"shielded" twisted pair dsl modem cables

Discussion in 'Broadband' started by Denis McMahon, Jun 4, 2011.

  1. I'm kind of wondering, what's the point of the shielding on these? It's
    not as if it's grounded anywhere, and this bear of little brain wonders
    if such shielding would (a) have any effect, and (b) if the effect could
    even be detrimental.

    The question is prompted, in part, by the fact that I get over 1 Mbit/sec
    more using a "flat" rj11-rj11 than I do using a round shielded twisted
    pair.

    Could be that there's something fundamentally wrong in the twisted pair
    cable too I guess, but I'm wondering if we're being sold a crock of shit
    for the last 2 metres.

    After all, it's unshielded twisted pair over most of the path from the
    exchange. Surely all that STP on the last 2m is going to do is provide
    more capacitive coupling between the two wires in the pair than
    equivalent utp would, for no significant gain in noise reduction, tp
    being fairly resilient to noise anyway.

    Rgds

    Denis McMahon
     
    Denis McMahon, Jun 4, 2011
    #1
    1. Advertisements

  2. Denis McMahon

    Peter Able Guest

    Don't quote me Denis, but I could imagine that when the cable runs parallel
    with cabling that might be carrying signals/noise there could be common-mode
    induction that might be outside the modem's working range (I've seen it)
    and, possibly some differential induction due to relative proximity -
    although the latter is what twisting is meant to minimise! There are STP
    cables where the connection of the shield continues through one or both
    cable end connectors - which makes sense - but, as you say, many STP cables
    just float the shield.

    PA
     
    Peter Able, Jun 4, 2011
    #2
    1. Advertisements

  3. Denis McMahon

    Graham. Guest

    Your logic appears perfectly sound to me, and I suggest you trust your own
    instincts rather than other peoples sales hype.

    I've seen plenty of twisted pair RJ11 cables, there is one bundled
    with every Cisco router I install in the course of my work. I don't recall
    seeing a twisted one coming with a domestic grade router, they tend to
    be parallel.

    What I have never seen myself, is a shielded twisted pair RJ11 cable,
    and as you say, where are you supposed to connect the shield?
     
    Graham., Jun 5, 2011
    #3
  4. The shield is floated, in fact it's not brought out at the ends of the
    cable at all.

    The cable is a "belkin high speed adsl modem cable" which I purchased
    because I wanted twisted pair all the way - but as I've said, I actually
    get lower performance, 5500 kbits / sec vs 7000 kbits / sec, when I use
    this cable compared to a "flat" (d section) rj11-rj11 cable for the last
    2 metres.

    So either the belkin cable is duff (and I don't have test equipment to
    test beyond basic continuity / resistance) or stp with a floating shield
    is actually detrimental .... and I remember enough of my electronics
    engineering to realise that the capacitive coupling between a twisted
    pair inside a floating screen will be much more complex than that between
    a normal twisted pair.

    Rgds

    Denis McMahon
     
    Denis McMahon, Jun 5, 2011
    #4
  5. Denis McMahon

    Graham. Guest

    years ago I was involved with a contract that required a data link
    between a satellite Rx and an 8 bit computer, I think the data was
    9600bps with no handshaking. We has some special cable to wire
    the link, one pair of wires with a screen. The pair was not twisted,
    and the screen was left floating at both ends. I never could see the
    point, and I am sure telephone cable would have done, It would have been
    one less roll to carry in the car.
     
    Graham., Jun 6, 2011
    #5
  6. Yeah, and like I said before, my recollection of electrical engineering
    from my ONC and HNC is that if you put a twisted pair inside a floating
    screen, then not only do you have the mutual capacitance of the wires to
    each other, but you'd also have capacitance between each wire and the
    screen, so your total capacitance between the wires per unit length would
    be c1 + c2/2 where c1 is the direct capacitance and c2 is the capacitance
    between each conductor and the screen.

    Or am I mis-remembering?

    Rgds

    Denis McMahon
     
    Denis McMahon, Jun 6, 2011
    #6
  7. on average the interconnector capacitance will be the same whether or
    not the shield is grounded.

    The grounding of the shield helps remove external interference, thats all.
     
