• Fibre Broadband causing unexpected extra work

    From liz@liz@poppyrecords.invalid.invalid (Liz Tuddenham) to uk.comp.sys.mac on Sun Nov 8 19:30:47 2026
    From Newsgroup: uk.comp.sys.mac

    Following the installation of a fibre/broadband/VOIP router, I found
    that I was getting interference on VHF radio in the 2-metre band and
    beyond. The router was on a shelf and the co-ax coming down from the chimney-mounted aerial ran right past it.

    Most of the noise seemed to be coming from an Ethernet cable, so I tried clipping a ferrite filter around the cable - but it made no difference.
    Moving the cable seemed to have the greatest effect and the positioning
    was quite critical, probably due to a null in the interference pattern.

    Evetually I decided I would have to spend an evening unclipping the
    aerial cables and re-routing them so they ran along the picture rail and dropped down well away from the router shelf. Some interference is
    still there but it is now at a low enough level to make little
    difference to reception.

    This was an extra job caused by the change to fibre broadband that I I
    hadn't anticipated.
    --
    ~ Liz Tuddenham ~
    (Remove the ".invalid"s and add ".co.uk" to reply)
    www.poppyrecords.co.uk
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Theo@theom+news@chiark.greenend.org.uk to uk.comp.sys.mac on Mon Nov 9 00:01:32 2026
    From Newsgroup: uk.comp.sys.mac

    Liz Tuddenham <liz@poppyrecords.invalid.invalid> wrote:
    Following the installation of a fibre/broadband/VOIP router, I found
    that I was getting interference on VHF radio in the 2-metre band and
    beyond. The router was on a shelf and the co-ax coming down from the chimney-mounted aerial ran right past it.

    Most of the noise seemed to be coming from an Ethernet cable, so I tried clipping a ferrite filter around the cable - but it made no difference. Moving the cable seemed to have the greatest effect and the positioning
    was quite critical, probably due to a null in the interference pattern.

    Evetually I decided I would have to spend an evening unclipping the
    aerial cables and re-routing them so they ran along the picture rail and dropped down well away from the router shelf. Some interference is
    still there but it is now at a low enough level to make little
    difference to reception.

    This was an extra job caused by the change to fibre broadband that I I
    hadn't anticipated.

    Possibly the change from 10Mbps to 100Mbps signalling? It think that's changing from 10MHz to 31.25MHz in signalling bandwidth. Perhaps the
    harmonics are covering the 144MHz band? I think you said your wiring was
    old, which suggests it's maybe only cat3, cat4 or cat5? That doesn't have
    very many twists to reject common mode noise.

    You can get shielded and/or foiled ethernet cables (STP, FTP, S/FTP instead
    of UTP - various permutations available) if that might help. If you are replacing your cabling it's worth investing in cat6a cabling - not only is
    it typically foiled, it's also good for 10Gbps ethernet when those speeds become commonplace (we're almost there now).

    Theo
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From liz@liz@poppyrecords.invalid.invalid (Liz Tuddenham) to uk.comp.sys.mac on Mon Nov 9 09:38:54 2026
    From Newsgroup: uk.comp.sys.mac

    Theo <theom+news@chiark.greenend.org.uk> wrote:

    Liz Tuddenham <liz@poppyrecords.invalid.invalid> wrote:
    Following the installation of a fibre/broadband/VOIP router, I found
    that I was getting interference on VHF radio in the 2-metre band and beyond. The router was on a shelf and the co-ax coming down from the chimney-mounted aerial ran right past it.

    Most of the noise seemed to be coming from an Ethernet cable, so I tried clipping a ferrite filter around the cable - but it made no difference. Moving the cable seemed to have the greatest effect and the positioning
    was quite critical, probably due to a null in the interference pattern.

    Evetually I decided I would have to spend an evening unclipping the
    aerial cables and re-routing them so they ran along the picture rail and dropped down well away from the router shelf. Some interference is
    still there but it is now at a low enough level to make little
    difference to reception.

    This was an extra job caused by the change to fibre broadband that I I hadn't anticipated.

    Possibly the change from 10Mbps to 100Mbps signalling? It think that's changing from 10MHz to 31.25MHz in signalling bandwidth. Perhaps the harmonics are covering the 144MHz band? I think you said your wiring was old, which suggests it's maybe only cat3, cat4 or cat5? That doesn't have very many twists to reject common mode noise.

