Ethernet-APL: Time for action!

Planning and installation of Ethernet APL segment is no more complicated than a traditional fieldbus installation, for example. So that planners, integrators and installation technicians are supported from the outset, a comprehensive engineering guide was created at the same time as the technical specifications, which covers planning- and cabling-related aspects, as well as explosion protection.

To build an Ethernet APL network and include the field devices, what are known as field switches are needed. A field switch is quite simply a switch, as is required for any Ethernet installation. In addition to distributing and coupling data streams, it also takes on additional tasks for Ethernet-APL. It supplies the connected field devices with intrinsically safe auxiliary power, which occurs via a spur line, or "spur". These spurs are limited to a length of 200 m, which limits the installation location of the field devices to the field switch. Depending on the supplier, these field switches can be installed in the control room, Zone 2 or Zone 1. The last conformance tests for field switches are currently being carried out. The devices are expected for the market around the end of the year. Corresponding devices are already available on the market or are about to be launched. For example, the R. STAHL Field Switch recently passed the conformance tests and is already available for test installations.

Essentially, Ethernet-APL offers two options for networks. On the one hand, Ethernet-APL can be installed in the star topology usually used for Ethernet. This means the field switches are directly connected to a 4-wire Ethernet such as 100BASE-TX over a maximum 100 m distance. If fibre optics are used, what is optionally supported by most field switches, longer distances (2 to 20 kilometres) can also be bridged. This installation has three main advantages: Firstly, planning is very straightforward and only the cable length has to be taken into account. Secondly, up to 250 field devices can be connected per network, as the field switches are supplied with energy separately. And finally, the field switches can also be operated in a ring, such as the PROFINET MRP, so that the availability of the system can be further increased.

Alternatively, the trunk/spur topology can be used, which are well-known from the fieldbus world. The network transition from 100BASE-TX or 100BASE–FX network occurs via a power switch here, which likewise must be supplied with auxiliary power. The power switch converts a 4-wire network to a 2-wire network. Furthermore, it supplies the Ethernet-APL network via the trunk line. The field switches are connected to the power switch. The field switches, in turn, supply the field devices with intrinsically safe energy. The field switches can be installed up to Zone 1; the power switch is generally located in the control room or in Zone 2. The trunk line can be up to 1000 m in length and is therefore well suited to the extensive systems used in the process industry. Due to the voltage drops on the trunk, around 50 field devices per segment can be connected. However, the trunk cannot be extended as a ring, so that only a linear structure is possible here. Generally, trunk/spur technology is always a compromise between distance and number of devices. Trunk/spur technology therefore requires good network engineering, as well as a sophisticated shielding/earthing concept, above all for extended networks with copper cables.

The cabling in an Ethernet-APL network requires particular attention. The type A cables (IEC 61158-2) used in fieldbus installations until now can continue to be used; replacing them in existing systems is often linked with high expenditure. However, the quality of these cables should be critically scrutinized in older installations instead of the usual 31.25 kbit/s fieldbus, it is now 10 Mbit/s. Generelly, when using a fieldbus type A cable, the user is always on the safe side, since Ethernet-APL is adjusted to this as regards the requirements of shielding, conductor length and cross section. You can rely on tried-and-tested method for the connection types, too. Screw or spring clamp terminals, M12 or M8 plugs are still used. Generally, the connection technology is defined by the device.

Port classification has been introduced to simplify installation and limit the number of variants. There are currently three power classes: Type A has an output of 0.54 W and is designed for Zone 1 devices. A type B is already being planned and will apply to high-power Zone 1 devices with an output of 1.17 W. The R. STAHL Field Switch, for example, already offers the option of operating 12 type A and 4 type B devices. Type C is suitable for Zone 2 devices with an output of 1.11 W.

The following scenarios show how the changing over or setting up an Ethernet-APL network may look. Here, the calculations assume a maximum current consumption of an APL field device of 50 mA. The real values for the future field devices are probably significantly lower (around 30 to 35 mA). It follows that in practice there will be somewhat greater distances and the availability will be verified again.

Migrating a PA system:

• Scenario 1: The scenario is based on a fieldbus installation in which individual field device couplers (FF H1 or PA) are used and a maximum of 12 devices are connected per segment. The spur lines to the fieldbus couplers are then up to 90 meters. This results in a trunk length of up to 800 m when replaced by an Ethernet-APL field switch with e.g. 16 ports and the trunk track topology. With the star topology and the use of fiber optic cables, a significantly greater distance of several kilometers is possible, but requires a local auxiliary power supply for the field switch.
• Scenario 2: In a plant section, there are maximum twelve fieldbus devices, which are connected to three available field device couplers. The configuration is to be transferred to Ethernet-APL. To do so, three Ethernet-APL field switches are connected to the Ethernet-APL power switch. The length of the spurs are accepted at 80 m each. The distance between the Ethernet-APL field switches is 100 m. Four Ethernet-APL field devices are connected to each field switch. For the configuration, the distance between the Ethernet-APL power switch and the first Ethernet-APL field switch can be 600 metres when using a trunk cable with a cross section of 1.5 mm² (16 AWG) or 1000 metres when using a trunk cable with a cross section of 2.5 mm² (14 AWG).

New Ethernet-APL network:

• Scenario 3: The user wants to accommodate as many Ethernet APL devices as possible in one segment. When using the star topology, the field devices must be distributed to the corresponding number of field switches and these in turn must be integrated into the Ethernet backbone. The limits of this installation are primarily determined by the limits of the controllers used, i.e. the amount and speed of the data to be processed. For trunk lanes, the number and installation location of the Ethernet APL field switches as well as the quantity and power consumption of the Ethernet APL field devices must be taken into account and the maximum permissible trunk length derived from this. If 16 spur lines of the Ethernet APL field switches are used, a total of approx. 50...60 Ethernet APL field devices can be installed. The track length is then 200 m and the trunk lengths e.g. 300 m each.
• Scenario 4: The user would like to determine the number of field devices at a maximum distance with one segment. The scenario assumes a length of 1000 metres between the individual field switches, the length between the power switch and field switch is likewise 1000 metres. With an APL trunk cable with a cross section of 1.5 mm² (16 AWG), up to two Ethernet-APL field devices can be connected to each of the two Ethernet-APL switches. With a cross section of 2.5 mm² (14 AWG), the number of devices increases to three per Ethernet-APL field switch. With the star topology, you would have to use fiber optic cables and provide local auxiliary power. However, the number of field devices would then be virtually unlimited, as described in scenario 3.