ADSL RFI - Ferrite is not the only answer!

Discussion in 'EMC Matters' started by G3YRZ, Jan 9, 2016.

  1. G3YRZ

    G3YRZ Member

    Ferrite as part of RFI handling seems to be presented as something of a panacea these days. Wonderful as common mode chokes can be, my little story shows that it is not the only thing to try.

    I have ADSL broadband, best speed about 2.5Mbps, with a wireless network. Line loss shows up as about 55db despite being in a built up area in a popular large town. I operate using only indoor aerials, a full size 80m dipole and a 20m folded dipole in the attic. Using these I cover all HF bands from 80m to 6m excluding 40m, on the odd multiple halfwave basis. I run 100W on all these bands. Results are quite good for such a compromise system, although that isn't the point of this note. There is a lot of RF around and I use a tuned counterpoise under carpets. My router is a Netgear DGN2200.

    My partner is very supportive where radio is concerned, but she does a lot online and the re-radiation and so on can knock out the router although only on 80m and 30m. No other bands are a problem.

    I made a 3-pole choke for the line in using a cheap RJ11 modem extension cable from e-bay and three recommended toroids from the RSGB shop. There are 14 turns on each toroid. The power lead to the router is threaded a few times through two cylindrical ferrites as a binocular core.

    So the leads into and out of the router are pretty much encased in ferrite! This reduced the problem and short CW QSO's became pretty much OK on both those bands. Longer CW sessions and certainly SSB on 80 carried a high but intermittent risk of killing the internet.

    Wanting to operate without restrictions, I was open to other ideas. Discussing this with Eddie M5PYE while mobile on 2m, he told me of a published article for a successful low pass filter by Steve G3VMW and kindly e-mailed me a scan. I have tracked down a .pdf version online and it is available here:

    I built this filter in a small Eddystone box on matrix board, using home wound T50-2 toroids. I have a Peak LCR analyser so was able to adjust the inductors to the design value and select appropriate capacitors.

    On testing, with the ferrite chokes still in place, the RFI was cured. Sadly, the ADSL speed reduced by over 1 Mbps which on an already slow line of 2.5 Mbps at best impacted on the service too much. A faster service could accommodate the loss but not at a speed which I compare to a snail on the back of a sloth!

    So I looked further, and found that my loss of 1 Mbps ADSL speed was mirrored in an article by Gavin, GM0GAV who designed an improved version of the filter with better matching to the ADSL line. His article on his website is available here:

    This time I used Murata inductors which I purchased from Rapid Electronics ( who are happy with small orders although you have to open an account which is no problem). I used their part number 88-1600 which are Murata PS 22R103C 10uH inductors. Again mounted on matrix board in the same small Eddystone box with the same cable.

    On testing with the ferrite chokes still in place, again there was no RFI on 80/30m. This time there was no loss of ADSL signal at all. Although I cannot say the speed improved the line loss now shows at 52db rather than 55db and the overall impression over a couple of weeks is that stability of the ADSL signal is improved a little.

    So thanks to Eddie M5PYE, Steve G3VMW and Gavin GM0GAV I can now operate on 80 and 30m with no restrictions at 100W, SSB or CW. These post is public thanks to them all!

    I do have to confess that the focus these days on ferrite did rather blinker me to other solutions, including internet searches, so Eddie did me a real favour! There was a time when ferrite was not available in the way it is today and the only vaguely similar thing was around the neck of CRT televisions. What was used then - filters, high pass, low pass, bandpass and notch.

    My current RFI handling of ferrite chokes and low pass filter is working and I plan no further experiments. I offer this note for comment or discussion if wanted, and in the hope that someone else may find it of use.


    John G3YRZ
  2. G3NYK

    G3NYK New Member

    Hi John well done, I cannot say I am surprised you had problems with 100W to an indoor dipole! Can I ask if your phone line enters your house underground, or on a "drop wire" from a DP (sorry distribution pole)?

  3. G3YRZ

    G3YRZ Member

    Hello Alan.

    I have a second floor flat in a purpose built block, with a decent size attic. The phone line is underground, enters the block into a utility cupboard where there is a distribution board. From there it rises in the cavity wall (I think - I don't believe there is a conduit) and then to my (and other) old style master socket. So that constitutes a 20 foot vertical! Because of the layout here the router is in the centre of the flat, so there is an horizontal ADSL cable from the microfilter plugged in to the master socket. I invested in an high quality screened ADSL cable but it is about 20 feet long. So from ground level to router is a 40 foot inverted "L"! Interesting potential for resonances there...

    Using the router at the master socket would mean long LAN runs and/or a degraded WiFi signal. As flats go this is a pretty generous size.

    Nobody else around has complained of RFI, touch wood, and I have been radiating RF from this QTH for many years now. The only problem has been my own router.

Share This Page