pass transistor in regulated supply?

[theAnonymous1]

I need to build a regulated supply for 35v @ >5A and would like to use parts I already have.

I plan on using a LM317 with a pass transistor, but I only have NPN 3055's and not any PNP 2955's.

I have found this diagram of using a 3055 instead, but it looks a bit over simplified and I would like to know if it would even work or if theres anything else that needs to be taken into account.

Let me warn you I am about as novice as one can get.

http://www.diyaudio.com/forums/attac...amp=1118719729

[richie00boy]

That would work in a fashion, it's just making an emitter follower where the output voltage will be about Vin - 0.65V. This would probably be better than the usual zener diode voltage reference as it can source enough current easily to drive the hungry transistor base.

[theAnonymous1]

Thanks richie00boy, I'll give it a go then. I have enough parts to experiment and not worry about blowing them up in the name of progress.

[theAnonymous1]

OK, I tested this with a 7912 and 3055; no other components, and it works great. The 7912 stays completely cold and the 3055 heats up as expected.

Is there any disadvantaged to doing it this way over the conventional method of using a 2955? This is so much more simple I can't see why you would want to use a 2955.

Also, if I wanted to parallel more 3055's do I need to put a current sharing resistor anywhere?

Thanks

[BWRX]

Ideally you'd want to put very low value resistors in series with the emitters of the transistors so they share the load current.

A neat trick would be to add those resistors in series with each emitter then connect the LM317s feedback resistor to that new output node.

That approach has some nice benefits.

[jackinnj]

PNP's are used in low-drop out regulators, NPN's are used in standard regulators. In the low-dropout version you burn fewer watts -- but the output capacitor has to be selected more carefully.

Consider the LM317 as a "error amplifier with reference" -- this is what BWRX is suggesting in essence.

you can get away with an LM317LZ which is a 100 milliamp TO-92 if you use 2 of your pass devices in Darlington configuration --

[theAnonymous1]

Thanks for the help guys.

I have lots of LM317T and LM317LZ , but I think I'll stick with using LM317T to keep things simple... I need simple.

There are lots of examples of using a 2955 as a pass transistor for a LM317, but the only example of using a 3055 I can find is that simple diagram found on this forum.

[BWRX]

Originally posted by jackinnj
Consider the LM317 as a "error amplifier with reference" -- this is what BWRX is suggesting in essence.

That's what I meant to suggest, but the way I described it in my last post will not work. I keep forgetting that these regulators mainting a constant voltage between the output and adj pins which causes a current to flow through the upper resistor and also the lower resistor which sets the reference voltage.

Attached is the basic version of what you want to construct, Mr. Anonymous1

The 100V was arbitrarily chosen. Obviously do not use an input voltage that high or parts will fry.

[theAnonymous1]

Thank you Mr. BWRX!

[BWRX]

This variant I just came up with simulates better (steady state or constant loading) in terms of line and load regulation.

The values in the schematic should give you about 35 volts output with a lot of current capability.