MARTIN - G8JNJ

ECLECTIC AETHER - Adventures with Amateur Radio

VDSL interference

About 50% of my amateur radio friends are currently suffering severe interference from Broadband internet VDSL services that are carried by conventional copper phone lines.

More technical information on VDSL can be found at kitz.co.uk

The RSGB also have a useful guide to the subject

This page summarises my investigation into various aspects of interference arising from VDSL services in the UK.

Block diagram of a VDSL 'Drive-By' interference monitoring system built by me.
The system consists of:-

A roof mounted active antenna based on the Chris Trask circuit featured on my Active antennas webpage

The antenna has been calibrated so that it can be used to measure field strength

The active antenna feeds a distribution amplifier and splitter.

The split signals feed:-

!. An up-convertor, to convert the signals to a suitable frequency for the RTL USB SDR dongle which is used by RTLSDR Scanner running on the Laptop
2. A Signal Hound USB spectrum analyser
3. A 3-4MHz Band Pass Filter ahead of a broadband RF signal monitor which gives an audio indication of signals levels whilst driving (so that I can watch the road)

The GPS antenna provides positional information to the Laptop so that RTLSDR Scanner can plot the route

The dash mounted WebCam allows me to record the position of the overhead phone lines and VDSL street cabinets as I drive past


Screen grabs of the system in use



First VDSL Downlink (Cabinet to Modem) spectrum
VDSL Uplink (Modem to Cabinet) spectrum



Videos of the system in use


Some videos taken whilst the system was in use


Identifying interference on drop wires


Once the general location of the VDSL nterference source has been determined, individual phone lines can be checked with a small tuned loop and RF detector.

The detector is based on an AD8307 logarithmic RF detector

http://www.analog.com/en/products/rf-microwave/rf-power-detectors/non-rms-responding-detector/ad8307.html

The loop is a single turn series tuned circuit with a impedance transformer to match to 50 Ohms. The tuned frequency can be changed to 3.1 or 4.1MHz by means of a small toggle switch.

An additional 3MHz to 4.5MHz band pass filter is placed ahead of the RF detector to help reject other signals that may be present.



The loop can be mounted on a telescopic fishing pole so that it can be used to test individual drop wires.


Ideally the loop should also have had an electrostatic screen to reduce E-Field pickup. I may build a new active detector probe with the ability to switch between E and H field modes, so that I can investigate which is the dominant transmission mode.


Observations


General observations (so far) please note these are generalisations and I have added further comments (from others) in red:-

1. Buried phone lines generally have much lower levels of VDSL radiation than those fed via overhead line
However buried lines can radiate a strong magnetic field component which may not be so obvious when using an E field antenna

2. VDSL street cabinets and nearby cables and distribution poles generally have very low levels of VDSL radiation
However some multicore cables emanating from cabinets can radiate

3. The multicore twisted pair cables that are used to connect between distribution poles, generally do not tend to radiate
Common mode currents can be measured on the multicore cables running up the poles and also in underground ducting using a current transformer

4. The overhead lines and drop wires from the distribution poles to the premises can radiate interference (depending on the line balance)
The length of the drop wire can also modify the radiation pattern

5. Not all overhead lines and drop wires from the distribution poles to the premises radiate (depending on the line balance)
All lines will radiate to a certain extent, if there is a standing wave on the drop wire it will depend upon if you are close to a current maxima or minima

6. Premises near the VDSL cabinet tend to radiate more in the Downstream spectrum
But this could also be partially due to the greater number of cables radiating at that point

7. Premises much further away from the VDSL cabinet tend to radiate more in the Upstream spectrum
This is because Upstream power backoff reduces power when close to a cabinet. Greater power is required when further away from the cabinet

8. Levels of up to 77dBuV/m (>40dB above system noise floor) have regularly been observed (antenna approximately 3m below overhead lines)
Some signals at even higher levels have been observed by others