variable DC power supply

[flippityflop1]

i've always wanted a variable power supply; not just for testing... possibly powering my other projects, some of the electrochemistry that i dabble with, charging my lipo and lead acid batteries, etc, etc...

problem is, good ones are in the $90-$150 range (even on ebay) and the cheap-ass as i am, i'd rather build myself one, as i think i've salvaged/collected enough components throughout the years.

i first set my eyes on the 0-30 VDC STABILIZED POWER SUPPLY WITH CURRENT CONTROL 0.002-3 A (http://electronics-lab.com/projects/power/001/index.html) from the project section... so mods, no need to move/delete this thread as this also concerns a project from the project section. originally wanted a SMPS, for efficiency (like i said, it's for extended use), but seeing as very few good design are posted online and for complexity issues, i'll just settle for a linear one, i guess.

anyways, it's simple enough and actually just as good as the ones i looked through ebay. i then came across a user called xristost's modification from the very first original project:

http://diyfan.blogspot.ca/2013/03/adjustable-lab-power-supply-take-two.html

so i was just about settle on his ver 2.

but when i started looking at other sites, there are designs with larger voltage/current specs... there's one that's 1-37V @ 30A max (no current regulation, pretty common to psu project out there, but i guess i can just rip off the current regulation from the 0-30V project in the project section).

i really don't have the time to look at all the good psu projects out there, so i would have to ask.... are there any better variable DC linear power supply designs out there in the web?.. with similar complexity as the one in project section (i.e. same expense to build)? and, by "better" i also mean a larger maximum for voltage and current, and more efficient.

as for the transformer, i have this AC welder -- i've posted online over and over that it's a DC welder... then i actually dared measure and  it's AC. by it's price, i suppose it's just an oversized transformer. quite lately i've been finding more use for it...

there's 2 mode for it: 25V @ 50A and 23V @ 70A. (i measured mode 2 once and at no load it was about 33V.  only dared once coz i thought it might fry my multimeter... it didn't... duh, it's only 23V)
says max output is at 40V, doesn't say the amps.

EDIT:

just remeasured, mode 1 is 34VAC and mode 2 is 38VAC without load.
also, these 50A and 70A can't be sustained very long, that's why there's a cycle limit for both. but i guess it's fair to say that at mode 2 the welder can sustain extendedly around 30A, which is more than plenty enough for me.

 

[flippityflop1]

btw, why isn't a PMIC (Power Management IC) a better solution other than using the more complex designs found on the web that doesn't use PMIC? is it heat dissipation, lower ripples, better control, what?

 

[xristost]

It's always amaze me how people want more and more amperage and voltage from a variable PS and most of the time they haven't any idea for what purpose they want that.

Any way, I would suggest you to get finished module like this: http://www.ebay.com/itm/ZXY6010-DC-Constant-Voltage-Current-Power-Supply-Module-60V-10A-600W-/120963323698?pt=LH_DefaultDomain_0&hash=item1c29f9d732

I hardly believe you would find simple circuit with your requirements.  

 

[flippityflop1]

It's always amaze me how people want more and more amperage and voltage from a variable PS and most of the time they haven't any idea for what purpose they want that.

i just like pushing the envelope, even if i hardly know anything about the theory (software, not hardware, really is my specialization).

alright, i'll be building your ver 2...

i'd like to have 2 different outputs, with that, i suppose that you can always share the rectifier. but, i was wondering what other sections can be shared if i were to do so?

 

[xristost]

i just like pushing the envelope, even if i hardly know anything about the theory (software, not hardware, really is my specialization).

alright, i'll be building your ver 2...

i'd like to have 2 different outputs, with that, i suppose that you can always share the rectifier. but, i was wondering what other sections can be shared if i were to do so?

Maybe the regulators (LM317L and LM337L) for the opamps can be shared.
And for the two channels you must double the C1 capacitor.

 

[flippityflop1]

Maybe the regulators (LM317L and LM337L) for the opamps can be shared.
And for the two channels you must double the C1 capacitor.

i'll be purchasing 2 of each components, maybe i'll

1.) build a second output or
2.) maybe i'll find a way to increase amperage (wherein past a level, the voltage has to start going down, or increase voltage (wherein past a level, the current has to start going down)... there is only as much power as i can draw, even for my over sized  transformer...

whichever one is easiest for me...

