Earlier this year we at Cabling Installation & Maintenance had the opportunity to interview several individuals involved in the development of standards and technologies related to single pair cabling systems. Single pair cabling is intimately related to Single Pair Ethernet, a technology that has proven successful in other industries, particularly automotive production. For the past couple years, standards and product development activity have brought single pair cabling to the precipice of market introduction.
This roundtable-style conversation features the following individuals, whose companies are members of the Telecommunications Industry Association’s Single Pair Ethernet Consortium (SPEC): Adnan Ahmed, technical advisory and solutions lead, Reichle & De-Massari; Valerie Maguire, distinguished engineer, Siemon; Harshang Pandya, general manager, test and measurement solutions, AEM; and Bob Voss, distinguished engineer, Panduit.
CI&M: The ANSI/TIA-568.5 standard is a work in progress. Can you tell me what the standard will cover, and where the document is in the standard-development process?
Maguire: The TIA-568.5 standard is the sixth in the series of 568 standards. It is a new document, called Balanced Single Twisted Pair Telecommunications Cabling and Components Standard. The standard defines two categories of Single Pair Ethernet cabling. The first is SP1-1000, defining a channel with 1000-meter reach. The second is SP1-400, defining a channel with a 400-meter reach. The channel configurations allow a different number of connectors. The SP1-1000 allows 10 connectors, and the SP1-400 allows 5 connectors.
Another interesting aspect of this standard is it will allow conductors as large as 18-AWG. So conductors from 18 gauge to 23 gauge are allowed. The thinking is that with a 1000-meter channel, you’re going to need the larger-gauge conductors, with more copper to meet the insertion-loss requirements. And with the 400-meter channel, we can use a copper-conductor diameter that looks more like a Category 6A or Category 7.
Ahmed: It is also worth noting that the SP1-1000 permanent link includes 9 connectors, while the SP1-400 permanent link includes 4 connectors. The standard also specifies the measurement procedures for all transmission parameters.
CI&M: Does the standard define or require any type of topology?
Maguire: The standard does not recommend a topology; there isn’t a standard crossconnect kind of configuration. What these connectors represent is a way to manage essentially an extraordinarily long channel. In this case, it makes sense to have a consolidation point, where 100 meters from the telecommunications room, you have an accumulation of connectors housed in its own enclosure, that break out to other areas where building automation systems are deployed.
So although it is not specified in the standard, we at The Siemon Company envision the 400-meter channel as similar to an expanded star, where the first connector would be your interconnect in the telecom room. If you had two connectors it could be a crossconnect, but in most cases there would be just one. Then you’d go out 100 meters or so to a zone enclosure. From there you could break out to a second zone enclosure that might be 300 meters away, for example. Then your fourth or fifth outlet would be the equipment outlet. But it remains to be seen what kind of configuration users will choose. The advantage with structured cabling is, you can precable. That’s a different approach than traditional Fieldbus-type solutions and RS-485 and control cable solutions, where you don’t precable. It will be interesting to see where this goes.
CI&M: What can you tell me about the commercial availability of single-pair cable and single-pair connector products?
Voss: I can speak for Panduit. We have had a Single Pair Ethernet solution, both cabling and connectors, since December 2020. And there are other solutions out there. As an industry, we are at the early-adopter stage of single-pair cabling now.
Ahmed: We officially launched our single-pair cable and connector products in June. Two interface designs are available, one that is in compliance with the IEC 63171-1 standard and another that is in compliance with the IEC 63171-2 standard. The R&M single pair cabling solution consists of a complete channel, including patch cords, connector modules, and cables. It is based on the reliable IDC termination technology—the insulation displacement contact—and has compatibility to existing 19-inch panels and outlets. So if you are using a 19-inch framework for your RJ-45-based Category 6/Category 6A installation, you can use the same frame for single-pair modules.
Maguire: At Siemon, we are approaching the single pair ecosystem with a particular focus on the opportunity in building automation, Internet of Things, and application convergence. So we are looking at the SP-1, 400-meter implementation. To support convergence between 1-pair and 4-pair cabling, we launched a system in August 2020 that is based on our TERA Category 7A solution. We have quite a bit of experience with Category 7A, and one advantage of a 7A system is each pair is individually foil shielded. Because of its very robust insertion-loss performance, each pair meets all the stringent requirements of the 568.5 standard. We believe there is a real advantage to being able to deploy four single pairs over one cable. There may be cost advantages, and there certainly is a pathway-sharing advantage.
