Ethernet 10BASE-T1S and 10BASE-T1L are Reinventing Vehicle and Industrial Connectivity
by Dustin Guttadauro, product line manager, L-com
Since it was first conceived in the 1970s, the number of IEEE 802.3 Ethernet standards has grown to nearly three dozen with more likely to come. So, does the world need another variant when Ethernet already appears to serve every conceivable need? The answer is yes; the reason isn’t speed, but to plug a “hole” in the standards that are becoming increasingly important, especially for the auto industry and industrial applications. The solution comes in the form of Ethernet 10BASE-T1L and 10BASE-T1S that fall within the domain of Single-Pair Ethernet (SPE). To understand why the IEEE created these Ethernet variants, we need to begin with how they were created and what forces were driving them.
First, it’s important to recognize that industry and automakers have long been relying on a broad range of wired connectivity solutions, each one contributing to more complexity, size, cost, and weight. As time went by, more were added, and today the result is that there are simply too many of them, especially in vehicles where size and weight reduction are increasingly important. It would obviously be ideal if a single connectivity solution could be used for everything, and this not being an ideal world, that doesn’t seem likely.
However, Ethernet is ubiquitous and serves an extensive range of applications. Even with this many variants, each one is compatible with all the others, and this alone makes it extremely appealing. For example, a facility might need only a low-data-rate solution to serve some devices, but it’s quite likely it would also need the ability to offload aggregated data at extremely high speeds to a cloud data center or some other central source. The only solution to serve both is Ethernet.
While it’s logical to assume that the common denominator for all new versions of Ethernet will be the data rates they can achieve, for many industrial applications blazing speeds aren’t needed or even desired. This is because the connected devices require only modest data rates as they serve everything from control cabinet wiring to temperature sensors, HVAC actuators, elevators, fans, switches, voltage monitors, DC-to-DC converters, and other modules.
These systems have been getting by using field bus systems such as Profibus, Modbus, CAN Open, Device Net, CC-Link, and IO-Link that have maximum data rates of a few megabits per second but more typically kilobits per second. An increase to 10 Mb/s would be more than enough to serve these devices. With its inherent backward compatibility, every kind of device could be supported, from the simplest low-data-rate switch to high-data-rate sensors such as cameras that produce vast amounts of data and require gigabit-per-second speeds. Ethernet is also invariably present in almost every facility, so moving outward from the edge to the cloud could be accomplished with a single standard.
Filling the Hole
In 2019 the IEEE released the P802.3cg-2019 standard, the initiative for which emerged from an IEEE task force studying how to create a low-cost, lightweight, point-to-point Ethernet variant that could cover distances up to 1 km, deliver a maximum data rate of 10 Mb/s, and do so with a single balanced twisted pair of wires (Figure 1) or even on PC board traces.
Automakers in the task force requested a shorter-reach solution up to about 25 m. It also needed multi-drop capability in which multiple devices can be connected throughout the cable path. When the task force finished its work, the result was 10BASE-T1S for short-reach (25-m) applications and 10BASE-T1L for longer-reach (1-km) scenarios in which only origination and destination points are supported (point-to-point).
Automakers benefit because the vehicle connectivity scenario is being redesigned to become a largely Ethernet-based zonal electronic/electrical (E/E) architecture. This allows sensor data to be aggregated into a single link from the zonal gateway to a backbone. Ethernet’s plug-and-play capabilities also allow devices to be connected and disconnected in real-time with no downtime, which is a significant benefit versus other technologies such as the CAN bus. In addition, long distances generally are not required because all the connectable nodes are within 25 m of each other.
Settling on a single protocol for most functions has monumental benefits for automakers currently faced with supporting not just the CAN bus but multiple application-specific standards (Figure 2). Every model year brings enhancements to ADAS systems, often requiring new cameras, radar (including the latest 4D advancement), ultrasound sensors, and possibly lidar), as well as changes to infotainment and navigation systems.
The result is that in today’s vehicles, it’s common to have 40 different wiring harnesses, 80 to 100 Electronic Control Units (ECUs), and 300 wires that collectively span 2.5 mi. that weigh up to 250 lb. The multiple types of cables required for various applications also present electromagnetic compatibility (EMC) challenges as each one has its own requirements. That said, the CAN bus has served the auto industry well for many years and continues to do so for those applications for which it is best suited, so it will likely be retained for years.
Two Standards, Different Applications
There are considerable differences between the two new Ethernet variants (Table 1). For example, only 10BASE-T1L allows full-duplex operation in a point-to-point topology, while the multidrop topology of 10BASE-T1S allows nodes to be connected along the way. Data generated by the nodes on 10BASE-T1S can be aggregated in an Ethernet switch and sent to the cloud using higher data rate versions of Ethernet for processing and analysis. No gateways are required because conversion is unnecessary thanks to Ethernet’s universal compatibility.
Of the two, only 10BASE-T1S employs Physical Layer Collision Avoidance (PLCA), a key ingredient when employed in real-time applications that require deterministic performance such as automotive, industrial, and building automation. PLCA is designed for half-duplex, multi-drop networks such as 10BASE-T1S and eliminates the problems with Carrier Sense Multiple Access with Collision Detection (CSMA/CD) that have kept Ethernet from being used in many industrial Ethernet applications because it can exhibit random latencies caused by data collisions. PLCA provides guaranteed maximum latency and other characteristics that overcome these limitations. With PLCA in place, the transmission cycle begins with a beacon sent by a master node (Node 0) that the network nodes use to synchronize.
After the beacon is sent, the transmit opportunity passes to Node 1. If it has no data to send, it yields its opportunity to Node 2, and so on, with the process continuing until each node has been offered at least one transmit opportunity. A new cycle is then initiated by the master node, which sends another beacon. To prevent a node from blocking the bus, a jabber function interrupts a node’s transmission if it exceeds its allotted time, allowing the next node to transmit. The result is that there is no impact on data throughput and no data collisions on the bus.
The Benefit for Industry
For industry, 10BASE-T1L overcomes challenges that have limited Ethernet for process automation, including minimal DC power, narrow bandwidths, and range that between devices in process automation applications has been limited to 100 m. Seamless, edge-to-cloud connectivity will be achieved in process automation with 10BASE-T1L from the field devices to the control level and eventually to the cloud. Unlocking field devices will result in rich datasets for advanced data analytics (Figure 3).
Single twisted pair cabling has the advantage of being less costly, smaller, and easier to install while also carrying power as well as data. With significantly more power available at the edge of the network, new field devices with enhanced features and functions can be enabled because power limitations of a typical field bus no longer apply. For example, higher performance measurement and enhanced edge processing of data is now possible with the additional power.
10BASE-T1L removes the need for complex, power-hungry gateways, and enables a converged Ethernet network across both IT and OT networks. The result is simplified installation, easy device replacement, and faster network commissioning and configuration. Using only Ethernet across an enterprise eliminates data islands and the network is no longer fragmented.
Ethernet standards ensure that all higher protocol layers with 10BASE-T1L work exactly as with 10BASE-T, 100BASE-TX, and 1000BASE-T. Unlike legacy standards, Ethernet allows sensors to be configured from a laptop or smartphone from any location. For example, a temperature transmitter has a USB interface that is used to configure the required converter. As some devices require up to 100 adjustments, this becomes a time-consuming task prone to error. In addition, the four-wire 20 mA connection with HART (for example) requires an external power supply but 10BASE-T1L’s higher DC power level makes this unnecessary. The adoption of 10BASE-T1L is rapidly expanding in building automation, factory automation, the energy industry, automation of wastewater treatment, and a variety of other applications.
Moving In on Wireless?
Though by no means a key application for 10BASE-T1S, using it to replace short-range wireless solutions for IoT is certainly a possibility, at least for sensors that do not generate much data. A data rate of 10 Mb/s may seem laughable, but as noted earlier, it’s enough to satisfy the needs of many types of sensors. And owing to the substantial number of nodes that can be served because of its multi-drop capability, 10BASE-T1S may ultimately wind up as a viable alternative.
There are several reasons why this might occur. First, it can deliver DC power, so it could eliminate the need for additional wiring and could even power multiple sensors. Power over Ethernet (PoE) has already become prominent in other Ethernet applications, and there is no reason it cannot be exploited with 10BASE-T1S. In addition, interference is increasingly becoming an issue for wireless IoT deployments, especially when many wireless-enabled IoT devices are present. It is likely this problem will increase in severity as more IoT networks are deployed. Other problems include varying signal strength, noise, multiple incompatible standards, and security.
The PHYs Arrive
Implementing both 10BASE-T1S and 10BASE-T1L requires Ethernet transceivers that support it. The first of these, which serves 10BASE-T1S, was Microchip’s LAN867x family of Ethernet transceivers that support both multi-drop and point-to-point network topologies. Point-to-point link segments of up to at least 15 m are supported and multi-drop mode supports up to at least eight transceivers connected to a common mixing segment of up to 25 m. Since the launch of the LAN867X devices, other manufacturers such as Analog Devices have launched their own products as well.
In a nutshell, there is one overarching goal for the 10BASE-T1S and 10BASE-T1L standards: To provide a seamless, low-cost, 10-Mb/s solution for generating and aggregating data from the edge to where the data is sent to the cloud.The release of IEEE’s 802.3cg SPE standard will have an extraordinary impact on Industry 4.0 and the auto industry as it races toward simplifying its massive connectivity problem. In addition, 10BASE-T1S and 10BASE-T1L pave the way for extending the reach of Ethernet to the edge of the network, providing support for low-data-rate devices that have no need for gigabit speeds. There is also the possibility that 10BASE-T1S can find its way into industrial IoT scenarios, either to complement or replace wireless solutions. Only time will tell.
Note 1. Figures 2 and 3 are taken from the excellent Analog Devices technical article entitled “Enabling Seamless Ethernet to the Field with 10BASE-T1L Connectivity”, which can be viewed here. https://www.analog.com/en/technical-articles/enabling-seamless-ethernet-to-field-with-10base-t1l-connectivity.html