    You can get shielded and/or foiled ethernet cables (STP, FTP, S/FTP instead of UTP - various permutations available) if that might help. If you are replacing your cabling it's worth investing in cat6a cabling - not only is
    it typically foiled, it's also good for 10Gbps ethernet when those speeds become commonplace (we're almost there now).

    I didn't realise the faster types were better screened but it seems as
    though the main source of noise was the 100 Mbps cable between the
    router and the switch, so that is the one I will look into replacing.
    The flexible lead between the router and the wall socket is at least a
    metre longer than it needs to be, so I might try to buy a shorter lead.

    The noise is on spot frequencies from 140 to at least 160 Mc/s (possibly
    more, I didn't check), rather than a continuous hash. This morning I
    was listening to a weak signal with a lot of background noise so I
    switched off the router but found the noise didn't stop, so I think that
    moving the aerial downlead was sufficient - but it won't hurt to change
    the Ethernet leads too.
    --
    ~ Liz Tuddenham ~
    (Remove the ".invalid"s and add ".co.uk" to reply)
    www.poppyrecords.co.uk
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Graham J@nobody@nowhere.co.uk to uk.comp.sys.mac on Mon Nov 9 12:34:18 2026
    From Newsgroup: uk.comp.sys.mac

    Liz Tuddenham wrote:

    [snip]


    I didn't realise the faster types were better screened but it seems as
    though the main source of noise was the 100 Mbps cable between the
    router and the switch, so that is the one I will look into replacing.

    Is this the same switch that gives poor speed? Have you tested that
    cable with a cable tester that measures its performance? (Fluke or
    similar.) If one of the 8 conductors is broken then you could get a lot
    of RF radiation and unexpectedly prro network performance.

    The flexible lead between the router and the wall socket is at least a
    metre longer than it needs to be, so I might try to buy a shorter lead.

    I think a shorter cable is very unlikely to make any difference.

    The noise is on spot frequencies from 140 to at least 160 Mc/s (possibly more, I didn't check), rather than a continuous hash. This morning I
    was listening to a weak signal with a lot of background noise so I
    switched off the router but found the noise didn't stop, so I think that moving the aerial downlead was sufficient - but it won't hurt to change
    the Ethernet leads too.

    Is the amateur radio 2-metre band?
    --
    Graham J
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From liz@liz@poppyrecords.invalid.invalid (Liz Tuddenham) to uk.comp.sys.mac on Mon Nov 9 15:11:37 2026
    From Newsgroup: uk.comp.sys.mac

    Graham J <nobody@nowhere.co.uk> wrote:

    Liz Tuddenham wrote:

    [snip]


    I didn't realise the faster types were better screened but it seems as though the main source of noise was the 100 Mbps cable between the
    router and the switch, so that is the one I will look into replacing.

    Is this the same switch that gives poor speed?

    No, it is the replacement switch that is rated 10/100. It gives about
    9Mbps wheras the previous switch gave 1Mbps on a good day.

    The flexible lead between the router and the wall socket is at least a metre longer than it needs to be, so I might try to buy a shorter lead.

    I think a shorter cable is very unlikely to make any difference.

    It was lying in coils alonside the aerial co-ax. Now I have moved the
    co-ax, it probably won't make much difference but it will avoid the
    untidy coils of loose cable lying on the shelf.

    The noise is on spot frequencies from 140 to at least 160 Mc/s (possibly more, I didn't check), rather than a continuous hash. This morning I
    was listening to a weak signal with a lot of background noise so I
    switched off the router but found the noise didn't stop, so I think that moving the aerial downlead was sufficient - but it won't hurt to change
    the Ethernet leads too.

    Is the amateur radio 2-metre band?

