Electronics > Projects, Designs, and Technical Stuff

Circuit Peer Review: High Current LM317 CV Power-Supply

(1/6) > >>

Stephen Hill:

I'm still learning electronics and in order to improve my designs it's about time I got some feedback from the community.

Today I've whipped up a design for a high current LM317 power supply and I would very much appreciate constructive comments, criticisms and suggestions.

It's essentially a very basic circuit: transformer, bridge rectifier, filter caps, LM317, power NPN's. The voltage range will be 0V to 24V and I don't expect to use any more than 4A. A micro and LCD could probably be added to measure voltage and current. Here is the schematic:

The notes in the schematic say:

» D6 Discharges C4
» R1 Discharges Filter Caps
» R2 Limits Inrush Current into LM317
» R3 500ohms 22 Turn Trimmer
» R4 5K Potentiometer
» R5 Minimum Load, 1mA at 24.576V, 24K + 576ohms in Series
» R6 Current Shunt

I look forward to reading your replies :)

Many Thanks

My 2c.
Basically the circuit looks more or less OK. Regarding the voltage setting resistors of the 317 I would check the literature, so that you don't end up bending the rules out of shape. Also you might want to think some kind of range limiter resistors in series wiuth the pot. I am not sure if the regulator likes a short in the divider. But then again it may - can't remember offhand. Its been a while since i played with 317s.

Please throw the 2N3055s out and put in something invented during this millennium. The days of those dinosaurs is truly past. Something starting "MJ..." would be the thing to specify in my opinion. Or why not just use a Darlington in the first place.
Finally, you seem to have a really exact spec for the ballast load resistor. Do you really need it to be that precise? If you need a milliamp, OK but surely 1,2 mA or whatever you would get from one E12 series component would do as well.

probably... and most likely... your LM317 is sensing its output... not the "true" output. so if something whacky going on to the darlington, lm317 will not aware of it. and blimey, if the darling going whacky, its not pretty, esp when you have precious gem on the "true" output. my bet is put the R3, R4 et al at the "true" output.

D2 and D4 aren't rated high enough.  They need to be rated the same as D1 and D3.  All of the current that goes to your circuit's ground is going to go through either D2 or D4.

With the current circuit configuration, D5 and D6 don't need to be nearly as beefy.  Perhaps you swapped them with D2 and D4?  Right now the only current they will see is what gets discharged from C3 or C4.

Finally, like Mechatrommer pointed out, the LM317 isn't in the overall feedback loop so it isn't truly regulating the output.  The LM317 is really just serving as a voltage reference for the base of the Darlington pair in this configuration.  You might consider basing your circuit on one of the 'high current' regulator circuits that can be found in the LM317 datasheet.

Oops missed the diodes. Never crossed my mind to check that - why would they be different since they see the same current...
The comment about the feedback loop sounds right. Reason why i said nothing is that i seem to recall a paper that recommended doing it like this instead of directly from the output. There was some other reason except accuracy of regulation but i cannot remember what it was and can't locate the paper to check.


[0] Message Index

[#] Next page

There was an error while thanking
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod