VDSL Emission Investigation

Discussion in 'Investigations New Technology' started by Robin, Nov 20, 2014.

  1. Robin

    Robin New Member

    The EMC Committee is carrying out an investigation of the significance of interference to amateur radio from leakage from VDSL systems. The emissions from VDSL are continuous and indistinguishable from white noise and may go unnoticed.
    VDSL2 covers frequencies up 17.66MHz. The downstream and upstream bands alternate and the easiest way to detect VDSL is to look for the change in noise level at the band transitions. These are nominally 0.138MHz, 3.75MHz, 5.2 MHz, 8.5 MHz, 12MHz, and 17.66MHz. The band that predominates depends on distance of the house from the street cabinet. Most problems are seen when the connection to houses in the area is by overhead cables.
    The effect will be that received signals seem weak and the background S meter reading higher than expected. To check for VDSL emissions tune about 75kHz either side of the nominal transition frequencies. Ignore off air signals and local incidental noise. Tuning around these frequencies and listening while watching the S meter will show a noise step if there is VDSL interference. If an SDR is available the step will be clearly seen and the 4.3125kHz patterning of the VDSL signal may be visible on a waterfall display, but careful adjustment of the contrast may be required for this.
    If you want to see some examples look at the 2014 RSGB Convention paper http://rsgb.org/main/files/2013/10/2014-Convention-EMC-paper-emerging-threats-v5.pdf There is also an item on VDSL emissions in in EMC Matters in RadCom December 2013.
    The key facts we would like you to report are the frequencies and size of the steps. If you have an SDR, spectral and waterfall plots would be very useful. If you are not QTHR please give some indication of your location.
    Please reply via this forum, to emc.chairman@rsgb.org.uk or to myself directly.
    Thanks and 73
    Robin G3JWI
    Acting Secretary EMCC
    G3VPW likes this.
  2. Robin

    Robin New Member

    Re the VDSL Investigation.
    Hello everyone
    There has been a gratifyingly large number of “views” but no replies on the forum. Apologies to everyone who has tried unsuccessfully to sign in to the forum. We have had a number of email. replies to emc.chairman for which many thanks. Please don’t wait for the forum to be fixed to report VDSL experiences but email to emc.chairman or to my personal email address which is (suitably disguised to avoid web crawlers): rmpj1937ATbtinternetDOTcom
    Thanks and 73
    Robin G3JWI
    Acting Secretary EMCC
  3. G3VGZ

    G3VGZ New Member

    I am suffering from this even though I'm in a (virginmedia) cableled area and all BT lines are underground. I get a 10dB increase in noise on 10MMz which is reduced over a small range when I transmit, to be replaced by a carrier about 20dB above floor. If I QSY and transmit again another carrier will appear about 4KHz away from the first. I have a feeling I also get it on 18MHz. It is not significantly higher than a lot of other local QRM I get, particularly something broad recurring every 31KHz and pulsing at 1HZ some of the time. I've no idea what that is.

    Brian G3VGZ
  4. Robin

    Robin New Member

    Thanks Brian.
    To confirm that the noise on 10MHz is VDSL leakage and to get some idea of the strength could you listen on the upstream 2 (US2) band edges. This may be tricky because of the off air (wanted) signals - particularly at the 12MHz end which is in a broadcast band.
    At my qth the actual frequencies of the band edge steps are 8.545MHz and 11.960MHz. The US1 band edge frequencies are 3.90MHz and 5.160MHz so it would help if you could listen there also. I just used available 7MHz and 10MHz dipoles as antennas.
    I will attempt to upload a plot of the 12MHz step at my qth. I,m not really familiar with this forum so don't be surprised if I'm not successful!.

    Attached Files:

  5. M0JAV

    M0JAV Moderator

    We have compiled the initial list of reports received. The levels shown are the rise in noise floor reported.
    We have also had some success through collaboration with BT in getting the interference levels reduced by
    BT improving the line to the subscriber. We need your reports to collect enough data to prove the extent of the problem.
    Please encourage amateurs with problems to post here or to email emc.chairman@rsgb.org.uk
    The attached file contains the reports so far
    John Rogers M0JAV
    Chairman RSGB EMCC

    Attached Files:

