Elevator controller enclosure design: failures and the fix

Elevator controller enclosure design usually goes wrong in the machine room, not in the cabinet. The drive derates, the controller faults, the car parks at the landing, and the fault clears by itself — which is the hardest kind of fault to defend to a building owner.

The governing numbers are simple. A 7.5 kW lift with a regenerative drive and a controller dissipates about 355 W. A 600 × 400 × 1200 mm cabinet in coated steel has 2.88 m² of surface and sheds that at roughly 5.5 W/m²K, which puts the inside 22 K above the machine room. At 40 °C ambient the electronics sit at 62 °C, and most of them are rated to 55 °C.

This guide covers elevator controller enclosure design as four decisions: the thermal path, the IP rating set by location, the earthing and bonding across coated joints, and the door and gland details that keep water and dust out. Each failure mode comes with the number that causes it and the callout that prevents it, so the drawing that goes to fabrication already accounts for the machine room it will live in.

Elevator controller enclosure design: how hot does the cabinet run?

The cabinet runs 20–25 K above ambient on natural convection, which is fine in a ventilated machine room at 30 °C and a problem at 45 °C.

The heat load is the sum of three parts. The drive loses roughly 3% of its rating, so a 7.5 kW drive contributes 225 W. The controller electronics add 50–80 W, and a small control transformer another 30–50 W. Call the total 355 W for a 7.5 kW installation, and scale it for larger drives.

The cabinet sheds heat through its surface. For a painted steel enclosure with no forced air, the coefficient is about 5.5 W/m²K, including radiation and convection, and only the external faces count. A 600 × 400 × 1200 mm box gives 2 × (0.6 × 0.4 + 0.6 × 1.2 + 0.4 × 1.2) = 2.88 m².

The rise is then ΔT = P / (k × A) = 355 / (5.5 × 2.88) = 22.4 K. If the machine room reaches 45 °C on a bad afternoon, the cabinet interior is 67 °C.

Two levers change that. A larger cabinet adds surface: 800 × 500 × 1400 mm gives 4.44 m² and a 14.5 K rise, so the same 355 W runs at 59 °C in a 45 °C room. A filter fan does better. The airflow needed for a 10 K rise is Q = P / (ρ × c_p × ΔT) = 355 / (1.2 × 1,005 × 10) = 0.029 m³/s, which is 106 m³/h.

That is the trap. A filter fan rated at 180 m³/h free flow delivers about half that once the filter is fitted, so 90 m³/h — less than the 106 m³/h the cabinet needs. Specify a 300 m³/h fan, not a 180 m³/h fan, and check the derated figure on the supplier curve rather than the catalogue headline.

The same cabinet does not scale with the drive. A 15 kW lift contributes about 450 W from the drive and 130 W from the electronics, so 580 W through the same 2.88 m² surface runs at 580 / 15.84 = 36.6 K above ambient — 77 °C in a 40 °C room. No fan rescues that inside an IP54 box. The cabinet has to grow, or the heat has to leave through a heat exchanger with its own rated capacity.

Cabinet size, IP and the limits for a lift control panel

The middle column is the specification for a machine-room controller; the right column is what the fabrication route holds.

The IP row is the one that gets argued. IP2X is a finger-safe requirement: no opening large enough to admit a 12 mm sphere, achieved with a formed lip on the door and no louvres. IP54 adds dust protection and splash resistance, which means a gasket, a sealed gland plate and no unfiltered opening. The two are different products, not two settings on the same drawing, and the location should decide which one before the sheet is cut.

Five failure modes that show up as lift faults

Thermal derating inside the cabinet

The drive derates on its heat-sink temperature before it trips. The car gets slower on the top floors in the afternoon, nobody reports it, and then a busy period pushes the drive into a fault. The mechanism is the 22 K rise plus ambient, not a faulty drive.

Measure the machine room, not the design ambient. If the room reaches 42 °C in summer, the cabinet needs forced ventilation or a larger surface. Adding a louvre to an IP2X cabinet is acceptable; adding one to an IP54 cabinet is not.

Water ingress through a gland plate

A gland plate without a gasket is a 200 mm wide gap at the bottom of the cabinet. Water runs down the trunking, sits on the plate and finds the clearance holes. The failure is a wet terminal block and a nuisance earth fault that clears when the cabinet dries.

The plate needs a closed-cell gasket, captive screws at 150 mm pitch and a drip edge on the inside so water cannot run along the plate into the cabinet. On an IP54 cabinet, the cable entries are the weakest point in the whole assembly.

