Series Pass Regulator

[Simon123]

I was looking at some ON Semi PDF and this circuit:
http://www.onsemi.com/pub/Collateral/HBD854-D.PDF
(Page 36.)

I was wondering why would they use PNP Error amplifier and reference it towards + output instead of common ground and use NPN?

Thanks.

[lordvader88]

IDK, but I was just starting to make some basic regulator circuits to learn multi-BJT circuits, and them parts arrived for a SMPS repair, and I've only been at that since. (But that's my backup computer PSU, so I better stay at it)

[SeanB]

I would guess it makes driving the pass element easier from the differential pair, as the collector voltage swing has a much higher gain than using a single transistor to compare the sample with the reference, and the differential pair is inherently temperature stable as well, only needing a good stable reference voltage, and all the temperature drift is cancelled out by the differential pair gain for the rest of the circuit. flipping it around to use NPN differential pair would need another transistor or two to do the level shifting needed to drive the pass element, and this extra stage will add more phase shift and reduce stability of the circuit as well. The first circuit as well can easily have current limiting added to it, as the current sense amplifier can easily feed current into Q3 to control the current in the circuit, using only one extra PNP transistor.

[Simon123]

What about this then?
It uses same darlington pair and shunt transistor.
I know this one is using a negative rail, but that is only to get output to 0V, also this one has Linear response opposed to the 1st circuit.

But the thing is, it uses NPN, without any extra shifting transistors?
Kinda confused now ...

So i guess the question is, why would they reference the regulator/comparator to + Output instead of ground?

[Simon123]

Ok i see the problem now.
The differential pair has a voltage across Emitter resistor and this voltage would bias the "Shunt" transistor.
Thanks, forgot about that ...