    The Natural Philosopher, Jun 7, 2011
    #7
  8. Yes, I know all of that.

    I'm speculating that stp with any shield will have *more* inter-connector
    capacitance than utp.

    I'm also speculating that an the ungrounded shield on the wall socket -
    modem connection will, regardless of effect on inter-connector
    capacitance, have no additional noise reducing benefit over that which is
    achieved by using utp.

    Finally, I'm observing that a belkin stp (floating screen) "high speed
    adsl modem cable" drops my downstream rate from approx 7000 Kbits / sec
    to approx 5300 Kbits / sec compared to a flat unscreened rj11 - rj11 lead.

    Now, I might have a duff belkin "high speed adsl modem cable", as the
    only tests apart from observing its actual performance that I can do are
    basic continuity, but the observed performance of the cable combined with
    my own understanding of the electrical engineering involved suggests to
    me that the shield on the belkin "high speed adsl modem cable" is
    essentially snake oil.

    Rgds

    Denis McMahon
     
    Denis McMahon, Jun 7, 2011
    #8
  9. I think even an ungrounded shield will remove most of the interference.
    Essentially, the interference can still only get inside it at the ends.
    Or, which comes to the same thing, only very low frquencies would get
    through (wavelength comparable to the length of cable I should imagine).
     
    Charles Lindsey, Jun 8, 2011
    #9
  10. Denis McMahon

    Sam Guest

    Not sure of the technical terms but I recently replaced my faulty RJ11
    cable and my download speed almost doubled.

    Not sure of the difference between the two cables.
     
    Sam, Jun 11, 2011
    #10
  11. Well, I recently installed an adslnation splitter at the NTE, and have a
    separate cat 5 run from the adslnation "unfiltered" punchdowns to near my
    adsl modem / router.

    The original phone wiring is attached to the filtered side punchdowns on
    the adslnation faceplate.

    I've tried the following:

    2 different rj 11 modules
    2 different rj11 - rj11 flat cables
    an rj11-rj11 stp (floating screen)
    an rj11-rj11 utp

    I also tried using a BT secondary jack and cutting one rj11 connector off
    of one of the rj11-rj11 flat cables and attaching a BT431A in its place.

    So far, none of the results have approached the original rates I was
    getting with microfilters, despite the fact that there's a long parallel
    connection from the NTE to 3 sockets, and the computer etc is located
    near the middle one of the three.

    At the moment, I'm starting to think that the adslnation faceplate isn't
    all that it's cracked up to be. I may look for a passive filter faceplate
    with filtered and unfiltered punchdowns and try that instead.

    Rgds

    Denis McMahon
     
    Denis McMahon, Jun 12, 2011
    #11
  12. you are damed right. Its nothing special. But its neat.
     
    The Natural Philosopher, Jun 12, 2011
    #12
  13. Denis McMahon

    Windmill Guest

    Not sure if anyone mentioned this already, but there used to be
    security issues with unshielded UTP in the long gone days of 10Mb/sec.

    Apparently it was possible to decode the signals from some distance away.

    Whether that is still true with 100Mb/s and 1000Mb/s networks I don't
    know.
     
    Windmill, Jun 13, 2011
    #13
  14. I'm now considering the Pressac and Austin Taylor faceplates as stocked
    by Solwise. Anyone here got opinions on these two?

    Downstream attenuation is 42 dB on 2+, which I believe indicates 3km ish
    of cable between me and the dslam (which looks about right from street
    maps, exch is about 2.5 km by what I would consider the most logical
    route for cables).

    Rgds

    Denis McMahon
     
    Denis McMahon, Jun 13, 2011
    #14
  15. Yes, any radiated signal can be received, and any received signal that
    contains some sort of encoded data can be decoded if the encoding is
    known.

    google "tempest security"

    Rgds

    Denis McMahon
     
    Denis McMahon, Jun 14, 2011
    #15
  16. Denis McMahon

    Phil W Lee Guest

    (Windmill) considered Sun, 12 Jun
    2011 23:00:48 GMT the perfect time to write:
    Let's just say that there are good reasons why those who are highly
    security conscious prefer fibre, as well as encryption of all network
    traffic.
     
    Phil W Lee, Jun 14, 2011
    #16
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.