  6. mm0zbh

    mm0zbh New Member

    Similar problems here, VDSL arrived in my village about 6 weeks ago. I was connected early on and noticed a minor (1s point) increase on 10Mhz to about s2-3. I have the telegraph pole in my back garden but using the rotary dipole can work round the noise. But in the last week 10Mhz has been unusable due to a broadband noise of about S7-S8. Tried powering down the house and running off a battery to no affect. Following some tests when I transmit a couple of dits the noise drop back to normal S2-3 then comes back after a short period of about 10 seconds. A longer transmit causes the noise to drop and be replaced with a S9 tone every 4kc on 10.108, 10.112 etc which continue for about a minute before stopping and the broadband noise reappearing again about S7-8. I am guessing I am reseting someones VDSL modem although my concious is clear as my modem is not affected as proved by continuous ping tests from the laptop when transmitting. Additionally I successfully use my MFJ1025 to remove the broadband noise using a longwire as a sense antenna wrapped around the phone line where it comes from the pole to the house with the null affect being the same on both my mine and my neighbours external lines. Guess someone has some very poor quality kit emitting into the local phone system and is a getting an equally poor FTTC service.
  7. Ken G3SDW

    Ken G3SDW Moderator

    Paul, make sure you do contact John M0JAVas requested on previous post.
  8. John G3XRJ

    John G3XRJ New Member

    Currently no VDSL within 250mtrs so figures may be interesting for comparison.
    Rural location with K9AY receiving system 150mtrs from nearest property.
    Preamp at loops to raise signal to equivalent to that from a directly connected dipole.
    Midday noise floor figures :-
    160m -121db
    80m -119
    40m -116
    30m -116
    No 4khz spaced tones whether or not I transmit on 30m.

    telephone lines rise from the ground within 35mtrs of K9AY system so I await the arrival of VDSL to the hamlet with trepidation !
    John G3XRJ
  9. Terry George

    Terry George New Member

    Significant problems here; RSGB and Ofcom aware.
    Bit of video at

    I am advised that the interference that seems to be generated by the modem is in fact the modem activating the cabinet or exchange equipment and noise then radiated back onto the ?ADSL2 telephone line.

    Terry G4AMT
  10. Ken G3SDW

    Ken G3SDW Moderator

    Terry G4AMT and John G3XRJ, please get in contact with John Rogers M0JAV on emc.chairman@rsgb.org.uk you need to talk to him ASAP
  11. Robin

    Robin New Member

    Hello John
    Thanks for the figures. We are short of on figures for the actual noise measured in a moderately quiet locations so these are most welcome.
    I assume they are dBm at the receiver input..
    Robin G3JWI
  12. John G3XRJ

    John G3XRJ New Member

    Yes Robin. db Measurement made on screen of Elecraft P3 calibrated for 1uv = -107db
  13. Paul gi4fzd

    Paul gi4fzd New Member

    Can anyone confirm what may be causing the 10db or so increase in noise floor just above 3.7mhz and right down to bottom end of 80 meters in the attached frame grab, Thanks
  14. Ken G3SDW

    Ken G3SDW Moderator

    I think with your very limited info Paul it will be quite impossible to give you an answer.
    Please can you give us some more back ground on your 80m noise problem.
  15. Paul gi4fzd

    Paul gi4fzd New Member

    Hi Ken,
    This is a permanent feature at my location for over a year or so, I have tried a vertical antenna and an inverted v dipole ,both have the same increase of noise floor at that frequency,
    there is no such problem with antenna removed also I am getting the same rise in noise floor on my kx3, so it is being picked up on the antenna and not mains borne. I can't be certain when exactly it started
    but certainly wasn't like this a few years ago.
    I don't see any other issues on other bands. There are a number of mobile masts within 100 yards and a few solar panel installations within 600 yards or so.
    I thought maybe someone would have seen something like this before and immediately recognised it. Any advice would be helpful.
  16. Ken G3SDW

    Ken G3SDW Moderator


    First thing to do is to turn off your mains supply at the main fuse box and see if the noise is still there using a portable receiver, this will prove that it is not coming from your house, this is because if you do have to call Ofcom then they will not give you a hefty bill, some three figures.

    With the receiver assuming it is not there when you turn off the mains supply then take a walk around and try to locate were it is coming from.

    At the stage let me know and we will take it one stage further.

  17. M0JAV

    M0JAV Moderator

    Many thanks to those of you who responded to our request for interference reports on VDSL. As we expected it is only people with overhead cables
    carrying the connection to their or nearby houses that suffer badly.
    We have two clusters of reports those near cabinets suffering 20m interference and those further away suffering at 40 and 80m. Unfortunately both
    groups suffer interference at 30m with many seeing the retraining carriers at 10.104 and 4kHz steps from those when they transmit.
    We still need more reports only 21 so far so please encourage others to look and report. Here is the latest summary with postcodes given but not callsigns.

