Holding Brake

The holding brake is a spring-applied, electromagnetically released friction brake bolted to the tail of a servo motor. Springs engage the brake by default; a 24 or 90 VDC coil pulls it clear only while the axis is powered and enabled. Cut power for any reason (e-stop, contactor open, cable pulled) and the brake clamps down before the arm can drift.

It's fitted to axes that fight gravity: J2 and J3 on articulated arms, sometimes J5 on wrist-heavy models, and both linear rails on gantries. Naturally balanced axes like J1 and J6 usually don't carry one. On a healthy machine you'll hear a distinct click from J2 and J3 at power-up as the coils energize and the springs let go.

Trouble shows up in three patterns. The arm settles a few millimeters after e-stop (worn friction material or weak springs), the axis won't move at power-up and throws a position-deviation alarm (brake is dragging or slow to release), or the brake-unit fuse pops every time the coil pulls in and the controller logs SRVO-215.

Selection is by motor part number, brake voltage, and torque rating. FANUC brakes are integrated with specific motor frames, so a J2 brake from an M-710 will not fit the J2 motor of an R-2000 even when the physical mounting looks close. Order the assembly by motor part number and confirm the DC coil voltage against the drive card in the cabinet.

Compatibility limits: the brake circuit is designed for the release current the brake release unit delivers on that specific amplifier. Retrofitting a higher-torque brake onto an axis whose release unit was sized for the original coil will either fail to release cleanly or cook the driver.

Troubleshooting is mostly measurement. Meter the coil resistance cold (typically tens of ohms; verify against the motor sheet), check for continuity of the release wiring back to the amp, and watch the axis at power-up.

A healthy brake makes a crisp click at enable; a dragging brake makes a slow scrape. If the pendant reads holding torque loss during a stop test, the friction material is at end of life.

Practical example: a J3 axis on a palletizer that started settling 3 mm every time the cycle stopped had worn its brake pads flat after eight years of eight-second stop cycles. A rebuilt motor with a fresh brake stack cured it. The motor power cable is separate from the brake wiring, but a tech chasing an intermittent brake-release fault should ring both out at the same time since they share a connector housing on many robots.