Opening the circuit as soon as it reaches 1.2 amps???

Well, a fuse will work. But both a fuse and a circuit breaker take time to operate.

One option is to have a current sense resistor with a transistor across it. Correctly configured, this will turn the transistor on when the voltage drop reaches about 0.65V.

This can be used to trigger an SCR which can be used as a crowbar to blow a fuse or to pull something down to hold the power supply off.

This has the advantage of having a very sharp trigger point compared with a fuse or circuit breaker.

I'm not particularly fond of crowbar circuits. If you have another alternative, it is best to use it.

 

Have you checked the specifications for the fuse? You might find it takes a minute of more to blow at just over its rated current.

 

I still urge you to find the datasheet and look at it.

1A fuses do not blow at 1A. Here is a sample datasheet for a fast blow fuse. A 1A fuse requires more than 1.5A to blow, requiring 1.75A in this case to blow in 1 second.

These figures may vary significantly for another fuse type.

 

From what I can tell, they're slower acting than the ones I pointed to. They require twice the rated current to guarantee to blow in 1 second.

Also note that they *may* blow in 4 hours at the rated current.

Here you see another factor with fuses -- fatigue. They will eventually blow at a current lower than you expect if you run them near their rated limit.

This is a reason why you typically gave a fuse rated 2 to 5 times the expected current in a circuit.

None of this means that a fuse is inappropriate for your application, however it doesn't match the requirements you initially laid out.

 

That looks very clever.

I once designed a PSU using a 723 which used a similar mechanism, although it used a real SCR and the sense resistor was in the ground return. It shut off the 723 by pulling a comparator input to ground or something similar.

It didn't actually limit the current though, so the 723 ended up dead (due to it's limited ability to provide base current).

That was in 1980 or something around then, so forgive me my lack of subtlety and poor reading of datasheets.

That circuit is a lot better. I wonder if you could get away with a smaller voltage drop across the current sense resistor if you used a discrete differential pair (two jfets perhaps?) I guess matching would be an issue.