Access Official Technical PDF

Share Summary:LinkedInEmail

Safe torque off (STO) is a widely used safety sub-function in industrial motor drives. It prevents torque-producing power from being supplied to a motor when a safety demand occurs, such as an interlocked guard opening. Designing STO is more than switching off a drive: engineers must consider fault tolerance, diagnostics, response time and the requirements of the applicable functional-safety standards.

What safe torque off does—and does not do

IEC 61800-5-2 defines safety sub-functions for adjustable-speed electrical power drive systems, including STO. In practical terms, STO removes the drive’s ability to generate motor torque by interrupting the energy path used by the power-stage switching devices. It is intended to help prevent hazardous motion, for example when an operator accesses a protected production cell.

STO is not a controlled stopping function. When it is activated, the motor is no longer driven, but it typically coasts to a stop. The time required to reach zero speed depends on the motor and load conditions. If an application requires a controlled deceleration or a predictable stopping time, that need must be addressed in the broader machine safety design rather than assumed from STO alone.

How STO relates to functional-safety targets

Drive safety designs are assessed within standards and frameworks that include IEC 61508, IEC 61800-5-2, ISO 13849-1/2, IEC 60204-1 and IEC 62061. The applicable requirements depend on the system and its safety function. A Safety Integrity Level (SIL) expresses a target level of risk reduction; it is not established simply by selecting a particular component or circuit.

Several architectural and reliability measures affect the level a safety-related subsystem can support:

  • Probability of dangerous failure per hour (PFHD): a measure of the frequency of dangerous failures that prevent the safety function from being performed. Lower values correspond to more demanding SIL targets.
  • Safe failure fraction (SFF): the proportion of failures that are either safe or detected as dangerous. Improving diagnostic coverage can increase SFF.
  • Hardware fault tolerance (HFT): the ability of the architecture to continue performing its safety function despite faults. HFT of 1 indicates that one fault does not, by itself, cause loss of the function.

Subsystem classification also matters. Type A subsystems use components with well-understood failure modes, determinable fault behavior and dependable failure data. Type B applies when one or more of those conditions is not met, such as for more complex components. The SIL that can be supported depends on the combination of subsystem type, SFF and HFT; these factors must be evaluated together.

STO architecture: independent channels and diagnostics

One implementation approach uses two independent hardware channels to disable gate-driver power. In a 1-out-of-2 architecture, either channel can initiate the safety action, while the second channel provides the ability to maintain the function if one channel fails. Adding diagnostics produces a 1oo2d arrangement, intended to detect certain faults as well as provide redundancy. The architecture must still be evaluated as a complete safety-related system; redundancy alone does not establish a SIL claim.

In an isolated gate-driver design, separate STO inputs can control power switches associated with the gate-driver supplies. When both inputs are at their specified active logic level, the drive can operate. If either input goes low, the relevant supply path is disconnected, removing the gate-drive capability needed to produce motor torque. A controller may also use an STO signal to disable pulse-width modulation, but software action should not be confused with the independent hardware shutdown path.

Additional cross-channel logic can help strengthen the shutdown arrangement. For example, combining signals from both channels and a monitored output in a hardware logic function can provide another means of enforcing shutdown. The specific circuit and its diagnostic behavior must be analyzed for the target system and safety requirements.

Diagnostics and fault analysis

Diagnostics help identify failures that could otherwise prevent STO from working when demanded. A system-level fault analysis should consider components such as digital isolators, low-side and high-side power switches, gate-driver supplies and monitoring circuits. Pin-level failure information can contribute to a system failure mode and effects analysis (FMEA), which examines how faults affect the safety function.

Examples of faults that require consideration include an isolated STO signal stuck high, a power switch that fails to turn off, or an output short. A signal stuck low or a switch that fails to turn on may cause a safe shutdown, while a signal stuck high or a switch that remains on could be dangerous because the drive may continue operating despite an STO demand. Monitoring, diagnostic pulses and fault reactions can help detect or control specific conditions, but their effectiveness depends on the implementation and timing.

Reliability calculations also require care. Failure-in-time (FIT) expresses a failure rate of one failure per billion hours, but a functional-safety FIT assessment is not simply a component-level headline number. Failure distribution within a device matters: if a particular function has a higher likelihood of failure, the surrounding hardware and diagnostic design must account for it. Process safety time is another key constraint—the safety system must respond quickly enough after a hazardous failure to prevent the hazardous event.

Response time is part of the safety case

STO response time includes the time from an input transition to the relevant gate-driver supply or status response. In one illustrated implementation, the measured interval from STO1 going low to activation of the gate driver’s RDY signal was 2.8 ms; for STO2, the corresponding interval was 7.6 ms. These are design-specific measurements, not universal STO response times. The first interval depends on capacitance at the power-switch output, while the monitored supply reaching its undervoltage-lockout threshold triggers the RDY indication.

Engineers should distinguish electrical response time from the time the motor takes to stop. STO removes torque-producing drive power; the motor’s coast-down behavior remains dependent on the machine and load. The safety analysis must account for the relevant timing and hazards in the actual application.

Design implications for drive teams

For system integrators and drive designers, a robust STO implementation begins with the machine-level hazard and required safety performance, then traces that requirement through architecture, component selection, diagnostics and validation. In particular:

  • Define whether the application needs torque removal, controlled stopping, or both.
  • Assess channel independence, fault tolerance and the consequences of single faults.
  • Use FMEA and failure-rate data to support subsystem-level analysis.
  • Verify diagnostic coverage and fault reaction timing against process safety time.
  • Validate the complete implementation against the standards and requirements applicable to the machine.

Safe torque off is a foundational motor-drive safety function, but its effectiveness depends on the complete design—not on the STO label alone. A considered combination of independent shutdown paths, diagnostics, fault analysis and timing verification helps teams determine whether the drive’s safety function is appropriate for its intended role in the machine.

Access Official Technical PDF

About The Author