I'd like to present some measurement of short circuit and transient response. For the first measurement a simple low Rdson mosfet switch is used to short output terminal with 0R010 power resistor:
Output signal from signal generator (SIG trace) is limited in duration (SW trace) and output (Vout trace) is monitored:
Recovery time from short circuit is ~1.2ms:
Measurement of voltage drop on the Rsense (connector X10) can give some idea about current:
Zoom in on the rising edge:
... and zoom in on the falling edge (some oscillation is visible?):
To test transient response I built a Closed Loop Transient Load Tester as described in LTC's app note
AN104-15. I was not able to follow procedure completely since I cannot afford any current probe. Therefore I just monitored Vout when the tester is connected and active.
Tester is housed in cute and easy to drill Hammond
HM1590U enclosure (if somebody is interesting in PCB, I can publish it):
At the beginning I tried to test output voltage up to 50V and (of course!) that cost me few output transistors (D44H11) since that violate badly SOA. I noticed that PSU perform almost uniformly from 10 to 50V (below 10 and definitely below 7V tester starts to generate some bad oscillation. First I blamed my PSU but when I tested the same range with i.e. 3-pin standard regulators I realize that something is wrong with tester by design or because I made some bad mistake during assembly). Anyway I proceed staying mostly in the middle of the scale, to Vout=25V.
I also noticed that measure point is important. I tried that with both AC probe described in above mentioned LTC document and with 10x probe that comes with Rigol MSO. Also connecting some capacitor in parallel with probe can change results. My AC probe looks like this:
Tester have possibility to adjust DC bias of little above 1A. On top of the another 1A can be added depending of the input signal level that is used for switching on and off additional load current. Transient response vary in accordance with selected continuous (DC bias) and "switched" load (signal level). If we goes with "full scale": 1A DC bias and 1A switched load on the chassis binding posts I got the following results, 146mV over-/undershot (measured at the tester's binding posts):
Zoom in on the rising edge (undershoot is 78mV or 0.31% at 25V):
... and zoom in on the falling edge (overshoot is 72mV or 0.28% at 25V):
If I move probe from testers binding posts to the post-regulator PCB output terminals, results looks better when over-/undershot drops to 54mV:
Maybe is good to add what MSO capture when load tester output is switched off but signal from the signal generator is still present: