View Full Version : One for the telecomms folks...

Shaggy Sheep Driver
22nd Feb 2015, 11:34
Years ago my employer had a second phone line installed at chez SSD so I could work at home, including many conference calls, without hogging the domestic phone. Later, business broadband was added to that line.

When I retired the line became disused though physically left in place, and domestic broadband was added to the home phone line.

A couple of years later the home phone line went dead, and a technician diagnosed a faulty line between house and the distribution pole across the road, so swapped that pair for the disused former business line pair. All was well for a couple of years. Then we got noise on the phone line and poor broadband speed and another technician diagnosed water in the cable between the distribution pole and the roadside cabinet, so swapped our line for a spare 'dry' pair in that cable. It got rid of the noise on the line, but killed the broadband. So swapped it again for another dry pair - again, no broadband. As the voice line now checked out as OK he declared the line to be fine and reported a fault to his broadband colleagues.

Broadband technician came out and confirmed the broadband signal was OK in the exchange, OK up to the roadside cabinet, but wasn't reaching the house. So he said it had to be the line between the cabinet, via the pole, to the house (he wasn't qualified in pole-climbing so couldn't check the signal there). Line man came again, and did another cable-pair swap. Still no broadband.

After much complaining to BT we finally got a broadband techy who could climb poles. He discovered the broadband signal was OK at the pole, so he dismantled the junction box where the line entered the house and confirmed the broadband signal was OK there. Just downstream of this junction box was a small white plastic box. he opened this and discovered - an RF filter! No wonder the broadband signal wasn't working - being very high frequency it was being suppressed by the RF filter! When the filter was removed, everything worked just fine.

Now the mystery; that filter was installed as part of the business phone line some years before, and that line subsequently supported business broadband with no problems.

How had the business broadband worked OK with that RF filter in the line? And indeed had supported domestic broadband for a couple of years when that former business line was utilised between pole and house to bypass the first line fault. It only stopped working with broadband when the pair was changed in the cabinet-to-pole cable.

22nd Feb 2015, 11:56
See here (http://forums.broadband-finder.info/technical/t/4106535-rfi-filters-on-bt-lines.html)

Shaggy Sheep Driver
22nd Feb 2015, 12:01
See here

Yes, but that just confirms broadband will be blocked by an RF filter in the line. The point of my post was how come that line (with filter in place) did support broadband with no problems, up until the pair was changed in the cable between the cabinet and pole? I do have a theory, but I'm not a telecoms engineer so could be wrong.

Shaggy Sheep Driver
22nd Feb 2015, 16:30
I was wondering if a fault on the original pair made the circuit operate unbalanced. When that pair went faulty and were replaced by a correctly balanced pair the filter did its stuff and killed the broadband?

tony draper
22nd Feb 2015, 18:00
I can climb poles,ere well I could a few years ago,well a few decades ago would be nearer the truth,ere, not sure about now. :(

22nd Feb 2015, 19:15
The original dial up modems used audio frequencies to transmit data. As data speeds increased it was discovered that the cables between the junction box and the customer could support greater speeds but the regulations regarding possible electrical interference being radiated from the phone lines required the fitting of filters. Data is transmitted essentially using on/off keying and the sharp edged square waves used could generate harmonic frequencies capable of interfering with radio and television transmissions. The filters fitted at the phone socket and in-line were designed to keep radiated interference levels as low as possible and also served to restrict the maximum data speeds, protecting the income stream of dedicated data cabling companies.

The latest techniques use modulated radio frequency carrier waves to convey the signals. These high speed connections do cause low level interference across a wide band of frequencies, resulting in an overall increase in the background level of radio noise. This is especially annoying to radio amateurs and radio astronomers who are generally interested in trying to receive very weak signals.

For the majority of the population, the increase in the background levels of radio noise is of no great consequence. Such is the price of progress.

John Hill
22nd Feb 2015, 19:24
Data is transmitted essentially using on/off keying

Are baseband modems really still in common use?

Shaggy Sheep Driver
22nd Feb 2015, 19:52
ADSL and VDSL (the broadband transmission technology between the router in the home and the fibre optic of the internet) uses multiple carrier frequencies, hence the term 'broadband'. It's these that use the copper telephone lines between the exchange and the home, or the cabinet and the home in the case of FTTC.

Fibre broadband is an oxymoran - fibre uses a single frequency laser light source so is really narrow band - uni band, even!

The Flying Pram
22nd Feb 2015, 19:59
One for the telecoms folks...Having been (gainfully) employed by GPO/POTel/BT from 1973 to 1996 I suppose I qualify?:)

Those RF filters were brought out to deal with medium wave radio interference which affected many of the first generation electronic push button phones, and some electronic microphone inserts. Prior to that, rotary dial & carbon microphone type phones were virtually immune to such problems. However, bad joints (particularly between different conductor materials), could also lead to unfortunate subscribers (customers weren't invented back then), hearing radio programmes when trying to use the phone.

Having said all of that, you mentioned that you had two separate lines at one point, and also described several occasions when engineers "swapped pairs" over. I'm wondering whether your originally working line ended up on the other pair, including the filter. Later types of overhead "dropwire" contain two pairs - if these had been continued into the internal cabling (possibly via the filter), it maybe that one of the engineers swapped them over completely from pole to internal sockets.

The mere mention of noise means corrosion somewhere in the system, and a high probability of a simple diode being formed at the fault. You then have a perfect "crystal set" with an excellent long wire aerial. Filters were often fitted in the hope it would solve radio pickup, rather than trying to locate the real problem. As for "Another technician diagnosed water in the cable between the distribution pole and the roadside cabinet, so swapped our line for a spare 'dry' pair in that cable" All non pressurised cables (which means from cabinets onwards) should be grease filled so it is very unlikely for water to get far, even from damage to the sheathing. But joints are another matter - various methods have been utilised over the years: Lead "Plumbed" sleeves, Expanding plugs in the end of plastic sleeves, Self Amalgamating Tape to seal cables to tapered sleeves, or end caps, "O" ring seals with clamps, and latterly heat shrink sleeving. All of these are fine when new, but after a few opening and closures leaks develop, and that's normally where water gets in.

From memory, the filters were simple low pass designs, allowing the speech band of 300-3400kHz through, whilst attenuating higher frequencies. Since broadband makes use of mainly audio bandwidth (both below this segment, and above it), it's possible that one would not cause significant problems, particularly on a longer line which already limits the bandwidth and download speed. Those filters were pretty basic and cheap with undoubted wide variations in component values, so I doubt that any two would be alike.

I agree it all sounds rather odd, but with SO many possibilities of cable pairs to be swapped out, joints cut and reconnected, to say nothing of variations of internal wiring and master/slave sockets, I often wonder why broadband works at all...

Edit: Having typed this lengthy tome out, I now see the extra comment from G0ULI - I think there may be some confusion over these filters. The ones I've described are what I suspect the O/P was referring to, and were about well before broadband was in common use. The "microfilters" that are required with all current installations are to separate the speech band from the rest of the available bandwidth. Try your phone without one of these and you will hear a "rushing" sound when the router has established a connection with the exchange. I confess I don't know what actual transmission methods are now used, particularly with ADSL2, but I think the interference G0ULI mentions is not broadband itself, but the widely deployed "Powerline" networking plugs, which put unscreened, non twisted pair, mains cables to uses they were never meant for...

22nd Feb 2015, 21:06
Quote "Fibre broadband is an oxymoron - fibre uses a single frequency [email protected] light source so is really narrow band - uni band, even!"

Yes, but only if the carrier wave is carrying no information i.e. it is unmodulated.

As soon as you start to send information - voice or data or whatever - the spectrum occupancy becomes wide in proportion to the rate at which you are sending that information. This is what the "broad" in broadband refers to.

Shaggy Sheep Driver
22nd Feb 2015, 21:16
In broadband, the media carries multiple carrier frequencies, each of which is modulated to transmit data. A single carrier frequency, even when modulated, isn't broadband.

22nd Feb 2015, 21:47

From Wiki: "In telecommunications (http://en.wikipedia.org/wiki/Telecommunications), a broadband signaling method is one that handles a wide band of frequencies."

So I agree with what you say about the multiple carrier frequencies etc.

However Wiki goes on to say that:

"Broadband is a relative term (http://en.wikipedia.org/wiki/Relative_term), understood according to its context."

In my particular context, the band of signals I receive via Beardy's fibre optics is much broader than when I had an old style dial-up system - so I call it broadband.

This may not be strictly the correct term to use - maybe wideband is a better one - but it has become accepted through common usage.

Shaggy Sheep Driver
22nd Feb 2015, 21:59
Indeed. Hence the oxymoronic term 'Fibre broadband'. Technically incorrect, but 'accepted'.

