I am attempting to limit the current to a circuit card to 2A. I initially intended to use a fast acting 2A fuse, but none that I am finding are very quick; it will take anywhere from 30s-1hr to blow @2A. It will blow much faster at 4A, but I do not want to allow 4A into my circuit card.

What is the best way to limit the input of the circuit card to 2A? Do I get a 1A fuse that blows quickly at 2A and just hope that it won't be over 1A for an extended period of time? Do I got with a current limiting circuit?

 

What would happen if/when your circuit wants 3A but is throttled to 2A? In other words, the answer depends on the application.

 

What would happen if/when your circuit wants 3A but is throttled to 2A? In other words, the answer depends on the application.

With the fuse I had previously picked out, it would take up to 5 seconds to blow at 3A. I need it to blow whenever it passes 2.

 

I can state with reasonable certainty that a fuse is not the way to go. A pass transistor in a common base configuration will give you the speed you need.

 

Well, you could use a comparator to watch current by way of the voltage across a shunt resistor, to shut it all down quickly if the current goes over whatever level you want to set. Or, you could use a similar approach to throttle the current without shutting it off. Depends on the circuit needs.

 

I can state with reasonable certainty that a fuse is not the way to go. A pass transistor in a common base configuration will give you the speed you need.

Thank you for this reply. I will look into this, but from what I have seen so far, it seems as though it will fit my needs.

 

Below is a common transistor current-limit circuit which responds very rapidly.
The current limit is approximately 0.65V / R_sense.
You may need to heat-sink Q1, depending upon how long it has to limit the current.

 

What do you want to happen when the current exceeds 2 A? In crutschow's circuit the voltage will decrease as the current increases above 2 A. This is constant current limiting. Another option is current foldback, where the output current snaps down to a much lower value and is latched there until the circuit is reset (usually by cycling power to the system). Third is an electronic circuit breaker - current shunt, comparator, latch, pass transistor. These can respond in microseconds if necessary, and the off state current is microamps. Linear Tech and Maxim make parts specifically for this, including some that retry periodically, a self-resetting electronic circuit breaker. One of my mil designs has one in it, very reliable performer. Or you can grow one from the usual suspects.

ak

 

Below is a common transistor current-limit circuit which responds very rapidly.
The current limit is approximately 0.65V / R_sense.
You may need to heat-sink Q1, depending upon how long it has to limit the current.

I know this is an old thread but I would like advice on selecting the value of R2 in this circuit. Does this resistor give us "foldback" instead of just current limiting?

 

I know this is an old thread but I would like advice on selecting the value of R2 in this circuit. Does this resistor give us "foldback" instead of just current limiting?

Just current limiting.

 

Just current limiting.

How do you select a value for R2?

 

How do you select a value for R2?

It just needs to be small enough to turn on Q2.
Start with the current needed thru R! to turn on Q1 and work backwards.

 

It just needs to be small enough to turn on Q2.
Start with the current needed thru R! to turn on Q1 and work backwards.

Okay, so voltage across R1 minus E-B of Q2, select a current that gives me the collector current I want in Q2 considering the gain of Q2?
We are no longer limited to the 700 mV or so of the E-B of Q2? R1 voltage can increase beyond that. So what does this accomplish over R2 just being a wire?

With no R2, just a wire, if we try to draw more current our current stays at the limited value (about) and the output voltage starts to drop.

With "Foldback" out output current decreases ??? or is it voltage that decreases??? What alteration in the design gives us foldback?

 

Look at the datasheet for an LM723 chip

 

Okay, so voltage across R1 minus E-B of Q2, select a current that gives me the collector current I want in Q2 considering the gain of Q2?
We are no longer limited to the 700 mV or so of the E-B of Q2? R1 voltage can increase beyond that. So what does this accomplish over R2 just being a wire?

With no R2, just a wire, if we try to draw more current our current stays at the limited value (about) and the output voltage starts to drop.

With "Foldback" out output current decreases ??? or is it voltage that decreases??? What alteration in the design gives us foldback?

You want Q1 to be turned on hard so you don't drop to much voltage across it. Usually a gain of 10 is used, so 200 ma. of base current.
The only reason there is an R2 is if there is a short at the output before Q1 can turn off there can be a lot of current thru Q1 and then the base emitter of Q2. It all depends on the parts you use so it's usually put in there just in case.
Usually with foldback both current and voltage decrease.
What would you like the circuit to do?