Dual rail current regulated power supply advice.

[TheBrick]

Hello all,

I have a question regarding the current limiting par of a bench power supply I am trying to design. Please excuse the poor quality of my schematic. I'm only just getting to grips with Kicad and having problems finding all of the symbols I need. Please also excuses the poor question but this is my first foray into analog systems.

I am trying to build a dual rail bench top power supply, I am awear this has been done before but reinventing the wheel is a valuble learing experience IMO. My loose requirements so far are.

1. Adjustable voltage from 1V - 12 v
2. Current regulation, not sure of the achievable figures yet.

The power will be supplied by a centre tapped transformer -> frull bridge rectifier-> smoothing caps.
Voltage and current regulation with be achieved via LM317 and LM337 for +ve and -ve rails respectively.
I would like the current for both rails to be adjustable by one multi-turn potentiometer and the voltage for both rails via another multi-turn potentiometer.


I am initially trying to work on the current regulation. I have looked at some designs on-line and at the datasheets and come up with an idea but I'm not confident I'm heading down the correct track. This is my schematic and I will try explain  my thoughts below.



IC1 operates in current limiting mode by setting a constant voltage difference between the adjust pin and the output pin. This voltage difference is adjusted via RVAR.
RVAR is tied to the negative output rail to provide
i) a small constant current source.
ii) The reference voltage to be pulled more than 1.25 v lower than V_out so as to allow greater range on the current regulation. (I got this from the data sheet). I'm not 100% on the point of the diodes but they are in the datasheet as  I understand to provide a 2 * 0.7 voltage drop over RVAR. I'm not sure if they are needed as RVAR is tied to the -ve voltage rail

IC2 is an op amp with its positive power tied to GND, which is referenced to the centre tap of the transformer. R2 and R3 are of equal value giving the op amp a gain of -1 as V- is more +ve than V+ on the op amp. This means the voltage going into IC3 is -ve of the voltage going into the adjust pin of the IC1 which should mean the current output of IC3 is equal in magnitude to that of IC1.

I am hover not sure I am correct in the arrangement of the op amp. Nor am I confident that feeding IC3 a -ve reference voltage to IC1 is the correct thing to do as there is no feedback on IC3 any more.

In short I have little confidence that I understand what is going on. So please correct me and help explain.

Any advice, feedback or discussion on the subject is greatfull recived.

Notes on resister values.
R5 is of the order 10 * R1 so as to minimise the current flow from the output of IC3.
R1 and R4 are of equal magnitude.
R2 and R3 are of a simlare order to R1 and R4.

[TheBrick]

On top of general feedback on the circuit , is the op amp idea to control both regulators valid? I would also like some advice on what and how to choose an op amp for this job. The number available is bewildering and I don't know where to start.

[Thor-Arne]

I'm not a expert on either OpAms or psu designs, but here is one I've been looking at.

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

[LEECH666]

If you're going to build that PSU take a look at the two threads in the forum as well.

The frontpage article is outdated and the supply circuit has received a few improvements to make the circuit more stable.

http://www.electronics-lab.com/forum/index.php?topic=19066.0
http://www.electronics-lab.com/forum/index.php?topic=7317.0

Florian

//EDIT
Current parts list and circuit diagram can be found here:
http://www.electronics-lab.com/forum/index.php?topic=19066.msg1010703#msg1010703 (you need to be registered and logged in to see the pictures)

[TheBrick]

Thanks they look interesting.

I'll think I'll probably end up with something like that but I want to try some analysis of my circuit first then I'll through it together on a bread board and see what it wrong and ask some more questions, just as a learning experience. As I said, I need to reinvent the wheel to learn personally. I'll be back once I've done some analysis for people to check my work then build it up.

[Thor-Arne]

Thanks they look interesting.

I'll think I'll probably end up with something like that but I want to try some analysis of my circuit first then I'll through it together on a bread board and see what it wrong and ask some more questions, just as a learning experience. As I said, I need to reinvent the wheel to learn personally. I'll be back once I've done some analysis for people to check my work then build it up.

That's a good idea.

Please post the progress/result.