Zero Volts, Proven: A Safe-Isolation Routine for Bosch Drive DC Buses

Why a stopped motor proves very little

A drive that has stopped, a motor that is silent and a display that has gone dark are useful observations. They are not evidence that the DC bus is dead. Inside a Bosch drive, the DC link capacitors store energy at a potential that can be substantially above mains voltage, and that energy can remain present after the supply has been locked off. On a 400 V system the DC bus is nominally in the region of 565 V DC, before ripple and before any regenerative contribution is considered. Higher-voltage drives and multi-axis arrangements can be less forgiving still.

Retrofit work deserves particular caution. Existing cabinets may contain undocumented modifications, shared DC bus bars, external capacitor banks, brake choppers or regenerative units. A permanent-magnet motor can generate if the load moves, and a common DC link can allow one drive to backfeed another even when the obvious incoming supply is isolated. Safe isolation therefore has to be proved at the point of work, not inferred from the state of the control panel.

The UK safety frame

In the UK, the Electricity at Work Regulations 1989 require electrical systems to be constructed and maintained so as to prevent danger. For practical work, that means a documented safe-isolation procedure: identify all sources, isolate, lock off, prove the test instrument, prove the circuit dead, and prove the instrument again. Test probes and leads should be suitable for the environment and for the voltage present, in line with HSE guidance note GS38. A standard multimeter may be useful for recording a measured value, but proving dead should be done with an approved voltage indicator and a proving unit appropriate to DC measurement.

This is not administrative decoration. The DC bus is a stored-energy hazard, and the instrument used to declare it dead must be capable of detecting the voltage that could be present, including DC, and must be rated for the category of installation. Check the meter's CAT rating, check the leads, check the calibration status and check the proving unit before the enclosure is opened.

Where to measure, and what to expect

Use the drive manufacturer's instructions as the primary reference. The exact terminal designations vary by product family, so the following is a method rather than a universal terminal list. Measure between the DC link positive and negative terminals, between each DC rail and protective earth, and across the incoming and outgoing terminals as required by the site's isolation procedure. On a retrofit, also check any external DC bus bars, brake resistor terminals, shared-link connections and capacitor assemblies.

Test all combinations. A single measurement between two phases can miss a live DC rail referenced to earth, and a measurement at an internal test point can miss a backfeed from a parallel drive. Where a drive has a shared DC bus, isolate the group and prove the whole group dead, not just the unit being worked on. Where a permanent-magnet motor is connected, secure the load against rotation before relying on the electrical isolation.

Discharge times are not universal. Some drives use active discharge circuits whose behaviour may depend on control power being present; others rely on bleed resistors. Small drives may fall to a low voltage within a few minutes, while large capacitor banks and multi-drive line-ups can take considerably longer. Do not apply a fixed five-minute rule across every Bosch drive. Read the manual, assess the installation and record the time at which the measurement was taken. If the voltage is falling but not yet at a level agreed in the site's safe system of work, keep waiting or use the manufacturer's approved discharge method. Do not short the DC bus to earth.

A dark display is not a measurement

Indicator lamps and displays can fail, and a blank screen says nothing about the charge held by the capacitors. A DC bus LED that is off may be evidence of a healthy discharge circuit, or it may be evidence of a failed lamp. The only defensible proof is a measured value, taken with a proven instrument, at the point of work, after the specified waiting period.

Take a reading, wait, and take another. A falling trend is useful information, but it is not proof of zero. A stable low reading deserves scrutiny: check the meter range, check the probe contact, check for induced voltage and confirm that the measurement point is the correct one. Where the reading remains above the threshold in the site's safe system of work, treat the circuit as live and escalate. The manual, not habit, should govern the next step.

A pre-work checklist for retrofit teams

• Identify the drive model, the relevant manual and the DC bus topology, including any shared link, external capacitor bank, regenerative unit or brake chopper.
• Confirm the isolation points, the lock-off arrangements and the permit-to-work requirements.
• Check that the voltage indicator is suitable for DC, rated for the installation category and in calibration, and that a proving unit is available.
• Isolate every source, including control supplies where they affect active discharge, and lock off.
• Secure any permanent-magnet motor or driven load against rotation.
• Wait for the manufacturer's stated discharge time before the first measurement.
• Prove the tester, measure all required terminal combinations, and prove the tester again.
• Record the measured values, the time and the instrument used.
• If work is suspended, re-prove the circuit before anyone re-enters the enclosure.

The evidence, not the assumption

As UK plants continue to retrofit Bosch drives into existing cabinets, the safe-isolation routine has to be explicit and repeatable. Dead is not a state that can be assumed from a stopped motor, a dark display or a quiet cabinet. It is a measured condition, supported by the right instrument, the right terminals and a record that another engineer can follow.