Using Regulator ICs In Parallel to Increase Output Current?

[swuster]

Can I use regulator ICs in parallel in order to increase the output current? I don't see any logical problem with it, though I may be missing something. Thinking of using L4902A but its output is maxed at 300 mA. Can I just string a few of them in parallel (adequately heatsinked, of course) to push my current up a little higher?

 

[berkeman]

Can I use regulator ICs in parallel in order to increase the output current? I don't see any logical problem with it, though I may be missing something. Thinking of using L4902A but its output is maxed at 300 mA. Can I just string a few of them in parallel (adequately heatsinked, of course) to push my current up a little higher?

The main issue is that regulators are pull-up devices only. So whichever one has a slightly higher setpoint within tolerances will try to source all of the current. As long as they all have a simple current limit mechanism (as opposed to fold-back current limiting, which many linear regulators have), then yes, they could be directly paralleled.

To parallel power supplies, you generally need to do something, either to address the current limit/sharing aspect, or to address stability issues (usually in switching power supplies). What is the internal circuit like on the L4902?

 

[swuster]

Actually, looking at the block diagram, the current limiting is not very robust at all - I may just end up using another chip altogether. It's essentially just two linear regulators and a timing circuit; the only reason I chose this particular one was for the added convenience of having the two outputs. In either case, could I just route it through separate current regulator diodes? I feel like the regulator with the higher setpoint would still try to source the current and burn out the diode or something.

 

[Maccara]

Why not just use a transistor/darlington with the regulator to handle the load?