Earth bond lost at a powder-coated joint

The door is bonded through its hinges and the hinges are powder coated. At 60–80 µm the film is an insulator, so the door reads open circuit to the body. The fault appears as noise on the controller or as a failed insulation test at commissioning.

Bond with a masked M6 stud on the door and one on the body, joined by a braided strap. Masking has to be on the drawing, because the booth will coat everything that is not taped.

Door distortion and gasket set

A 500 × 1,100 mm door in 2.0 mm steel bows under the gasket load unless it has a return on all four edges. Bow at the latch means the gasket compresses at one corner and not the other, and the IP54 seal becomes a dust seal within a season.

Vibration from the traction machine

A machine-room controller sits on the same slab as the traction machine and the rope sheave. Vibration at the machine's running frequency loosens self-tapping screws, wears clearance holes and, over years, cracks a solder joint on a terminal. Weld or rivet the internal mounting rails, and use thread-forming screws with a locking feature rather than a plain self-tapper.

Condensation in an unheated machine room

A sealed cabinet in a room that cools overnight draws its own weather. At 35 °C and 80% relative humidity the dew point is 31 °C, so a cabinet that has dropped to 28 °C by morning is below the dew point and its internal surfaces are wet. The water collects on the cold back plate, runs onto the drive terminals and shows up as an earth fault at first start.

A 50 W thermostatically controlled anti-condensation heater keeps the cabinet above the dew point for a few watts, and it is cheaper than a sealed breather or a heater inside the drive. Give it its own circuit and a thermostat that switches on below the calculated dew point — 31 °C in this case — rather than a fixed setting copied from another site.

What Ganesh actually runs

Controller cabinets are box-bent on the 200 T, 5-axis brake with the 4,450 mm bed, which forms a 600 × 400 × 1,200 mm body from one blank and keeps the corners square without a welded seam. Doors, gland plates and mounting rails are laser cut on the 12 kW fibre laser with the 40 ft × 16 ft bed, and the rails are welded or riveted on fixtures built in our tool room. Coating runs through seven-tank pretreatment and the 18 ft × 14 ft powder booth at 60–80 µm, with DFT and cross-hatch checks per batch and the earth studs, gland lands and hinge faces masked to drawing.

Enclosures for lift programmes have gone through this route for TKE and Eros Elevators, and the elevator parts page lists what a controller drawing needs to include. The manufacturing route guide covers the blank size and the forming sequence for the same cabinet.

We build to the enclosure drawing. We do not assemble the drive, wire the cabinet or certify the IP rating; those belong to the panel builder, and the IP test is done on the assembled enclosure, not on the sheet metal.

What to put on the drawing

1. Drive rating, controller model and the total internal dissipation in watts.
2. Machine room ambient maximum, and whether the position is shaft, car top or outdoor.
3. Required IP rating with the test method, and which faces are sealing faces.
4. Cabinet size, panel thickness and door size, with the thermal calculation attached.
5. Ventilation type — louvre, filter fan with derated airflow, heat exchanger or air conditioner.
6. Earth stud positions, strap route, and the masked-face schedule.
7. Gland plate size, gasket type and cable entry direction.
8. Finish, DFT range and touch-up procedure.
9. Tolerance split: ±0.5 mm on hole positions and ISO 2768-1 class m for formed features.

When a fabricated cabinet is the wrong route

A cabinet that has to carry a type-tested assembly to IEC 61439 is a panel builder's product, and the enclosure is only one input to that test. A controller for a machine-room-less lift that has to sit in a 600 mm shaft often needs a purpose-designed compact unit with a sealed heat exchanger, which is a different enclosure family. And an outdoor or car-park position may need stainless steel or a double-skin sun shield rather than a thicker mild steel panel.

For machine-room and shaft positions where the cabinet is a formed steel box on a wall, fabrication is the right route and the drawing decides the outcome. The elevator component hub collects the related guides, the box bending guide covers the corner details, and the welding inspection guide covers the records a lift OEM audit will ask for. Send the cabinet drawing and the drive data through request a quote and we will check the thermal path and the bonding schedule before pricing.

The two lines that prevent most callouts

Most controller complaints end in one of two drawing notes. The first is the dissipation figure in watts with the ambient it was calculated against; without it, the fabricator cannot size the surface and the panel builder cannot size the fan. The second is the masked-bonding schedule; without it, the door is coated and the earth path disappears.

Neither note costs anything at the drawing stage. Both cost a site visit once the lift is in service and the building manager has stopped taking calls.