Six more.
The part was never the problem.

Soft foot. One motor foot sat proud of the baseplate. Torquing it down sprang the frame, so the motor was bolted on under strain from day one. Run it, heat it, shake it, and a sprung frame relaxes and creeps. Thermal growth would have shown hot and come back cold. This was out cold.

The tell: alignment that won't repeat. Put a dial on the shaft and slacken one foot bolt at a time. Any foot that moves it more than 0.05mm is soft.

Fix: check for soft foot before you align, every time. Shim the soft foot flat, recheck all four, then align and torque in sequence.

A flat accumulator. The bladder had lost its nitrogen precharge. With no gas cushion the accumulator stores almost no oil, so every clamp stroke pulls pressure straight down and the pump has to cut in to cover it. A leak would bleed pressure at rest. This held. The switch was telling the truth, the system had lost its buffer.

The tell: a pump that cycles with demand but holds pressure at rest, and a gauge needle that swings hard on every stroke. Oil at the gas valve means the bladder has split.

Fix: isolate, drain the oil side and check the precharge with a charging kit. Recharge with dry nitrogen only, never air or oxygen, to the figure on the plate. If it won't hold, replace the bladder, then add a precharge check to the PM.

Running off the curve. Throttling the discharge pushed the pump far back on its curve, well left of its best efficiency point. Down there the radial load on the impeller climbs, the shaft deflects, and the seal faces stop running square to each other. Seals go first, bearings follow. The pump was sized for the old duty and the valve made it fight itself.

The tell: a discharge valve barely cracked open, a gauge sitting close to shut off head, and a wear track on the seal faces heavier on one side.

Fix: stop controlling flow with the valve. Trim the impeller or put the motor on a VFD so the pump runs near its best efficiency point at the new duty. Then fit the seal.

A dragging brake. The brake lining had worn until the air gap was too wide for the coil to pull the armature plate fully clear. The brake only half released, so the motor spent every lift driving against its own brake. The heat at the fan end was the brake, not the windings.

The tell: heat and brake dust at the fan end, a discoloured brake disc, and an air gap past the maximum on the plate when you put feelers in it.

Fix: with the lift secured and isolated, reset the air gap to spec, or replace the disc if it's below minimum thickness. Check the brake rectifier while you're in there and put the air gap on the PM.

The wrong cable. When the panel moved, the run was extended with ordinary copper multicore instead of thermocouple extension cable. That made the junction box by the barrel the real cold junction, but the controller was compensating for its own terminals in a cool panel. It read low by roughly the difference between the two, so it drove the zone hotter to make up for it.

The tell: a steady offset that moves with the temperature around the junction box, not with the setpoint. Check the display against a calibrated probe, then follow the cable. Proper extension cable is colour coded to the thermocouple type. Grey multicore is the giveaway.

Fix: run matched extension or compensating cable all the way back to the controller, with the right thermocouple connectors, then verify the zone against a calibrated probe.

A slipping coupling. The coupling was held by a single grub screw on a plain shaft. Every hard reversal let the hub creep a fraction. The motor and encoder did exactly what they were told, the load just didn't follow. The drive can't see slip that happens on the far side of the encoder.

The tell: drift that builds one way and clears on homing, with no following error. Paint a line across the shaft and hub and watch it part.

Fix: fit a clamp hub or keyed coupling, torque it to spec, and leave a witness mark so any future slip shows at a glance.