We feel, at this point, that market deployment and availability of Single Pair Ethernet equipment is quite a ways out in the future, maybe even as far as 36 to 48 months. So the system is designed to be compatible with one-pair screw terminal and one-pair pluggable interface connections. But in the absence of the availability of those connections right now, deployment of these systems would be for screw terminals only, or for futureproofing. An advantage of this type of deployment is that a Category 7A permanent link can be run to futureproof for either 4-pair or breaking out for future 1-pair. And the existing installed base of 100-meter Category 7A cables can be added to, with additional lengths of cables to support Single Pair Ethernet devices in the future.
So we are comfortable with this implementation, and we like the opportunity for 400-meter single-pair channels in the commercial building environment. And we will continue to adjust and monitor as devices become available in the marketplace.
CI&M: Valerie briefly mentioned termination styles. Adnan and Bob, can you also comment on the termination of your single-pair systems?
Ahmed: Our system is IDC field-terminated, just like RJ-45s. The connection modules are very much alike.
Voss: Panduit’s connector design does not require fancy tools. It is terminated in the field, with a tool as simple as a pair of pliers that an electrician would have. And that’s noteworthy for single pair, because today the edge of the network is supported by electricians. Once Ethernet gets more engagement at the edge that will change, but today, the network’s edge where these devices will be deployed is serviced by maintenance staff and electricians.
CI&M: Harshang, can you discuss the need for, and the practicality of, testing single-pair cabling systems in the field?
Pandya: One of the things about Single Pair Ethernet is, although it is a relatively simple technology, and as has been explained, has familiar termination practices, it does need to meet transmission performance requirements—for 10 Mbits/sec to begin with, and then for future technologies that might be used on the same cabling. So there is a significant amount of effort being done in the standards committee. PN-5071, which will be published as TIA-5071, Requirements for Field Test Instruments and Measurements for Balanced Single Twisted-Pair Cabling.
There are some unique challenges related to testing for single pair. The first is that the cabling is specified for 400 meters with 5 connectors, or 1000 meters with 10 connectors. So it’s a very long distance—10 times the distance of the standard 4-pair cabling. As a result, the testing needs to be configured differently in terms of frequency spacing, because for longer distances, in order to accurately spot where defects are, we need much finer resolution at frequency points. Currently the discussion is having a frequency resolution of 20 kilohertz, as opposed to 150 kilohertz, which is the minimum resolution in current standards.
The 1000-meter length of single pair systems brings its own challenges. As we increase frequency, insertion loss increases. For longer distances, like in this case up to 1 kilometer, it is impractical to use much higher frequencies. In order to accommodate data rates, Single Pair Ethernet technology will make use of the lower frequency spectrum, and as a result, testing also will need to be done at the lower frequency spectrum. In this case, we’re talking about a low-end frequency of about 100 kilohertz or even lower, as opposed to 1 Megahertz.
One other factor related to test equipment is that the main unit and the remote unit must communicate with each other. Synchronizing two units that are a kilometer apart, and having them talk to each other, is a technical challenge for the tester manufacturers.
At AEM, our TestPro is equipped with software, and we have hardware in the form of an adapter, to test single pair cabling.
CI&M: One technology that has been associated with Single Pair Ethernet is PoDL—Power over Data Lines. Can you explain what PoDL is? Maguire: IEEE 802.3bu introduced us to PoDL. It is different from PoE, which has power over two or four pairs. So in an of themselves, PoDL and PoE are not inherently compatible. Within 802.3bu, 10 classes of PoDL are specified, classes 0 through 9. They run over a variety of voltages—12, 24, and 48 volts. While different systems are meant for different implementations, all these 802.3bu PoDL applications have one thing in common: They are designed for cabling systems that have a DC loop resistance of 6.5 ohms or less. Those familiar with structured cabling may know that our loop resistances are much higher than that. So in essence, that means the classes 0 through 9 specified in 802.3bu are not PoDL applications that would be used to support either 400- or 1000-meter single pair.
That brings us to the IEEE 802.3cg Single Pair Ethernet Amendment. That amendment does want to supply power over these longer data lines—the 400-meter links, as well as links as short as 15 meters. 802.3cg introduces six additional classes of PoDL power, classes 10 through 15. There are essentially two ‘buckets’ of these classes; one bucket has a 30-volt maximum voltage from the PSE [power sourcing equipment], and the other bucket has a 58-volt maximum from the PSE. In sub-buckets are how much power can be delivered based on the DC loop resistance of the cabling.
