Current Limiting Questions

[pop]

Hi,

I've been thinking about this circuit for a while now and I need some help.

I am trying to achieve current limiting and perhaps a short circuit protection (I don't want to use a fuse) using the attached circuit.
The voltage regulator is LM350 and it will be used to regulate voltages from 30 V to 1.2 V. However, I would like to force the LM350 to conduct a current of 1A max in the worst case. I would like the output current to be 3A max (I am aware that LM350 can handle 3A but I would still like to limit it.).

1. How can I achieve current sharing between Q1 and the voltage regulator, if Q2 is placed in the circuit as shown? Before I decided to use a current limiter circuit, I used a small resistor instead of Q2, and the ratio of that small resistor and R_sense determined current sharing. Can something similar be done here?

2. When voltage drop accross R_sense reaches 0.7V, Q2 will turn ON and current from Q1 will be shunted through Q2. How can I prevent this current from exceding 1A through the regulator? I don't want to insert a large series resistance between Node C and regulator, because my output voltage will drop.

3. Also, I was thinking about adding a larger resistor (100 ohms) between the base of Q2 and emitter of Q1 to protect Q2 from damage during normal operating conditions of the pass transistor. Will this large resistor affect the turn-ON voltage of Q2?

I understand the npn-current limiting technique, but when I think of using a pnp pass transistor I can't seem to design the circuit.
Thanks for the help!

p

 

[dknguyen]

Have you considered using a PTC? One of them has the "Polyswitch" brand name. It's a resistor whose resistance sharply increases when it heats up too much due to current flow.

Maybe I'm a little brain fried, but what is the circuit supposed to do? What good is shunting excess current to the same output as the regulated current? Sure, you might save the regulator, but what about the load?

I'm also confused about how Q1 is supposed to turn on in order to allow current to flow through R_sense to produce the 0.7V drop to turn on Q2. Q1 seems to be biased so it's always off (or floating, rather).

Not sure if this works for BJT circuits...but for NMOS/PMOS circuits you can get the PMOS/NMOS equivelant by flipping it upside down (reversing all the voltage sources) and then replacing all the NMOS with PMOS/NMOS...at least that's what I think the rules were. The flipping refers to when the schematic is drawn in the non-looped up-down fashion.

 

[hjames]

The configuration is pretty common - pretend the Q1 is just a diode drop temporarily, and you can imagine that the regulator will work as desired. Put in the rest of the transistor and you can see that the transistor ends up amplifying the regulator's current. The regulator ends up conducting 1/Hfe of the load current, with the pass transistor(Q1) conducting the rest. Using a 3A regulator in this configuration doesn't make too much sense unless your pass transistor is a monster(30A+?).

So, back to the original assumption: why do you think that the output voltage will "drop" if you put a resistor between the voltage regulator's input and Q1/Q2? How much current do you think will flow through Q1's emitter-base into the voltage requlator under some normal load.

Incidentally in any case, I'd second the PTC recommendation. One "cute" trick is to use the PTC as the sense resistor. It'll do some pretty loopy power cycling stuff when it overloads, but it'll probably waste less power if it trips.

 

[audioguru]

Read the datasheet!
The LM350 reduces its output current to protect itself if the voltage across it exceeds only 10V. With 30V across it, its output current could be only 300mA.
The LM317 reduces its output current when it has more than 15V across it. The LM338 also reduces its output current when it has more than 10V across it.

 

[Hero999]

or is LM350 and it will be used to regulate voltages from 30 V to 1.2 V. However, I would like to force the LM350 to conduct a current of 1A max in the worst case.

Then perhapps consider using a LM317.

I would like the output current to be 3A max

Don't you mean 1A?

pass transistor.

Why are you using a pass transistor?

The whole point of a pass transistor is to boost the current capability, you want to reduce it so a pass transistor totally useless.

Connect a circuit like this in series with your reglator to limit the current.