Troubleshooting Magnus-21 Door Drive Commissioning Faults

On Magnus-21, a door that races, loses VKO, or keeps moving after a command drops can involve different layers: mechanical slip, encoder tracking, station wiring, model-specific behavior, or a failed output stage. Separate those layers in commissioning order; changing force or speed cannot repair an incorrect signal path or slipping belt.

Magnus-21 revision and parameter identity

Identify the exact drive variant before changing parameters or interpreting inspection-mode behavior. Field references distinguish Magnus 49L8 and 49L10 from newer 54L3 and 54L2 variants. The distinction matters: earlier drives may continue a movement to its endpoint after an inspection button is released, while the newer variants move only while the button remains pressed. That difference alone does not prove a fault.

Record the model marking and current parameter values before edits. A field parameter list is included below, but its numbers are not a universal factory baseline: compare them with the exact unit documentation and the installation’s accepted setup. In inspection mode, an open KBR circuit with the cabin stop button held was reported to release closed-door holding; treat that state as a mode-dependent behavior, not proof of a failed holding output.

1. Read the model marking on the drive and match it to the applicable documentation; the expected result is an exact variant identification, not just “Magnus-21.”
2. Compare inspection-mode movement with that variant’s described behavior; the expected result is movement consistent with whether the version requires a held button.

Door-leaf geometry and belt tracking

Correct the mechanics before increasing motor force. A slipping toothed belt moves the door without preserving the relationship between motor rotation and door position. The drive can then lose its calculated opening point, fail to issue VKO, or continue pushing against a stop. A belt that appears tensioned can still slip under an obstruction or incorrect linkage geometry.

With the cabin doors closed, both carriages should meet their rubber stops rather than stopping because the leaves collide with each other. At full opening, the larger leaf should meet the adjustable stop on the rail while the smaller leaf remains short of that stop. The shaft doors should open wider than the cabin doors, including the smaller leaf. Check the coupling and the smaller-leaf mechanical reverse for free movement; a binding or wedged actuating bar can delay reverse operation while the belt skips.

Inspect the complete travel for interference, including a displaced pin or other hardware contacting the linkage. Do not compensate for binding by raising force, cutting belt teeth, or disabling a leaf. Those workarounds conceal the mechanism that causes lost position or delayed reversal.

1. Move the doors through their full travel in the controlled commissioning mode; the expected result is free movement without binding, belt tooth skipping, or impact at an unintended stop.
2. At the closed and open endpoints, check the carriage positions against the stated stop geometry; the expected result is both closed carriages at their rubber stops and the larger open leaf at its adjustable stop while the smaller leaf remains clear.

Encoder feedback and endpoint recognition

Magnus uses encoder pulses to relate drive movement to door travel. Missing or intermittent pulses, a mispositioned encoder, damaged wiring, or belt slip corrupts that relationship. The result can be an overrun, a missing VKO or VKZ, or repeated pushing after the door has reached its mechanical limit. A door that reaches the physical endpoint is not proof that the drive has correctly recognized it.

Inspect encoder mounting and wiring, then observe position feedback during repeated opening and closing. Confirm that the indicated direction follows the actual door direction and that the count or position progresses consistently through the stroke. Check the end-stop positions on the door rail and any obstruction that acts on the linkage. If Magnus loses VKO after a belt skip, first restore the mechanics and encoder reference, then measure the opening again using the unit’s prescribed procedure; do not tune around a position reference that changes from cycle to cycle.

1. Run repeated controlled strokes while observing the available encoder or travel indication; the expected result is consistent position progression with no dropouts or unexplained jumps.
2. At each endpoint, compare the physical stop with the drive and station limit indications; the expected result is VKO at the fully open end and VKZ at the fully closed end.

24 V command wiring and station interface

The reported opening and closing command inputs are 331a and 332a, and the command level is 24 V. Do not connect station conductors L51 or L52 to those inputs, and do not apply 220 V. A mistaken connection can blow fuses or damage the drive. Similar-looking terminal labels also create a risk of landing L313A where a 331a or 332a command belongs.

Check the unit schematic and terminal labels rather than relying on wire color or a similar number. Verify the command voltage against the drive’s common, identified as -L in the field wiring description. When the drive operates in installation or inspection mode but remains inactive in normal operation, compare the actual open/close commands and door interlocks in both modes. That comparison separates a drive fault from a missing or misrouted station command.

For a reported case of blown opening and closing fuses, inspect for an unintended short from -L to chassis and verify the station outputs before condemning the power stage. Isolate circuits under the applicable elevator maintenance procedure; do not test by applying an improvised jumper or an unverified supply to a drive input.

1. Measure the command at 331a and 332a relative to the documented common while issuing each direction; the expected result is the specified 24 V command on the correct input, with no 220 V present.
2. Compare the station terminal-to-drive terminal mapping with the schematic; the expected result is that L51/L52 and L313A are not mislanded on the 331a/332a inputs.

VKO, VKZ, KZ, and station relay states

Separate the door-end signals from optional relay functions. VKO and VKZ report open and closed limit states. KZ is a separate relay function described for some installations as a shaft-protection bypass; it is not a general substitute for the drive’s open/close commands. If that relay is fitted, check its coil wiring and the polarity of any diode across the coil. Incorrect polarity can damage the VKZ output key.

