FAQ

Home > Support > FAQ

One Nicked Cable, Nine Dead Feeders: Rethinking 24 V DC Protection

Author: EFFSEN AUTOMATION | HONGKONG | Date: 2026-11-19

Take a conveyor panel that is not unusual in any way. A 20 A switch-mode power supply feeds nine 24 V DC branches: two valve islands, four sensor rails, a safety relay group, an HMI and a small drive. Each branch has its own device circuit breaker on the DIN rail, all of them 6 A thermal-magnetic types, and the whole arrangement has worked for years. Then a forklift clips a cable tray, the insulation on one 1.5 mm² feeder is crushed against the tray edge, and instead of one dead valve island the entire panel goes dark.

The arithmetic explains it. That feeder runs about 25 m to the valve island. Copper at 1.5 mm² is roughly 12.1 Ω per kilometre, so 25 m out and 25 m back comes to about 0.61 Ω of loop resistance, before the supply impedance and the return path through the PSU are counted. At 24 V that puts the prospective short-circuit current somewhere around 35 to 40 A on paper, and rather less in practice once the power supply starts limiting. A 6 A thermal-magnetic breaker of the 2210 type needs a multiple of its rating, typically eight to twelve times, before the magnetic element fires instantly. Forty amps is under that threshold. So the breaker waits for its thermal element, the PSU folds back to protect itself, and every other branch on the rail loses its 24 V at the same moment.

What the second version of that panel looked like

The fix was not a bigger breaker. It was a different trip principle. Each 24 V branch got an E-T-A ESX10-TB-101-DC24V-6A electronic circuit protector, still 12.5 mm wide, still on the same rail, still fed from a common LINE+ and 0 V busbar. The ESX10-T does not wait for fault current to build. It limits it actively to roughly 1.8 times its rating, so about 10.8 A in this case, and then trips on a short circuit in typically 100 ms. The power supply never sees the fault. The other eight branches keep running while the maintenance fitter walks out to the tray with a torch.

The overload side matters just as much. The ESX10-T is specified to break reliably at 1.1 times its rated current, 6.6 A here, which is exactly what you need on a long run of 1.5 mm² cable where the fault current is small but the thermal stress is real. The same device will start a capacitive load up to 75,000 µF without tripping, so the 24 V LED drivers and the DC/DC converters on the sensor rails no longer nuisance-trip at switch-on, which was the reason the original panel had 6 A breakers instead of 4 A ones. A status output carries the trip to the PLC, and a manual ON/OFF button lets a fitter isolate one branch without pulling a fuse.

Two details are worth flagging before copying this. The first is the ambient temperature on the rail. The ESX10-T family is rated 0 to +50 °C, and a closed cabinet in a conveyor gallery in August can sit above that. If the panel runs hot, the honest answer is ventilation or a device with a wider published range, not a bigger nominal current. The second is the failsafe element inside the electronic protector, which is matched to the current rating and to the cross-section of the cable it protects. Changing the cable from 1.5 mm² to 2.5 mm² on a branch without checking that pairing throws away part of the protection you paid for.

None of this makes the thermal-magnetic 2210 obsolete. On the AC side of the same panel, feeding a 400 V servo drive or a 230 V heater group, the 2210-T2 range still does the job it was designed for: one to three poles, 0.1 to 32 A in fine steps, AC 250 V and 3-phase 433 V, breaking capacity of 800 A above 6 A, and a trip-free mechanism that opens even if someone holds the toggle in the ON position. Device breakers belong on AC loads and on DC circuits where the fault current is genuinely high. On a long 24 V run with a current-limited supply, they are being asked to do something they were never built to do.