OK, I have somewhere over 40 years of hands-on commercial experience...
Unfortunately, it's not with electronics as that's only been a 'hobby'...
Nevertheless, here's a few more 'open comments' (please don't take them the wrong way?):
1:
I ADORE LTSpice!!!
As a program, it truly ROCKS!
(Although it's managed to crash VirtualBox under Slackware a few times now... LOL)
2:
I'm treating this whole project primarily as a 'learning exercise', and I'll no doubt make a MILLION mistakes along the way.
(In fact I actually WANT to make those mistakes so that I can learn from them!)
3:
I'm actually intending building my own 'lab power supply', but I'm betting it will be somewhat different as I have different 'criteria'.
I want more current and more voltage than this puppy is currently designed for.
It's my 'intention' to butcher a PC SMPS into becoming a microcontroller adjusted pre-regulator.
(I will probably build two 'identical' systems so that I can have a 'dual tracking' system - PC SPMS are dirt cheap anyway).
I know it's easy enough to get almost 30V from a PC SMPS after replacing all the output caps - I'm hoping to end up with about 25VDC at a reasonable current at the output of the linear reg.
4:
The current limiter in the design seems rather s-l-o-w to operate.
It's quite easy for me to simulate 10% or more 'current overshoots' during power up.
As far as I can tell, this is because the op amp in question is allowed to be driven to the rail (saturated to clipping) and it takes a finite time to 'recover'.
I'm still thinking how to address that aspect...
5:
As mentioned previously, I have already placed some small caps across the current sense amplifier feedback resistors.
Initially 10pF, I've since changed these for 100pF.
I also changed the 1N4148 diode for a Shottky (it was WELL worth the effort... LOL)
6:
Since I am aiming at extra amps, I've elected to drop the current sense resistor from 1 ohm to 0.1 ohm.
In order to keep the current reference input at 1V = 1A, I dropped the series 100k resistors feeding the current sense amp to 10k giving it a gain of 10 (thereby producing the same 1V per Amp output on this op-amp).
I am guessing the down side to this is that it also amplifies any input offset of the sense amp.
I can 'calibrate out' any such discrepancies in the microcontroller (and at present, it's a fairly linear 21.772mV of 'offset' across the full current range)
7:
Upon power up, the voltage manages to 'overshoot' by quite a decent margin (which I'm led to understand is common on LDO regulators)
I'll be able to address that by simply 'ramping up' the voltage reference from the microcontroller.
(More than likely, I'll be using a poor-mans DAC (PWM-DAC) for this anyway which will implicitly create the ramp up - I may not NEED to slow it down.
8:
Now that I am _beginning_ to understand the concepts of phase margin, I have a LOT more 'tuning' to do.
Summary:
I don't know if I should be THANKING Amspire, Dave et al for getting me started on this, or perhaps CURSING you all...
Only time will tell...