Don't use the external pass transistor configuration as shown in post #10. It impacts load regulation because the base-emitter voltage of the pass transistor isn't in the feedback loop. It always puzzles me why someone would do that.

In either case the base current passes through regulator. If it is adjustable you adjust the output for the desired voltage. It does not have to be the output voltage of the regulator. In a set voltage design where the desired voltage is the output of the regulator putting the transistor at the input is the right design.
If he wants 30 or 35 volts there is no LM7830 or LM7835, is there? So an adjustable design is called for.

 

You mean they're outside of the typical 4% tolerance? Or the A suffix parts which are 2% tolerance?

page 2 : ΔV.O -- Load Regulation . . . blah blah -- i never saw them outputting more than 30mA before dropping 100-s of mV
page 3 : V.IN (Min) -- Minimum Value of Input Voltage Required to Maintain Line Regulation 6.7 7 V -- this must be a typographic error
the LDO v. of those things is actually worse than regular (about it's V.IN Min) -- the chip is designed good but built poor and even worse (making it a useless joke)

 

Sir, following the schematic in fig 7.3 can i use a npn transistor for positive output voltage?

If you hang an emitter follower on the output of a 78xx, it isn't inside the feedback loop - regulation will be rubbish.

The way to include the pass transistor in the fb loop is to put a resistor in series with the 78xx input and use the volt drop to control a power transistor. A PNP transistor turns out to be the right way round for use with a positive regulator.

Some application notes describe a more powerful circuit - the 78xx controls a medium power PNP, which in turn drives a NPN power transistor.

 

If you hang an emitter follower on the output of a 78xx, it isn't inside the feedback loop - regulation will be rubbish.

The way to include the pass transistor in the fb loop is to put a resistor in series with the 78xx input and use the volt drop to control a power transistor. A PNP transistor turns out to be the right way round for use with a positive regulator.

Some application notes describe a more powerful circuit - the 78xx controls a medium power PNP, which in turn drives a NPN power transistor.

Rubbish? Your point isn't wrong, certainly. I guess it depends on how much regulation is needed.
The basic text book voltage regulator circuit is basically just a zener and an NPN. No feedback at all. Yes, better regulator designs put the feedback on the emitter but I don't think that would work on the LM317 designs.
All things considered, yes, the PNP, 78xx design is superior. But rubbish? How much regulation is needed? As emitter voltage and current varies so does base voltage and current, but the process is just not as good as your suggestion.

 

fig 7.3 can i use a npn transistor for positive output voltage?

If you hang

zener and an NPN

a slight modification of -- but in principle -- the Fig. 7.3 can be set/-tuned to work -- but it requires . . . quite a bit of work

fig 7.3 can i use a npn transistor for positive output voltage?

you can cascade transistors but you have to track 3 basic states of a voltage regulator

1. no or the minimum load on OUTP
2. the max. load on OUTP
3. the regulation/feedback( loop) from 78sxx to transistor switch (all transistors in) enabled
   + no load Start enabled
   + max. load Start enabled
   + current limiting enabled
   + OUTPUT short ciruit limiting (foldback) enabled
   + recovery from OUTPUT short enabled
   + perhaps i forgot something ....
   
4. Happy that day! -- keeping the required features enabled through modification of your power switch transistors
5. + i did forgot something -- keeping the power(currents) at all regulation/regulator paths in the right ranges for those paths - while achieving best regulation
   + i don't think it ends there
   + you need to do it at all times simultaneously (by year or so you get used to it , muhahaa)