"UTP Cabling and the Effects of EMI"
Robert Y. Faber Jr., RCDD and Valerie A. Rybinski
The Siemon Company, Watertown, Connecticut
A great deal of conjecture surrounds the issue of unshielded twisted-pair (UTP) cabling and the effects of electromagnetic interference (EMI). Just recently, national and European standards groups have begun to investigate the ability of UTP cabling to support high speed data applications in environments with excessive ambient noise, but many of these studies rely on abstract and hard to quantify field strength and radiated emissions measurements. These procedures are difficult to understand and, often, do not directly correlate to the effects of real-life sources of EMI such as high frequency transceivers, electric motors, or even common fluorescent light fixtures.
To demonstrate the sensitivity of UTP cabling to common sources of EMI, several channels were set-up to run a shared IEEE 802.3u 100BASE-TX1 Ethernet operating system and exposed to various sources of noise. Signal packets and network traffic were monitored for errors using a commercially available local area network (LAN) analyzer software program. Sensitivity is determined by evaluating the number of packet errors detected under peak operating network utilization.
ELECTROMAGNETIC COMPATIBILITY (EMC)
As communications speeds increase, concern grows over noise and its potentially disruptive effects upon balanced twisted-pair telecommunications cabling. Electromagnetic compatibility (EMC) describes a cabling system's ability to minimize radiated energy levels (emissions) and resist noise interference from outside sources (immunity).
Successful strategies for minimizing radiated emissions levels and improving noise immunity for UTP cabling are dependent upon the principles of balanced signal transmission along a pair of twisted wires. As a twisted wire pair approaches perfect "balance" (i.e. the two conductors appear to become geometric duplicates of each other), currents induced on the cable as a result of noise interference equalize and are subtracted out when detected by the receiver. Good pair balance also minimizes the tendency for a cable to radiate unwanted emissions. For example, currents induced on unbalanced cabling behave as miniature loop antennas, radiating a field whose magnitude is dependent upon the degree of mismatch between the conductors of a pair. Perfectly balanced cable will exhibit infinite noise immunity and radiate zero emissions.
Although the parameter of balance is not addressed in ANSI/TIA/EIA-568-A2 and is noted as an item for future study in ISO/IEC 118013, it is commonly understood that the higher the performance category, the better balanced the cable. For example, category 5 cables would be expected to exhibit better balance characteristics than category 3 cable. It is interesting to note that screened twisted-pair (ScTP) cables rely on the same mechanism of balance in order to realize good EMC performance.
Historically, definitions of environments with "abnormally high ambient electromagnetic interference" have been vague. The field strength guideline most commonly accepted as the threshold for high EMI environments is 3 Volts/meter4 (V/m). However, this de facto requirement was selected because interference levels greater than 3 V/m typically exceed the noise immunity levels of digital devices and are above the sensitivities of analog devices, not because of a direct relationship to the capabilities of UTP cabling. Many cabling installation practice guidelines address this issue by mandating separation requirements from suspected sources of interference. For example, the National Electrical Code specifies a 2 inch separation between communications cabling and conductors of any electric light or power circuits5. Other "rules of thumb" are simply passed along from installer to installer. Unfortunately, many of these guidelines are unnecessary and generated as a result of fear, uncertainty and doubt.
An Intel "EtherExpress" 10/100 Fast Ethernet stackable hub was selected as the platform for all of the EMI susceptibility testing. Two personal computers sharing the Ethernet hub ran video and file transfer protocols concurrently to simulate network utilization rates of approximately 30%. This rate was selected to be compatible with Ethernet guidelines that recommend a maximum utilization rate of 30% to ensure acceptable access times and collision rates.
To maintain compliance with both international and national cabling requirements, all cabling channels networked with the Ethernet hub were assembled in accordance with the worst case model defined in TSB-676 and the maximum length limits established in ‘11801 as shown in Figure 1.
Figure 1: Worst Case Channel Model
Experiments were conducted for ‘generic' category 5 and ‘enhanced' category 5, UTP cabling. Cables were loosely coiled and secured to a standard freestanding equipment rack. This worst case configuration ensures that the entire horizontal cabling length can be subjected to the test sources of EMI at the same time.
Ethernet is a "broadcast" network. This means that each device connected to the network listens for traffic on the network and then sends its "packets" when the line is quiet. Packets contain sequences of binary information and packet size is usually determined by the application that is running and the type of information that is being transmitted. Packet sizes can range from 64 to 1518 bytes. Typical packet sizes are 512 and 1024 bytes. In addition to the data being transmitted, each packet also contains source, destination, and parity (bit error detection) information. Errors occur when packets do not reach their destination or information is dropped from the data sequence. Common types of errors that may be associated with full network utilization and/or noise disturbances are:
Alignment: Packets do not end on an 8-bit boundary.
Collision: Two devices detect that the network is idle and try to send packets at exactly the same time. Collision errors are common in Ethernet systems and are expected as network utilization increases. Upon receipt of this error type, both devices hold, wait a "randomly" calculated amount of time (to avoid a second collision), and attempt to re-transmit.
Cyclic Redundancy Check (CRC): Packet size is correct, but the information contained in the frame check sequence (FCS) is corrupt.
