I think this has been hashed-out successfully - use a breaker with a DC rating. DC is harder to break, since there are no zero crossings to extinguish the arc.
However, I thought I would relate a relevant incident that happened at my first job at Intel, around 1978. I was a design engineer on the 8089 I/O processor which followed the 8086 design by about a year. (It never became very successful, but the 8086 - that's another story). At that time we prototyped microprocessors in a TTL breadboard to test & debug functionality before first silicon came back. This was in the days before logic simulation. I got called into the lab by the 8086 lab technician, who wanted me to look into a problem with the 8086 TTL breadboard. This was built with multiple Augat wire-wrap boards that were mounted in a tall rack cabinet. At the bottom of the cabinet was a shoe-box sized 5V 200A switching power supply with big (I guess about 8 gauge) insulated wires running to each Augat rack assembly. The complaint was that the breadboard had stopped working. Putting a scope on one of the bus lines showed a logic high of about 1.5V. The VCC was about 3V. As I was looking around for the problem, I started smelling something hot. I looked down towards the power supply at the bottom and saw that the +5V wires coming up had no insulation (it was burned off) and were glowing a deep red color. Tracing the problem, one of the +5V wires had shorted to the rack cabinet due to an over-enthusiastic cable clamp. At the lowest rack position was a panel with a single household circuit breaker in line with the +5V, rated at 30A, if I remember correctly. It was somewhat charred, but was otherwise happily passing hundreds of amps from the massively over-sized power supply. I asked the tech about this, and he thought that a 30 Amp breaker would work just fine. Two problems here: Don't use a power supply that is almost an order of magnitude bigger than needed (the bread board used less than 30 Amps), and no, a 60Hz 120VAC circuit breaker doesn't work on 5V DC. Putting in a smaller power supply with current limiting and fixing the short brought the breadboard back to life.
- John Atwood