So I periodically need to piece together my electronics lab in my garage with whatever I can get/build/fix because I'm too cheap to spend tons of money on expensive (and greatly more accurate) versions of these tools. It's fine and all because it's not like I depend on this stuff for income, but rather for fun.
Every proper EE lab needs an adjustable power supply. I had one.... like probably 15 years ago. It was cool because my uncle bought me a kit in Taiwan and after I assembled and tested it, my dad helped me make a metal case for it complete with a volt/ammeter on the top. 0-50Volts. I think either up to 3 or 5 amps. Built in overcurrent protection!
I used it to power anything and everything I built. And because I was young (probably in elementary or middle school), I also did stupid things with it like connect up some aluminium foil and make sparks. Lots of sparks. My own personal fireworks show. And then it died.
Sometime later, my uncle bought me the same kit. And I built it, and it didn't work. But for some reason I didn't understand why, nor did I take the time to figure it out. So I set the whole thing aside for over a decade.
In the recently months, I've needed an adjustable power supply again. I've been getting by with little adjustable boost and buck converters linked to battery packs, but I blew out both of the buck converters. Besides, they're kinda weak.
I dug out my broken old kit supply and tried to see if somebody had a similar schematic for the LM723 and didn't find one. I tried tracing from the output to find where the 50V was coming from. I disconnected the fat power resistor to see if voltages returned to normal in most of the circuit (and it did.)
I tested a bunch of the semiconductors in place and then pulled out all the transistors.
I discovered the following:
1) I thought the transistors had a pinout (while looking at the flat face) of EBC. They didn't. Playing with the meter, I determined them to be CBE (if I recall correctly). So I tested them all and inserted them back in. And the overcurrent LED's reset now responded correctly. (doesn't mean I know if overcurrent protection works though...)
2) I found that the 2N3055 power transistor had shorted out. This explains why 50V was going all over the place.
3) Removing the 2N3055 leads me to a B507 (2SB507) transistor which was driving it. Things were looking up when I saw an output that varied based on the voltage knob. So I tied a load to it and played with the voltage knob. BANG! A bit more testing leads me to believe I just blew up the B507.
On the upside, tweaking the voltage knob shows an varying voltage on one of the pins leading to it, which means things should be fine if I replace these two blown transistors.
I'll make a trip to HSC sometime this week.
In the meantime, I came across another idea. Converting an old AT power supply to also be an adjustable power supply. After cracking open two power supplies, one appears to have no PWM logic (weird), and the other has a KA7500B. Googling for schematics, I came across some close enough (but definitely not identical) ones. Finding that other people have thought similar ideas, I felt more comfortable with the idea of messing with the PWM.
By strategically tying a variable 10k pot into the reference input (pin2), I was able to get the voltage to vary somewhat. I'll need to figure out what my bounds are, but I expect that a 100k pot being used as a voltage divider to the reference ought to be fine. I'll try this later.