Latching Current Limiter (resettable fuse if you will!)

Posted by on April 8, 2010

As my first electronics-related entry, I would like to share a modification of the “current limiter add-on” by Ron J. The original circuit works as advertised, however I needed to disconnect the load immediately after the maximum current was reached, what I needed was pretty much a resettable fuse.

The following schematic depicts this concept:

With only a few extra components you can make a trusty yet simple electronic fuse.

( This schematic is also available in PDF: LatchingCurrentLimiter )

Should you attempt to draw over ~500mA, the circuit will disconnect the load. Even if the load no longer attempts to draw half an amp, it will remain disconnected until the reset button is hit or the load is physically reconnected.

BC337 was chosen in this case but any general purpose or medium power transistor will suffice as long as it can handle at least ~150mW (worst case scenario someone glued the reset button and Q1 blows). All resistors 1/4W 5%, the shunt resistor should be at least 1/2W and 1%.

(If you can’t find 1 Ohm resistors, make your own by winding thin magnet wire onto a 1 meg resistor)

The latching behavior is possible thanks to Q2. It’s job is to pull the base of Q1 to ground, this effectively allows us to turn the transistor off. Now that it ceased to conduct, the “OUT-” lead is no longer tied to ground, therefore the LED is allowed to conduct. R5 will now provide enough base current for Q2 to remain active. Discharging the capacitor through the reset button unlatches the circuit. ie. Now that R5 is tied to ground, the base of Q2 no longer receives any current and therefore Q1 conducts again.

I must thank the regulars at ##electronics on freenode.net for the invaluable help provided!.

If you find any mistakes or would like to enhance the circuit, feel free to do so! Keep in mind that I’ve successfully built and tested this; otherwise I wouldn’t be posting it 🙂

Cheers!

5 Comments on Latching Current Limiter (resettable fuse if you will!)

Closed

  1. Strynox says:

    I breadboarded your circuit and while it worked for a short period of time i found it doesnt work anymore

    after doing some testing i measured half a volt between the output leads!
    the red LED is always ON no matter how small or big the load is. do you have any clues?

    • GuShH says:

      Hi, yes chances are your pass transistor failed open. Did you, by any chance, hold the reset button while testing a load bigger than half an amp? — This would smoke the BC337 and possibly the shunt resistor as well!

      The original current limiter calls for a mid power transistor, here we’re using a low power transistor because the latch is supposed to keep it from blowing up, however if you continously bypass it by holding the reset button, then theres nothing to prevent Q1 from blowing up.

      You won’t believe this but I just built another one of these and the very same thing happened to me, Q1 smoked and I didn’t notice it until it failed to work (and then I smelled the enclosure…)

      Clearly a revision of this circuit is called for. However if you’re careful not to use the reset (or heck, don’t implement it at all) then things should work as planned.

      Cheers.

      • Strynox says:

        yes!! I changed both transistors and the circuit works again!
        no I do not have a reset switchbutton but im using a smaller value resistor for r2

        can I put two or three q1 in parallel?! I need 1 ampere at 5 volts and im afraid of killing my last two transistors. thanks

        • GuShH says:

          Cool. For 5V @ 1A you’ll need a beefier pass transistor (Q1) such as the BD135 or TIP31, however you’ll need to provide more current to the base, which means R1 would likely have to go down to 100 ohms or so. The shunt will need to be 0.5 ohms (just put two 1 ohm in parallel if you don’t have smaller resistors) and it’ll have to be at least 1/2W. This should give you a current limit of about 1A and it’ll latch off above that.

          You don’t really need to heatsink Q1 in this case. As per using two or more BC337 in parallel, you’d have to use resistors to do it properly, in the end it’s a better idea to use a proper high power transistor instead.

          I hope that helps.

  2. GuShH says:

    I forgot to mention this: the other day one of the limiters decided to blow up, it’s still not clear why it happened (since I’ve used it for months now) I’m thinking it didn’t latch in time and the pass transistor couldn’t handle the current, but why? I’ve no idea yet.

    Replacing both transistors didn’t help either, I believe the shunt drifted upwards and if that’s true, then the latching circuit needs a revision because the only way the shunt and pass transistor can go bad is if the latching does not occur in time.

    Magic blue smoke anyone?