Fragment: Packet is undersized and contains corrupt FCS.
Jabber: Packet is oversized and contains corrupt FCS.
Oversize: Packets are greater than 1518 bytes in length.
Runt/Pygmy: Packets are less than 64 bytes in length.
Depending upon the severity of the error, the network may ignore packets, re-transmit packets or, the network may halt or ‘crash' because the error causes all devices to appear busy.
Packet flow on the Fast Ethernet hub was monitored via NetXRay Protocol Analyzer and Network Monitor software7. The software not only monitors network utilization levels and packet errors, but is also capable of generating errors on the network.
The operating systems under evaluation were charged with the task of simultaneous video and file transfer to keep utilization rates at approximately 30%. The NetXRay software confirmed utilization rates. Under normal operating conditions (no known sources of EMI), no packet errors were detected for any cable type.
To demonstrate the effects of packet errors, the NetXRay software was programmed to generate CRC, fragment and alignment errors. Depending upon the size of the defective packet, resultant errors ranged from a slight pause or glitch in the video display to a complete network ‘crash'. Typically, CRC and alignment errors, associated with the video transmission protocol, resulted in the most serious network crashes. All network errors were visually detectable on the video display.
The following four items were selected as potential sources of undesirable electromagnetic interference:
- Electric Hand Drill
- Transceiver Set (154 MHz)
- Fluorescent Light Fixture
- Microwave oven
A Holaday Industries field strength meter verified the level of radiated emissions from each of these devices as shown in Table 1.
Measured Field Strength
Electric Hand Drill
Note that all four of the EMI sources show field strength levels at or above 3 V/m.
The two test channels were subjected to the EMI interference sources by placing the coiled horizontal cable at, near, or around the interference source and moving the interference source to various locations at the cross-connect and work area. Resultant effects upon the network were monitored via the NetXRay software.
No packet errors were detected for either the ‘generic' or ‘enhanced' category 5 channel configuration regardless of EMI source type, source location, or duration of exposure.
The TIA and ISO/IEC technical committees have acknowledged the robust nature of UTP cabling and are pursuing the study of power and electrical separation requirements and immunity characterization. The ISO group, in particular, has made great progress in the characterization of EMC test criteria for information technology equipment and generic cabling based on radiated emissions tests. The group is currently working to develop a standardized cabling set-up so that results between various test facilities can be compared and reproduced. Even though radiated emissions requirements for electronic equipment are well established8, immunity requirements for cabling are still subject to further study. It is anticipated that the ISO working group will specify a procedure that subjects the cabling under test to a radio frequency (RF) field (most likely 3 V/m) and monitors resultant noise voltages induced on the cabling. Coupling attenuation, which represents the ratio of conducted power to radiated power, may also be used to characterize immunity. Published studies9 that utilize this method confirm the ability of UTP and ScTP cable to support high speed data applications such as 100BASE-TX in environments with high electromagnetic interference.
Good research demands enormous effort, lots of time, and commitment to the project. The authors would like to thank Mark Catrone, RCDD and Matt Mailhot of The Siemon Company and Rudy Jordan of Cinco Networks for their efforts leading to the publication of this document.
- ANSI/IEEE Std. 802.3u-1995, "Media Access Control (MAC) Parameters, Physical Layer, Medium Attachment Units, and Repeater for 100 Mb/s Operation, Type 100BASE-T", 1995.
- ANSI/TIA/EIA-568-A, Telecommunications Industry Association Standard, "Commercial Building Telecommunications Cabling Standard", 1995.
- ISO/IEC 11801, "Information Technology - Generic Cabling for Customer Premises", 1995.
- Section 126.96.36.199 of ANSI/EIA/TIA-569, "Commercial Building Standard for Telecommunications Pathways and Spaces", 1990
- Section 800-52.2 of the NFPA-70 National Electrical Code, 1996.
- TIA/EIA TSB67, "Transmission Performance Specifications for Field Testing of Unshielded Twisted-Pair Cabling Systems", 1995.
- NetXRay Protocol Analyzer and Network Monitor Software, version 2.5, Cinco Networks, 1997.
- CENELEC EN55022/A1, "Limits and methods of measurement of radio disturbance characteristics of information technology equipment", 1995.
- Danish Electronics, Light, & Acoustics (DELTA), "Proposed cabling set-up for electromagnetic characterization of cabling and EMC measurements on LAN systems", Erik Bech, April 1997.
Robert Y. Faber Jr., RCDD received the B.S. degree in Economics from Southern Connecticut State University in 1982 and is the Manager of Educational Services at The Siemon Company. He has achieved the Registered Communications Distribution Designer designation from BICSI. Mr. Faber is an active member of the BICSI Standards Committee and has spoken at numerous international telecommunications conferences.
Valerie A. Rybinski received the B.S.E.E. degree from the University of Connecticut in 1991 and is a senior electrical engineer at The Siemon Company. Her research activity is mainly in the field of balanced twisted-pair telecommunications cabling and connecting hardware. Ms. Rybinski actively participates in national working groups responsible for the publication of industry standards such as ANSI/TIA/EIA-568-A and is the chairman of the TIA PN-2948 connecting hardware task group.