 

[flippityflop1]

i know this is a stupid question, but i'll go right ahead...

i understand that's IC1A, IC1B, and 2N3055 mainly are regulating the voltage and current, can i swap them in the future with switching mode components/assemblies to increase efficiency??

 

[xristost]

i know this is a stupid question, but i'll go right ahead...

i understand that's IC1A, IC1B, and 2N3055 mainly are regulating the voltage and current, can i swap them in the future with switching mode components/assemblies to increase efficiency??

Go and see this video:
I think that theoretically you can replace output circuitry with switching mode regulator.

And if you don't care much for constant current mode, then there are tons of different switching mode regulators like LM2596 or LT1074 for example. They are very efficient, with simpler schematics, have all sorts of protections build in and may suit your needs better.

 

[flippityflop1]

thnx xristost!

i'm thinking more of using 3 different PCBs for this design rather than print it all in 1.

1) rectifier, C1 section -- this plugs in directly to my oversized transformer and goes to a separate enclosure box i'll be putting the main PSU. this is done so that it can be reused for any future projects... i hope my 12mF capacitor can handle any general purpose rectifying

2) LM317, LM337 section -- like you said, xristost, these are the reusable parts, if i want to add several more outputs.

3) TL082P A/B sections and 2N3055 -- so i MIGHT be able upgrade them to a switch mode regulators someday. but i think i'll concentrate more on increasing output first (i'm hardheaded)


the only problem is how to segment the sections and which components to include with it.... obviously, i don't know which components supports which section. i don't know how to group them. what if a component is shared within the 3 sections i outlined?? this is where i need help.

 

[flippityflop1]

the first section, the rectifier, since it's gonna be general purpose, i decided to be a capacitor-input filter.

i already decided on 12mF on C1, since that's what i already ordered online (and the best i could get for a hefty $16, EDIT: but worth any penny! digikey, i love you! i'm sucking up coz you're my single go to place for electronics and your prices are great!).
i have a 470uF capacitor, which i salvaged years ago, so i'll use it for C2.
using the resonance formula:

L = 1/((2*pi*60 Hz)^2 * 4.7 * 10^(-4) F )

L = 14.97 mH.


i'm going to be using a toroidal inductor whose core is a permanent magnet, with the magnetization going with the current (right hand rule, you did again!). now... to get that core.... i have 2 more spare permanent magnet blocks that my mom hadn't "lost" when we were moving. i can get a big chip from it, then grind it to a toroidal shape. have to be careful not to break it in the process. i could water jet it, but that is god darn expensive.

also... i was wondering if i could use 2 inductors on each sides in series with C2 and the load (make sure both polarities are well-behaved)... if so, do i just divide their needed inductance to 2??

 

[flippityflop1]

i'm going to be using a toroidal inductor whose core is a permanent magnet, with the magnetization going with the current (right hand rule, you did again!). now... to get that core.... i have 2 more spare permanent magnet blocks that my mom hadn't "lost" when we were moving. i can get a big chip from it, then grind it to a toroidal shape. have to be careful not to break it in the process. i could water jet it, but that is god darn expensive.

also... i was wondering if i could use 2 inductors on each sides in series with C2 and the load (make sure both polarities are well-behaved)... if so, do i just divide their needed inductance to 2??

well, nobody replied to that suggestion, if silence means "yes", then i suppose no one's objecting. actually, instead of 2 permanent magnet toroidal core, i think i'll just share 1 core -- each polarity winded in reverse. that should save space and be more efficient.

i might also change from a toroidal core to a pot core. there should be several companies out there that specializes in general ferrite (both soft and hard) that can make me a pot core with the composition of a hard ferrite and permanent magnetization with little overhead. i'll try the company that cheaply sold me 4 blocks of magnets for $40 years ago...

so for the while i guess i'll just make the rectifier enclosure, with room to add an inductor in the future.

oh yeah, would making the inductance a multiple of 14.97 mH guarantee that the capacitor-input filter still be in phase for 60Hz??

 

[flippityflop1]

oh yeah, would making the inductance a multiple of 14.97 mH guarantee that the capacitor-input filter still be in phase for 60Hz??

instead of N*L, it's L/(N^2) and L = 3.742mH. sorry. my bad. can't even do simple algebra.

 

[remonx6]

variable DC power supply used in variable resistor. perhaps...Thanks
Author