    Yes, I haven't notices much of a problem on the HF bands but the QRM is
    so bad there, it probably wouldn't be noticeable. I moved the downlead
    of the longwire at the same time as the 2-metre feeder, just as a
    precaution.
    --
    ~ Liz Tuddenham ~
    (Remove the ".invalid"s and add ".co.uk" to reply)
    www.poppyrecords.co.uk
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From nospam@nospam@de-ster.demon.nl (J. J. Lodder) to uk.comp.sys.mac on Mon Nov 9 22:26:35 2026
    From Newsgroup: uk.comp.sys.mac

    Liz Tuddenham <liz@poppyrecords.invalid.invalid> wrote:

    Graham J <nobody@nowhere.co.uk> wrote:

    Liz Tuddenham wrote:

    [snip]


    I didn't realise the faster types were better screened but it seems as though the main source of noise was the 100 Mbps cable between the
    router and the switch, so that is the one I will look into replacing.

    Is this the same switch that gives poor speed?

    No, it is the replacement switch that is rated 10/100. It gives about
    9Mbps wheras the previous switch gave 1Mbps on a good day.

    Just curious: why bother with obsolete and slow switches
    when standard gigabit switches are cheap and readily available?

    Jan


    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From liz@liz@poppyrecords.invalid.invalid (Liz Tuddenham) to uk.comp.sys.mac on Mon Nov 9 22:01:38 2026
    From Newsgroup: uk.comp.sys.mac

    J. J. Lodder <nospam@de-ster.demon.nl> wrote:

    Liz Tuddenham <liz@poppyrecords.invalid.invalid> wrote:

    Graham J <nobody@nowhere.co.uk> wrote:

    Liz Tuddenham wrote:

    [snip]


    I didn't realise the faster types were better screened but it seems as though the main source of noise was the 100 Mbps cable between the router and the switch, so that is the one I will look into replacing.

    Is this the same switch that gives poor speed?

    No, it is the replacement switch that is rated 10/100. It gives about 9Mbps wheras the previous switch gave 1Mbps on a good day.

    Just curious: why bother with obsolete and slow switches
    when standard gigabit switches are cheap and readily available?

    1) Nothing on my system needs to run at gigabytes.

    2) A lot of the devices can only run at 10, so even 100 is overkill and
    may not be able to communicate with everything.

    3) Secondhand 10/100 switches are very cheap.
    --
    ~ Liz Tuddenham ~
    (Remove the ".invalid"s and add ".co.uk" to reply)
    www.poppyrecords.co.uk
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From David Brooks@boaterdave@hotmail.co.uk to uk.comp.sys.mac on Mon Nov 9 22:45:59 2026
    From Newsgroup: uk.comp.sys.mac

    On 09/11/2026 21:26, J. J. Lodder wrote:
    Liz Tuddenham <liz@poppyrecords.invalid.invalid> wrote:

    Graham J <nobody@nowhere.co.uk> wrote:

    Liz Tuddenham wrote:

    [snip]


    I didn't realise the faster types were better screened but it seems as >>>> though the main source of noise was the 100 Mbps cable between the
    router and the switch, so that is the one I will look into replacing.

    Is this the same switch that gives poor speed?

    No, it is the replacement switch that is rated 10/100. It gives about
    9Mbps wheras the previous switch gave 1Mbps on a good day.

    Just curious: why bother with obsolete and slow switches
    when standard gigabit switches are cheap and readily available?

    FYI, Jan

    A friend said this to me today .....

    That completely explains it. A five-year-old router will be the Smart
    Hub 2 (released in late 2018), and while it has gigabit Ethernet ports
    on paper, the underlying hardware simply cannot cope with a modern 900
    Mbps FTTP line under real-world conditions.

    The primary bottleneck is the system-on-chip inside the unit. Routing near-gigabit traffic requires significant processing power to handle
    PPPoE encapsulation, stateful firewall inspection, and Network Address Translation (NAT) without dropping packets. The CPU in that older hub
    pegs at 100% utilisation when pushed past 400rCo500 Mbps, effectively
    choking the throughput before it reaches the local network.

    On top of the internal routing limits, its wireless spec is strictly
    Wi-Fi 5 (802.11ac). In real homes, through standard walls and
    interference, Wi-Fi 5 rarely delivers more than 350rCo400 Mbps even right
    next to the antenna. Swapping to BTrCOs modern Wi-Fi 6 router brings a
    much faster dual-core processor that routes full gigabit speeds with
    hardware acceleration, alongside the bandwidth improvements needed for high-speed wireless.

