1. Possible Impact of VDSL on Stations Operating in the Amateur Radio Service Ed Hare, W1RFI ARRL Laboratory Manager 225 Main St Newington,CT 06111 w1rfi@arrl.org 860-594-0318 w1rfi@arrl.org

2. Part 15 Absolute Maximum Limits Unintentional Emitters Non-interference Manufacturer responsible for FCC authorization and maximum limits Operator responsible for harmful interference Both are important to mitigate possible harmful interference

3. Unintentional Emitter Emissions Radiated emissions above 30 MHz Conducted emissions below 30 MHz Good engineering practice Carrier-current devices must meet intentional emissions limits – 3 installations Meeting the intentional emissions limits on HF is good engineering practice

4. Intentional Radiator Radiated Emissions Limits - HF Sec 15.209 1.705-30.0 MHz – 30 uV/m at 30 meters This should protect users of the spectrum against interference, yes?

5. No! If the absolute emissions limits were set to offer unconditional protection to all radio services, the permitted levels would be unworkably low Amateur Radio Service, by design, uses very sensitive equipment and weak signals The “legal limit” will result in a strong signal to nearby Amateur HF installations On 3.5 MHz, a half-wave dipole placed in a 30 uV/m field will receive a –86.4 dBW signal (338 microvolts across 50 ohms) To hams, this is S9+16 dB – clearly harmful interference to typical amateur communications! Harmful interference at even greater distances than the compliance distance is likely The absolute limits are not enough to prevent interference to nearby receivers

6. Harmful Interference The repeated disruption of radio communications Merely hearing a signal is NOT harmful interference 30 uV/m at 30 m works somewhat well for discrete signals If from broadband device, however, will interfere with entire band(s)! 30 uV/m works somewhat well for isolated sources If from VDSL, interference potential exists for occur for entire length of wiring in areas where VDSL is deployed!

7. A Case History – Wireless Modem Jacks Carrier-current devices that use residential electrical wiring to couple modem signals between a computer and a remote telephone connection Phonex model PX-421 designed to operate on 3.53 MHz These were purchased in volume by TCI Cable and installed in conjuction with their digital cable systems Widespread S9++ levels signals and harmful interference Phonex responded promptly, redesigned product TCI, now AT&T still in midst of system-wide recall! Costs?

8. What Can Be Expected from VDSL? Better scenario than power lines and electrical wiring Phone system reasonably well balanced at audio At HF? 20 dB to 30 dB balance typical?

9. Amateur HF stations Bands at 1.8, 3.5, 7.0, 10.1, 14.0, 18.1, 21.0, 24.8 and 28.0 MHz Receiver sensitivity –135 dBm Ambient noise levels –125 dBm Antenna gain 2.14 dBi (F.S) on 3.5 Mhz Antenna gain 7.5 dBi (F.S) on 14-30 MHz

10. Balance EZNEC 3.1 used to model 50 feet of simple phone wiring (uses NEC-4, written by Lawrence Livermore National Laboratories) Terminated in 100 ohms – j0 Fed with two sources, one on each wire, 180 degrees out of phase, with 20, 25 and 30 dB of balance Showed “gain” of –16 dBi to –26.4 dBi Real-world installations – bigger radiators, but more loads on line Correctable defects

11. Balance 20 dB = -16.8 dBi 25 dB = -21.8 dBi 30 dB = -26.4 dBi Will use –21.8 dBi for following calculations

12. Frequency = 3.5 MHz Receiver bandwidth = 2500 Hz Transmit power in 1 Hz = -50 dBm/Hz Transmit power in 2500 Hz = -16.0 dBm Telephone-wiring antenna gain = -21.8 dBi Receive antenna gain = 2.14 dBi Distance to receiver = 0.03 km Free-space pathloss = -12.9 db E-field estimate in 30 kHz = 140.8 uV/m peak Receive Noise Figure = 24 dB (includes external noise) Calculated receive system sensitivity = -116.0 dBm Received noise in receiver bandwidth = -48.6 dBm peak Received noise in receiver bandwidth = S9 + 24.4 dB Receive system noise floor increase in dB = 67.4 dB

13. Frequency = 14 MHz Receiver bandwidth = 2500 Hz Transmit power in 1 Hz = -50 dBm/Hz Transmit power in 2500 Hz = -16.0 dBm Telephone-wiring antenna gain = -21.8 dBi Receive antenna gain = 7.5 dBi Distance to receiver = 0.03 km Free-space pathloss = -24.9 dB E-field estimate in 30 kHz = 140.8 uV/m peak Receive Noise Figure = 24 dB (includes external noise) Calculated receive system sensitivity = -116.0 dBm Received noise in receiver bandwidth = -55.2 dBm peak Received noise in receiver bandwidth = S9 + 17.8 dB Receive system noise floor increase in dB = 60.8 dB

14. Phoenix VDSL testing

15. Conclusions Pre-standard system, but approximately –50 dBm/Hz Underground wiring, relatively new homes No VDSL signals detected except near one neighborhood hub – just audible Other noises In areas of underground wiring, notching not needed However, this should NOT be extrapolated to determine that notching is not needed under other circumstances!

16. HPNA

18. HomePlug – Final Specification

19. Industry Measurements of Interference Potential? In general, industry has not made measurements of interference to other services Such measurements, made by Amateur Radio in other, similar disciplines show interference potential In all cases where industry groups have conducted joint studies with ARRL, they have chosen to include spectral masks in their industry standards or specifications To date, no widespread interference from protected products

20. What is Needed? Based on modeling and calculations, -50 dBm/Hz will probably result in harmful interference with overhead wiring and older homes Amateur Radio vs other services? These models not exact, but demonstrate clearly the need for further study ARRL wants to participate in field studies, especially any near New England, but I will travel!

21. What Does ARRL Want to Accomplish? Participate in field studies Protection to –~80 dBm/Hz in ham bands Such notches will exist naturally in residential telephone wiring, so product must be robust enough to function with small segments of frequency band attenuated Advisory language in standard about harmful interference, the need to use notches when needed and the need for manufacturers to include such advisory information in their product literature Not necessary to use filtering – HPNA standard achieves this level of protection by not using OFDM carriers in the ham bands. This will take care of MOST cases of harmful interference. Remainder are “correctable defects” or can be addressed on a case-by-case basis If frequencies not used, additional filtering can be added Why waste a good notch? Other services?

22. More Information Ed Hare, W1RFI ARRL Laboratory Manager 225 Main St Newington,CT 06111 w1rfi@arrl.org 860-594-0318 http://www.arrl.org/~ehare/rfi/vdsl/vdsl.html http://www.arrl.org/tis/info/part15.html