. So the udershoot would be [(1.3862-1.385)/1.385] X 100 = .0866%.
Regards,
Kral