    Bufferbloat and CPU overload on the Smart Hub 2: That five-year-old
    router lacks modern Active Queue Management (AQM). When a gigabit burst arrives, the router's underpowered processor pegs at 100%, fills its
    packet buffers, and begins aggressively dropping packets. Once TCP
    encounters heavy packet loss, its congestion control algorithms slash
    the throughput down to a crawlrCooften dropping straight into the 20 to 50 Mbps range until the queue clears.
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From nospam@nospam@de-ster.demon.nl (J. J. Lodder) to uk.comp.sys.mac on Tue Nov 10 15:14:05 2026
    From Newsgroup: uk.comp.sys.mac

    Liz Tuddenham <liz@poppyrecords.invalid.invalid> wrote:

    J. J. Lodder <nospam@de-ster.demon.nl> wrote:

    Liz Tuddenham <liz@poppyrecords.invalid.invalid> wrote:

    Graham J <nobody@nowhere.co.uk> wrote:

    Liz Tuddenham wrote:

    [snip]


    I didn't realise the faster types were better screened but it seems as
    though the main source of noise was the 100 Mbps cable between the router and the switch, so that is the one I will look into replacing.

    Is this the same switch that gives poor speed?

    No, it is the replacement switch that is rated 10/100. It gives about 9Mbps wheras the previous switch gave 1Mbps on a good day.

    Just curious: why bother with obsolete and slow switches
    when standard gigabit switches are cheap and readily available?

    1) Nothing on my system needs to run at gigabytes.

    Your fiber connection runs at that kind of rate.

    2) A lot of the devices can only run at 10, so even 100 is overkill and
    may not be able to communicate with everything.

    Neverthelss, slow swithes may drag down the whole network.
    And gigabit switches will communicate with old stuff.

    3) Secondhand 10/100 switches are very cheap.

    As are secondhand gigabit switches,
    typically 5-10 euro, around here,
    for 5-8 port ones, sometimes even 16 port.

    Jan

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From liz@liz@poppyrecords.invalid.invalid (Liz Tuddenham) to uk.comp.sys.mac on Tue Nov 10 15:54:32 2026
    From Newsgroup: uk.comp.sys.mac

    J. J. Lodder <nospam@de-ster.demon.nl> wrote:

    Liz Tuddenham <liz@poppyrecords.invalid.invalid> wrote:

    J. J. Lodder <nospam@de-ster.demon.nl> wrote:

    Liz Tuddenham <liz@poppyrecords.invalid.invalid> wrote:

    Graham J <nobody@nowhere.co.uk> wrote:

    Liz Tuddenham wrote:

    [snip]


    I didn't realise the faster types were better screened but it
    seems as though the main source of noise was the 100 Mbps cable between the router and the switch, so that is the one I will
    look into replacing.

    Is this the same switch that gives poor speed?

    No, it is the replacement switch that is rated 10/100. It gives about 9Mbps wheras the previous switch gave 1Mbps on a good day.

    Just curious: why bother with obsolete and slow switches
    when standard gigabit switches are cheap and readily available?

    1) Nothing on my system needs to run at gigabytes.

    Your fiber connection runs at that kind of rate.

    Web pages rarely arrive quicker than they did on 10Mbps Ethernet and
    ADSL, so there is a bottleneck somewhere else in the download path. My nominally 100 Mbps link between router and switch is only running at 9
    Mbps (nine times faster than before), so the speed of the switch is no
    longer the limiting factor; trying to go faster still is pointless.


    2) A lot of the devices can only run at 10, so even 100 is overkill and may not be able to communicate with everything.

    Neverthelss, slow swithes may drag down the whole network.
    And gigabit switches will communicate with old stuff.

    It seems to be working as fast as I need. I have noticed that sending documents to the printer is a bit faster, which is a welcome
    improvement.


    3) Secondhand 10/100 switches are very cheap.

    As are secondhand gigabit switches,
    typically 5-10 euro, around here,
    for 5-8 port ones, sometimes even 16 port.

    Same numbers but in UKP instead of Euros here. As you have suggested,
    there isn't much difference in price between Gigabyte and 10/100
    switches.
    --
    ~ Liz Tuddenham ~
    (Remove the ".invalid"s and add ".co.uk" to reply)
    www.poppyrecords.co.uk
    --- Synchronet 3.22a-Linux NewsLink 1.2