    Attached Files:

    Kev likes this.
  18. Paul M0XDX

    Paul M0XDX New Member

    My noise level increased by 18dB on 7Mhz and 15dB on 14Mhz suddenly a few weeks ago. The noise is on 24 hours a day 7 days a week and is white noise so just sounds the same as strong atmospheric noise.

    I am in LN12 postcode. The local cabinet is 100m away and was upgraded to VDSL in February this year. I have not upgraded to fibre broadband myself.

    I have tried DF’ing the noise and it appears to be coming from the overhead lines and especially the telephone pole outside my neighbour’s house.

    I have checked the transition frequencies on my SDR:
    6dB at 5.245MHz
    6dB at 8.5MHz
    4dBat 12Mhz
    The screenshots are attached.
    12Mhz transition.jpg 5.245MHz transition.jpg 8.5MHz transition.jpg
    Paul M0XDX, Mablethorpe.
  19. topbandman

    topbandman New Member

    Sent this to the EMC Chairman a while ago, but the mail bounced for some reason:

    My station has have been suffering from an elevated noise floor over part of the 80 metre band for the last few weeks. The noise is 24/7 and shows no sign of going away of its own accord. I suspect it may be related to a recent installation of VDSL / VDSL2. You can see the effect of the noise 'live' on my WebSDR at www.160m.net. You will observe that the background noise floor is some 10 - 13 dB higher in level below about 3707 kHz when compared with above. My normal dry, daytime noise floor on 160 and 80 here runs at around -104 dBm in an SSB bandwidth, as indicated on the WebSDR (uncalibrated).

    Part of the idea of running a WebSDR from here is to give others the benefit of listening to bands from a location where background noise is substantially less than in urban locations. The service I provide, not to mention my own enjoyment, is now being compromised severely.
    Our village (Colton, near Rugeley, Staffordshire) has been enabled for high speed broadband (BT) for a few months now, and we were one of the first homes to take a VDSL2 service. Our house is about 150 metres from the BT green cabinet and distribution from there to here is mainly underground, but with the last 20 metres being overhead, fanning-out from a pole which feeds ours and several other, adjacent properties. Our line speeds were measured by BT on installation as close to 80 Mbit/s down and 20 Mbit/s up. Our line was noted to be very well balanced. We generally get a good service from our ISP (PlusNet). There was no change in my receiving noise floor on any band that I could detect relating to our own installation.

    The centre of my dipoles (in the back garden of our house) is some 160 metres from where the overhead telephone lines fan-out from the distribution pole to the front of our house and those of our neighbours. Far closer (30 metres) to my dipoles is a telephone line that runs overhead to a water pumping station.

    Evidence is only circumstantial at this stage that the problem is VDSL-related. There was a BT / Kelly Communications white van and a linesman up the pole that feeds our house a number of weeks back and I have a suspicion that that is when the noise started. If so, then the radiation from the line(s) must be severe to be detectable at such a distance. There is a possibility that the telemetry equipment in the water pumping station that is far closer to my antennas has been upgraded to use high-speed broadband. I have tried transmitting in the band occupied by the noise, using the full 400 Watt, and was unable to discern any change in the noise spectrum and amplitude.

    I have added another temporary RX to my WebSDR (RX4), having a centre frequency of 8.5 MHz, chosen deliberately to correspond to one of the VDSL band edges. Around this frequency there is a lot of broadband noise. Multiple carriers are visible, spaced at around 8 kHz and, listening on AM, a tone may be heard mixed in with the noise. I wonder if others having more experience can confirm or therwise the likely source of this noise? RX3 on my WebSDR covers down to 5.2 MHz, which I gather is another VDSL band transition frequency. There is some increase in the amplitude of the noise floor when tuning HF of this frequency, but no real pronounced 'cliff edge' as visible on 80 metres.

    When time permits, I intend to make an RF sniffer and install on the end of a fishing pole and go for a walkabout, probing around the overhead lines and VDSL band edges. If I can identify the source with a reasonable degree of certainty, I will then refer the matter to OFCOM.


    Mark, G4FPH.
  20. Ken G3SDW

    Ken G3SDW Moderator

    Mark are you saying that you still have not spoken to John Rogers M0JAV the EMCC chairman? as until you do you will not get much joy out of Ofcom at the moment.
    I am about to send your details to him and get him to get in contact with you if that's OK with you.
    Kev likes this.

Share This Page