23rd Feb 2015, 00:10
The Flying Pram

I have used modems since 300 baud was regarded as fast. The certification process to legally use these in the UK was ridiculous. Thankfully technology has moved on and the necessary filters are just routinely built into equipment now.

The trick of swapping pairs of cables to get a better (less noisy) connection was one I've seen used many times. It was the easiest trick to try since all phone line connections are made with four core cable and only two are normally used in a domestic setup.

I have glossed over the actual technical details because they are pretty boring but I agree with the points made in your post.

23rd Feb 2015, 19:58
Quote "Fibre broadband is an oxymoron - fibre uses a single frequency [email protected] light source so is really narrow band - uni band, even!"

Pah ... you're wrong there my friend.

You can do WDM on fibre and send up to 40+ different light frequencies down the same fibre.

So therefore if you've got termination kit running at 10Gb a port, you could run 40+Gb down a single fibre pair using DWDM. :cool:

John Hill
23rd Feb 2015, 20:03
I have used modems since 300 baud was regarded as fast

Pffft...! I have been involved in digital communications since we made the bits of the binary digits individually by hand with a telegraph key!:hmm:

BTW, we were using Morse Code on AFTN channels in some parts of the Pacific until the mid 70's. I saw Morse being used in an airport communications centre in 1999 (that was in a country that is forbidden to be mentioned on this forum).

23rd Feb 2015, 21:35
G0ULI has a point. This pollution of the RF spectrum may affect a few people - although we know some of it affects broadcast signals - but that sort of justification suggests that because it makes life easier and cheaper for some people, we should allow raw sewage discharge into rivers from a major metropolis because it is financially advantageous for those inhabitants but a big problem for the small village down stream.....

An argument many intellectually lacking idiots cannot accept, including the usual incompetents in the EC!

Shaggy Sheep Driver
23rd Feb 2015, 22:17
You can do WDM on fibre and send up to 40+ different light frequencies down the same fibre.

Very true, but broadband to the home doesn't use Wavelength Division Multiplexing except probably to carry more than one subscriber's data over each fibre cable from the exchange. Each subscriber's connection uses only one frequency on the cable, though many subscibers may be connected through a single fibre cable..

ADSL and VDSL however runs mutliple frequencies on each subscriber's own dedicated copper pair to carry the data from the exchange (ADSL) or cabinet (VDSL) to the home. So for each user, the only part of their link to the internet that is truly broadband (using multiple frequencies) is the copper part of the link - ADSL or VDSL.

Dont Hang Up
24th Feb 2015, 13:11
A single carrier frequency, even when modulated, isn't broadband.

I beg to differ.

Simple amplitude modulation at a single frequency introduces two equidistant side tones, thus defining the bandwidth required to carry the modulation.

I agree that if the carrier is light (frequency order 100 TeraHz) then the frequency shift of the modulation side bands are likely to be very tiny in comparison, but they still exist.

24th Feb 2015, 15:08
I beg to differ.

So said the aptly named Dont Hang Up the "Air Traffic Research Scientist" from Malvern, UK....

I wonder what gives me the idea they might know what they're talking about when it comes to comms signals ? :cool:

John Hill
25th Feb 2015, 09:16
Simple amplitude modulation at a single frequency introduces two equidistant side tones, thus defining the bandwidth required to carry the modulation.

Are modulation techniques presently in use on fibre optics not all variations of ON/OFF keying? Is amplitude modulation used at all? Or is ON/OFF keying considered amplitude modulation?

Dont Hang Up
25th Feb 2015, 13:00
Are modulation techniques presently in use on fibre optics not all variations of ON/OFF keying? Is amplitude modulation used at all? Or is ON/OFF keying considered amplitude modulation?

ON-OFF keying is effectively a square-wave amplitude modulation of 100% modulation depth.

It is perhaps the greediest of all data transmission techniques in terms of bandwidth. For laser light in its own dedicated fibre this is not really an issue, which is fortunate because better techniques using frequency modulation or phase shift keying are technically challenging for a laser source.

However if you tried on-off keying as your preferred technique on a radio transmission then all the radio hams in the county would form a lynch mob and come and find you! :)

John Hill
25th Feb 2015, 18:45
Dont Hang Up,

I always thought ON/OFF keying is exactly what I do with my 'ham' transmitter and all the other transmitters I sent Morse Code during my early career on circuits that ICAO called 'Manual A1 Simplex'?