For our purposes, we care most about the 30-volt/up-to-65-ohm bucket, and the 58-volt/up-to-65-ohm bucket. Basically, class 10 will be the longer-reach 30-volt implementation, providing minimum power at the device of about 1.2 Watts. Class 13 is the second implementation for longer reaches, providing a minimum power at the device of about 7.7 Watts. So that’s the type of wattage that can be supplied over single pair. It is low because we have the same maximum voltage constraints, and must deal with additional resistance that the 400-meter and the 1000-meter topology bring in.
CI&M: What is the value proposition for single pair cabling and Single Pair Ethernet in manufacturing and industrial environments, as well as in smart buildings?
Ahmed: Single pair cabling had its origin in the automotive sector. In order to be prepared for the challenges of the future, single pair was developed as a lightweight, space-saving and low-cost transmission protocol. It is now finding its way into buildings. In industrial and process automation environments, Single Pair Ethernet offers and IP-based alternative to existing fieldbus protocols. The vision is an all-IP-based communication system from the sensor to the cloud.
Industrial environments operate using machine-to-machine communication that typically occurs between low-complexity components that are commonly found in control and automation systems. The fieldbus devices in these systems include sensors that detect light, heat, motion, moisture, pressure, or any other environmental phenomena and respond with information or an instruction. So they also include actuators that can move or control mechanisms.
These systems have been moving from older, proprietary communication protocols to Ethernet networks, and the new single pair standard provides a seamless migration path to ensure interoperability between diverse systems. So if I look at it from the perspective of an industrial network designer, an all-IP infrastructure will provide more bandwidth and power than previously available through legacy fieldbus technology. It is anticipated that 10Base-T1L technology will quickly facilitate the use of Ethernet to the edge of the network, to realize the ambition of Industry 4.0 smart manufacturing.
Maguire: If you look at the types of systems that 10Base-T1L is intended to evolve, the systems are very diverse. The interfaces are different; the cable is different. Sometimes it’s a one-pair system, sometimes it’s two-pair, depending on whether or not power is involved. The lengths are all different for these different applications. In the equipment room, the systems do not all plug and play with each other. Overall there is a lack of homogenization—all different cables, all different connectors, all different equipment that doesn’t play nicely with each other. And we don’t have the ability to converge these systems. So a first step is to get these automation systems in the enterprise environment communicating with each other, so I can tell what sensor A is reading, and initiate an action on controller B—two different systems, but the same cable and the same language.
The next thing we’d like to facilitate is communication between 4-pair and 1-pair networks. Right now you can’t take a 1-pair Single Pair Ethernet cable and plug it into a 100Base-T switch. They are not interoperable. We are going to need some sort of media conversion. That’s where I envision us being. All the devices, all the systems will be based on twisted pair, and they will be based on one standard structured cabling topology that has a maximum length and a maximum number of connectors. All these systems can then be networked, and can be accessed by a single, universal controller platform.
There are huge obstacles to overcome. These installers typically are siloed. They’re familiar with their specific connectors and their specific connectors. There is fear and uncertainty that converging these systems and networking them together is going to interfere with people’s livelihoods. So we have a significant obstacle to overcome there, in the sense that we do need networking people and building automation people to converge.
Ahmed: Valerie makes a great point. It’s not just the technology, but also the human factor. A building’s operational technology [OT] traditionally has been more focused on proprietary BACnet or other fieldbus communication systems, and now they are all coming to IP. So that is a change for OT workers and facilities workers to move on, and at the same time, change their thinking process from having a proprietary protocol and connectors, to more open network protocols like Ethernet.
Traditionally there are two separate teams in a building: OT and IT [information technology]. Now you’re bringing them together onto a single team, using technology that is going to be all IP-based.
Pandya: When Single Pair Ethernet first addressed the automotive space, some of the challenges faced were the large number of sensors. The number of sensors in a vehicle is increasing all the time. And a challenge lies in the need to connect a large number of sensors to processing units, the need to power the sensors from a centralized place, and the complexity of managing an assortment of different networking technologies and making them all work together. An additional requirement was to reduce the weight and cost of the cabling. These are very much the same factors that play out inside buildings. There are so many sensors to be connected. One characteristic of these sensors is that they do not have a high data rate requirement. That is true in automotive applications and in buildings. But in buildings, the sensors are spread out significantly, so the 100-meter channel restriction we are familiar with would be an inhibiting factor. Also, to use wireless communication would require a large number of batteries, which has its own challenges. Single pair cabling solves a number of challenges. It can sufficiently handle the low data-rate requirements. It is less costly and occupies less space than four-pair cabling. And it provides wired connectivity plus power to each network device.
Patrick McLaughlin is our chief editor.