Hi! I'm a mechanical engineering student, but I've been doing hobby electronics for a couple years now.
I, like most people here, am building a power supply from scratch mostly because I have a few of the components already lying around and I don't have a super great power supply to begin with and I can build it to my specifications and features that I want, like switched outputs on every voltage rail

(not something I see much on power supplies these days)
so here are my main design goals:
--3-4A output current
--adjustable to 1.5-20V give or take. This is fairly loose
--achieving 0V isn't a high priority
--very low ripple at max current output
--low voltage drift
--most of all: SIMPLE
Contraints:
--I have 3 multi tap transformers from old cash registers they each have a 25V,15V and 12V secondaries ( I think... its been a while since I measured. it's 25V for sure though) , no idea what the current limit is, its not labeled, so I'm giving it a conservative value of 4A, I can include pics if requested. The fuse on in the socket is labeled 3A on the 120V primary side. so not sure if that would help deducting the output current of said transformer.
My final design will have fixed 12V, 5V, and 3.3V output rails. What I will be focusing on here is the variable voltage rail.
So I did some research and settled on the LM138K as my voltage regulator and the circuit is taken almost directly from the datasheet. Super simple.
I modeled it up in Multisim and played around with the filter capacitor values and such to achieve a low to no voltage ripple at 4A at max and min voltage of the regulator. Im not sure how well the virtual scope in Multisim is but I'm thinking it gave me a pretty good ballpark. This seems like a great solution for me and my budget, only problem is navigating around all the fake LM138k chips. First question: Whats the average going price for one of those and where could I reliably find a genuine one? I'm finding them from $8-$60 USD so it's quite confusing.
I recently came across this circuit :
(not sure if I'm allowed to post links like this so be kind and give a guy a break its my first real post)
[url]http://electronics-lab.com/projects/power/001//[url]
This circuit meets my requirements, though current limiting is not essential it would be nice to have.
My only concern is the stability and voltage ripple at max load. It is also not nearly as simple as the LM138k design.
So on to my main question.
I am wondering which of the two power supplies would be the most stable, accurate,
and most of all easy to implement. I don't have ready access to board etching supplies/equipment so I'm mostly stuck doing strip board or point to point.
Also I have a pair of very large Mallory type CGS electrolytic capacitors
labeled "10000MFD 100VDC
POS + 85degrees Celsius "
These caps have been sitting in a box brand new, never used for over 10 years.
They look brand new to this day so I am planning to use them in my design for the main filter cap after the bridge rectifier.
My only concern is that if these caps need to be reformed.
I have read that electrolytic caps need reforming if they have been sitting for long periods, unused, to work at peak performance.
I personally do not have a power supply that goes to a high enough voltage to properly reform them for my purposes.
Would any of you know how I cold get this done? Is reforming even necessary?