Voltage overshoot and undershoot

[ThievingSix]

So i picked up a design for a 0-30V 5A variable PSU. It had some pretty serious flaws however i managed to find a revised version and make some small changes of my own.

So i plugged all the values into multi-sim and simulated the circuit, however i found a pretty serious problem with the output voltage. When i changed the pot to a higher resistance there was undershoot to 0 volts before it climbed back to the correct voltage, the same was true when i decreased the resistance, however there was overshoot to 40V before it stabilised. To make matters worse the voltage took around 2-3 seconds to become stable. Perhaps i've done something wrong(most likely), but another person who built a similar design seemed to have no problems with theirs. Could it just be a spice error? Any idea how i could fix it?

I've got little to no experience in actually designing a PSU, so its somewhat a first project, other than the standard LM317T and 7805 stuff and fairly basic experience with AC rectification. Any help is much appreciated!

This is what it looks like on the DC output when i decreased the resistance on the pot.


This is my schematic. The 10K pot, P2 controls the variable voltage.

[Rerouter]

based on the design the undershooting would be the the capacitor likely triggering the current limit, (as voltage climbs the capacitor draws current, thanks to that 330pF cap between its output and inverting input it will be slow to respond so will overshoot then the current will really back off, then cycle a bit before stabalising,

from my own experiences, change it to 10p or remove it entirely, the smaller those stabalisation capacitors (C8 and C9) the smaller those errors will be and the faster the output will settle, being how your only driving transistors there shouldnt be that great a need for them anyway, keep the pads on your pcb, but you should be able to get away without them,

[PeterG]

I would also try simulating the design in LTspice to confirm.

Regards

[ThievingSix]

Thanks rerouter, your suggestion worked perfectly, the voltage stabilises much faster on startup aswell, the undershoot is only about 1v for 2ms, which is far better than the 2-3 seconds i was getting earlier. The overshot is better too, although it spikes quite a bit while i vary the resistance, i'm sure there should be a way to minimise that spike, but i can't seem to figure it out. I tried a low value resistor in series with the input signal, and it helped but not significantly.

Thanks for the help!

Increasing the resistance first and then decreasing it using the 10k pot.

[Rerouter]

i have yet to find an actual function for the center op amp, (care to enlighten me?) personally it would seem to me it would be better used as a differential op amp for the current sense, but as i dont see its function as obvious i may be wrong,