Separate fieldbuses avoid EMI and other problems

Thomas Kramer, Technical spokesman at the SafetyBUS p Club, discusses the arguments in favour of separate fieldbuses for safety and standard data.

During the first design discussions about a safety fieldbus the arguments for and against a common bus for safety and standard (non-safety) data began.

Advantages of a combined network for safety and standard control are tempting.

Major difficulties can be found with respect to security, availability and performance.

In addition some common standards impact the design and usability of the standard components in a combined network.

In the following, the various issues are discussed.

As every given field bus, and even Ethernet, have limitations in bandwidth and latency times, there is a natural correlation between performance and availability.

Based on experience, the number of safety inputs and outputs (I/O) is about 10 to 20 per cent of the total I/O of a system.

Furthermore, it is a technological fact that the safety communication requires a communication overhead of factor 2x to 3x per I/O compared to a standard I/O.

The safety I/O would then account for 20 to 60 per cent of the total busload on a combined bus.

This is a percentage that is certainly significant and has a serious impact on the design structure.

Only small systems with a good performance do not suffer from this additional load.

On bigger applications this directly requires smaller production cells compared to an actual solution if a single network is used.

Using separate busses for standard and safety does not interfere with grown production units.

Knowing that there is a natural demand for modifications or expansions of a system during its life cycle, additional I/O on the standard component side, or even just higher signal frequencies caused by faster production, impact the typical latency times of communication telegrams.

On most fieldbus technologies this may activate the safety mechanisms due to telegrams arriving late and reduce like this the availability of the total system.

Customer expectations of a combined solution of safety and standard typically imply a common programming and diagnostics.

While a common diagnostics is possible even without a combined solution, the combined programming shows itself to be more critical.

EN and IEC standards require special training for personnel doing the safety programming or design.

Like this a combined programming platform may become an availability issue.

Only specially trained personal must be able to access even the standard programming part or interference of any kind must be mutually excluded.

The same is true for repair and maintenance as only especially trained personnel is aware of the difference between safety and non-safety.

In a combined system non-trained personnel may therefore cause unintended interference with the safety functions.

This leads to a reduced availability and what is of more importance, it can lead to safety problems of the plant.

Safety systems very often require to be approved by notified bodies.

Even if this is not the case the system must be checked after modifications.

Systems with separate safety and standard structures are in natural advantage, as possible interference is implicitly excluded.

This makes approvals and tests easier and faster and in case of modifications or expansions they will very often not be required at all.

This lowers set-up and modification times and the associated costs.

The EN 61496, a standard mainly known with respect to ESPE such as safety light curtains, states that the electronic computing and communication of the safety signals needs to meet the same EMI requirements as the protection device.

The general argument is that in a safety network this applies even to standard I/O devices connected to the same network as the safety protection device.

As the EMI requirements stated by the EN 61496 are significantly higher than those of standard I/O devices, the use of safety light curtains on the same network as standard devices expands the applicability of these EMI requirements to the standard devices.

The result is that the standard devices are required to meet a higher EMI immunity.

This mutually excludes the use of old standard devices in combined networks and makes new standard devices for combined networks more expensive.

Last but not least there is the benefit to choose a suitable standard network if a separate solution for standard and safety is selected.

The user has not to focus on one individual standard network.

He can always choose the network solution for his actual process requirements and separately standardise the safety network.

Especially manufacturers building subsystems for various big end users often face the problem that the standard fieldbus is preset.

With a separate approach, at least the safety does not require modifications depending on the end user, which reduces design time and cost.

In total, the advantages of one platform for standard and safety are, in most cases, far less significant to users than the impact on availability, cost, openness, reliability and administration.

Consequently SafetyBUS p has been designed to be a dedicated safety network with gateways to various standard networks.

This ensures best performance combined with excellent transparency on a factory level, and at an attractive cost.

Changing standards, growing safety awareness and standard devices with higher EMI immunity might change this.

For the time being, using a standard field bus and safety field bus like SafetyBUS p is definitely most attractive.

Back to top