DIY Power supply

[dio123]

I want to remake my power supply, improve the voltage adjust, and put current adjustment. Besides going for voltmeter and ammeter digital occupying less space than analog.
The features for my new power supply, are very simple. The source 0-15V 2A, stabilized, with adjustable current and voltage.
Will be to work with atmegas, build circuits in beardboard, basic things for beginners.

However I would like to know some circuit that recommend me to use.
I wanted to see if he used the LM358 I have to use the source if possible.

I thank you

[Paul Price]

Check it again.

[dio123]

edited the first post, please read

thank you

[herctrap]

http://www.electronics-lab.com/projects/power/001/index.html

[niflheimer]

You can try Dave's design as a starting point since he uses less discrete components than the psu herctrap linked. You can find it here :

I plan to improve mine soon as well so keep us updated please !

[logman]

thanks for the link/video

[niflheimer]

I think he linked a schematic in the fifth or sixth video of the same series. It's got a lot more components but it's got everything you need to interface it to some nice digital meters and for arduino/ other MCs.

[geraldjhg]

hi
use 338, replace voltage pot with 10 turn
keep analog ammeter (connected before the regulator)
use digital display for volts but can be ommited
add extra fixed regulator for +12 +5 and nowdays 3.3
current control is another story
it involves sense resistor and opamp and pass transistors
and a way of showing in what mode its operating in
not worth the effort
usefull add on:
a lm317 in current mode an a switch to select 1ma 10ma 100ma VERY usefull
saludos

[dio123]

Good afternoon,
I'm with ideas to build this circuit, because I have almost all the components, but would like to know your opinion.

circuit:  https://meocloud.pt/link/53912182-9ce7-4571-8b0d-af202f95e031/Lab%20Supply%20Final.PDF

thank you-

[codeboy2k]

Its pretty good actually. It has current control, voltage control, uses proper shunting of the base drive, has over temp shutdown, parallel tracking mode... everything you could want in a power supply.

However, I think maybe the designers had stability problems and couldn't get a good phase margin in the control loops, (or they didn't even measure it) but it's probably stable as drawn. I say this because they have put .01 uF caps across all the op-amps, from output to inverting input. This adds an output zero and slows down the control loop to the point where stability is not an issue. It probably has a poor (slow) load step response because of it. The output A and output B also has too much capacitance, there is a 5000uF on each output.  And that just tells me again, that the feedback loop is probably unstable with load changes, so they put that excess capacitance there to slow down the loop, dominate the load capacitance and stabilize the loop with those 0.01's everywhere. 

I'd personally want a PSU with less than 100uF of output capacitance. But if you build it and take that 5000uF out, it will likely become unstable and oscillate with just minor changes in the load. 

[dio123]

Good afternoon,
I made a power supply with cc and cv in LTSpice, and is working as I had planned.
After I mounted the breadboard circuit, and did not notice any problem.

Since I am a beginner in electronica, I would like to know opinions and suggestions to improve my  power supply.

I leave here the files from my power supply.

my power suppply:  http://cl.ly/1a3z2K342E2y

pics circuit and graphics
1--http://cl.ly/image/2x3j0O3w3b1x  2-- http://cl.ly/image/2M1d0H2s253W   3-- http://cl.ly/image/1e381l3s0R42

Thank you for your help.

[David Hess]

Your current shunt degrades the load regulation.  The frequency compensation when current limiting is uncontrolled (Q1 is a problem) and oscillation or at least very poor transient response is likely.

Things I would do for a simple lab power supply:

1. I would minimize the output capacitance to limit total energy under fault conditions.  This implies using an emitter follower output which is more stable but precludes low dropout regulator designs to some extent.

2. For relatively low power designs, I would use an integrated voltage regulator as a pass element because that provides thermal shutdown, safe operating area protection, and instant current limiting.

[Teledog]

Hate to be a "stick in the mud", but I've built many a variable regulated supply in my time..but for your time & cost, just BUY one!
http://www.leeselectronic.com/index.php?id_category=1952&controller=category
Can get the same units on eBay..for probably much less..
Set the current for overcurrent/shorts/continuous regulation (ei: charging batteries), etc.
Have both analog (for pulsed circuits) & digital units..no, not Agilent, but in the long run,  they essentially cost less than making one yourself.