Bench Power supply 0V ~ 15V CC/CV

[Scarionn]

I was wondering about build this power supply.
Do you have any suggestions for improve the design?

R03/R09 are pots.

[Rerouter]

Edit: silly fault, i now see how your doing it with a voltage doubler,

[sync]

At powered off the -5V rail can collapse before Vcc. This can cause an overshoot at the output because the LM317 adjust pin can't pulled negative any more. I build a power supply with this kind of negative rail in the past. It overshoots and fried my projects 

[Paul Price]

That one seems too complicated and has some overshoot problem. Try this one.

[steve30]

Scarionn, That's pretty similar to what I have built and it works reasonably well.

There is of course the overshoot issue mentioned. That can probably be fixed some how.

Only problem I have is that the switching action of the transistor (when current limiting) can cause the output to be noisy. So probably not perfect as a current source, but would be fine if used more as overcurrent protection.

Is the sideways LM317 just a current source so the supply has a constant load on it?

Build it and see how well it goes

steve30.

[Paul Price]

From the schematic you've posted, the circuit will barely work at all. R04 will limit the output current to a few mA.

[steve30]

Paul, are you sure? The sense resistor shouldn't dissipate too much power, although to be on the safe side, I would go for one bigger than one watt. Not sure how else R04 could be a problem.

I'm curious also to know of any improvements anyone can think of for this, considering that my design if very similar.

[Paul Price]

I misread the value, it looked to me at too fast a first glance to be 1K, not 1-ohm (1R).  At 1.5 A output this resistor would burn up, or change value to cause current regulation drift, if only rated at 1W but asked to dissipate 1.5W at full output current.  But why not use .01 ohm?

Your design is very similar to what other design?

[steve30]

Looking closely at R04, it looks like 1R, not 1k.

I mean my design is similar to the one in the first post by Scarionn.

[Paul Price]

Ok, I could not see the schematic while composing a reply.

There is a very good chance of only getting 10V out of this supply?  Is that enough?

The U02A op-amp could easily be burnt out by a voltage reversal on the output.

C10 100uF would severely lower the response time of the constant current or current limiting to kick in so slow as to fry your external circuit but likely would protect your power supply from overheating.

What do you think of my much simpler design?