A relay contact that remains open when the drive should assert it can result from a failed relay or output key, but first confirm the drive state and the contact circuit against the unit schematic. A simultaneous VKO/VKZ indication with Magnus unpowered can originate in the station: the described relay chain drops K1, which drops K10 and removes power from Magnus. In that condition, investigate the station relay sequence and supply interruption before treating the two missing limit signals as independent drive faults.

1. Observe VKO and VKZ at fully open and fully closed positions; the expected result is the corresponding endpoint signal asserted at each position, not both signals lost during a powered, stable endpoint state.
2. If E59 appears with both limit indications, check K1, K10, and Magnus supply state; the expected result is a confirmed cause in the station power/relay chain or a separately measured drive/output fault, rather than an assumed KZ failure.

Magnus-21 parameter map and revision limits

The values below reproduce a field parameter list. Treat the listed numbers as a reference snapshot, not as guaranteed factory settings or a recommended target for every door. Parameter distances have no physical unit specified here; do not convert them to millimeters. Preserve the current values, confirm the model-specific meaning, and change only parameters needed to correct a measured behavior.

1. Read and record parameters 00–28 available on the installed revision; the expected result is a complete baseline with no value outside its applicable range.
2. For parameters 20 and 21, confirm the model before editing; the expected result is use of those force adjustments only where the exact variant supports them.

Speed profile and deceleration correction

A door that completes its stroke in roughly half a second and does not slow down points first to the motion profile, travel reference, or encoder feedback—not automatically to a separate speed-regulation board. Parameters 06 and 07 set running speed in the listed map; 10 and 11 set opening and closing deceleration distance. Parameters 12–15 define start and final speeds. These settings act on a travel profile, so a wrong belt length or lost encoder position can make a valid deceleration setting occur at the wrong physical point.

After restoring mechanics and a repeatable travel measurement, adjust one direction and one relevant value at a time within the applicable range. Reduce speed or force if a mechanically sound door is skipping the belt under load; do not increase force to overcome a stuck linkage or slow the door by creating friction. On the cited 54L3/54L2 variants, do not expect parameters 20/21 to correct final closing force; use only the adjustment functions available on that revision.

1. Record the present travel reference, running speeds, and deceleration distances, then compare them with the exact model map; the expected result is a valid travel reference and in-range settings for both directions.
2. Cycle opening and closing after a controlled adjustment; the expected result is visible deceleration before each endpoint, without belt skipping, hard stop impact, or lost VKO/VKZ.

Output-stage faults and transistor replacement

A shorted output transistor is a board-level power-stage fault, not a parameter problem. One damaged-unit description gives the marking as 64BC30KD, while a later component search names IRG4BC30KD. Resolve that discrepancy from the actual device marking, board position, and correct schematic or parts documentation before ordering. The suggestion to use IRF540 was challenged because the cited device differs in power capability and lacks the described internal protection diode and short-circuit adaptation. Do not treat it as a drop-in replacement.

If qualified board repair is chosen, the service guidance in the field calls for replacing all output transistors together with the driver rather than mixing a new device into a set of failed or stressed parts. Confirm the approved replacement’s package, electrical ratings, pinout, protection features, and driver compatibility from authoritative parts data. If the correct parts or board documentation are unavailable, replace the drive with an approved unit rather than operating it with output devices removed; a motor that starts after transistor removal is not a serviceable repair.

Before attributing blown command fuses to KZ, isolate the command wiring and check the documented common for a short to chassis, then verify whether the fault remains on the drive side or station side. A separate KZ coil/diode wiring error can damage an output key, but does not by itself explain every command-fuse failure.

1. With the unit isolated under the applicable service procedure, confirm the failed component marking against the board documentation; the expected result is an exact replacement identity with no guessed cross-reference.
2. After repair, verify the full output stage and driver against the approved service procedure; the expected result is normal controlled motion and valid endpoint signals, with no fuse operation or output short.

End-to-end Magnus-21 commissioning proof

Commission the drive only after the mechanical path, feedback, station commands, and revision-specific settings agree. This sequence distinguishes a door that physically travels from a system that also recognizes travel and reports the correct state to the elevator controller.

1. Confirm model identity, preserve the baseline settings, and verify the values against the applicable revision; expected reading: exact model recorded and all used values within that revision’s stated ranges.
2. Inspect closed and open carriage geometry, belt engagement, linkage, and mechanical reverse; expected result: no binding or tooth skip and the leaves meet their intended stops.
3. Run a controlled travel and inspect encoder progression; expected reading: consistent direction and position progression with no missing or jumping feedback.
4. Issue open and close commands from the applicable modes and measure at 331a/332a; expected reading: the correct 24 V command reaches the correct input, without a 220 V supply or misrouted station conductor.
5. Observe acceleration, running speed, and deceleration in both directions; expected result: repeatable motion that slows before the endpoint without striking or driving against the stop.
6. Confirm the fully open and fully closed signals at the drive and station; expected result: VKO at fully open, VKZ at fully closed, and no unexplained simultaneous loss.
7. Restore normal elevator operation under the site commissioning procedure and call the car to a served floor; expected result: the station commands the door, the door closes fully, and the station receives the closed-door state required for elevator movement.

Magnus-21 troubleshooting FAQ