More ADSL/Phone Wiring Qs

[Rick Hughes]

BT also has gas discharge devices on every line at the exchange

[Andrew Gabriel]

That's done inside the ADSL modem. The microfilter passes the ADSL line straight through as a direct connection. (If you don't have a phone on the line, you don't need a microfilter at all.)

The original CD50's and Hayes Accura modems with BT callerid support were rather polarity sensitive.

Back in the days of BT Approval, one of the tests for equipment was that it was not polarity sensitive. This was abandoned with BABT Approvals (which only care about equipment not damaging the phone network, and not if the phone quipment will actually work or not).

[w_tom]

Which was posted:

A protector (ie Gas Discharge Tube) connects from each line to earth ground. Each GDT does not connect between two lines as with the Clarity device. That would do nothing useful. GDT during a surge must connect each wire to earth to provide protection. Furthermore, separation between GDT and BT's computer is also part of the protection. Protectors adjacent to electronics do not provide that additional protection. Just another weakness of that Clarity device.

[w_tom]

If this pick-up of noise was a problem, then all POTs lines would be filtered at the service entrance AND the DSL line would run separately to the modem. But JohnDW is discussing crumbs. These noise sources are microvolt and nanovolt sources. Totally irrelevant to a DSL millivolt DSL signals.

Furthermore, POTS phone equipment does not generate 10 Khz noise (of millivolt levels). Filter to stop 10 Khz noise from POTS equipment - also arguing over crumbs.

Andrew Gabriel has accurately described the only releveant purpose of that filter:

Without that filter, POTS equipment would eat the DSL signal.

Why are DSL filters close to a DSL modem? Convenience. Its just easier to install. When we install from scratch, we put one fitler right where the phone line enters a building and we run a dedicated wire to the DSL modem. It simplifies design. All POTS wires carry no DSL signals. When putting filters at every POTS phone, then other wires also carry and reflect DSL signals (and any 10Khz noise that may be picked up). Separation of a POTS and DSL line verifies DSL is only on its one intended wire. DSL modem connects directly to the DSLAM in a CO on one hardwire connection - no branch circuits to other dead end circuits. Clean, neat, and only one filter to leak off DSL signals (rather than many). Simpler installation where all other non-DSL phone wires are also completely separated from (do not carry any) DSL signals.

But for installers, a filter at each POTS phone with a new connector (rather than tracing wires and installing a new DSL wire) is easier.

Andrew Gabriel accurate described the only purpose of that filter. Those other noise sources too small (crumbs) to be relevant.

[Andy Burns]

yeah, I realised I'd described it ass-backwards, and said so a bit further down the thread ...

[The Natural Philosopher]

And IIRC on every master socket in every house..well there s SOMETHING anti-surge.

BTW surge protectors are only good against induced spikes of relatively low power from ground strikes adjacent to a line. Nothing will protect against a direct line strike. I know. I had one.

Most equipment that is connected to phone lines has adequate protection: If it didn't there would be too many returns on it. The cost of protecting against a full blown strike is more than the cost of giving away free replacements when it happens.

In short, don't worry about surge protection.

[JohnDW]

In my experience, it's those crumbs of noise that add up when you are working on installations with a marginal service anyway. If you have a good service, then, I agree, there is unlikely to be a problem with the small additional noise.

Cheap, plug-in filters must work somewhere. However, I've an installation with a reported negative s/n ratio... This was vastly improved by changing over to a faceplate splitter.

I find it better to at least try to use a single faceplate splitter to get the signals separated as soon as possible. It's easier than to try to debug the apparently intermittent problems in marginal installations when you introduce the unknown quality of in-house wiring into the equation. As BT have found, it gives you a convenient point of demarcation :-)

[Huge]

Hear, hear. Much easier to use a faceplate splitter and go ethernet ASAP.

[w_tom]

There may be one downside to a single filter - when all POTS phones connect to a DSL line at one point via one filter. Too many phones (too many ringers) may saturate a DSL filter inductor causing a short reduction of DSL signal when POTS phones ring. I have heard it reported - but never saw the problem. DSL modems I used would monitor DSL signal levels - no signal degradation displayed. Apparently these filters did not suffer inductor saturation with four phones.

Meanwhile, I detest some DSL (or cable) modems that do not provide a status table with signal strength. How can anyone see a problem if signal strength (in dBs - not is silly bars) is not reported on a status page?

[Huge]

Because 99.997% of ADSL users wouldn't know a dB if it hit them on the head.

[Dave Liquorice]

Yes but it is only across the pair so any common mode surges will not be supressed, as described by Mr Gabriel. Lightning induces common mode spikes and surges that are "looking for" earth.

[Dave Liquorice]

When trying to squeeze every last bps out of a DSL connection any additional noise of distortion is a bad thing. The interference from distant broadcast MF stations at night (and close ones during the day) is enough to affect noise and sync rates on DSL. Plotting the number of symbols v carrier I can see the notches that the close by MF stations cause. I just wish I knew what punched a blooming great notch at around

300kHz...

A single filter as the line comes in then either site the DSL modem mext to that filter and run ethernet to the computer/network switch or if you must a single pair direct to and only to the DSL modem. This keeps all stubs and bad terminations to a minimum and stops the unbalanced ring wire injecting noise just to one wire via the ringing capacitor in the normal master socket.

[w_tom]

That reasoning tells us nothing useful. 99.97%+ don't even know what octane is but make sure their gasoline has enough.

dB is even simpler. If dB is 0, then signal is too small. If the db is 8, then signal is mostly good. If dB approaches 20, then DSL is solid. Nothing complex. Speedometers with Kph and Mph numbers are more complex. But then many will deny what they never learned or used. Back to the original problem. Some modems are designed with dB on status screens as if the user was an idiot. In another post, the benefit of that dB number on status screens should be apparent.

[w_tom]

1) Using a faceplate splitter lets all other phone line in the house act as antennas for more DSL noise. 2) Faceplate splitter is not as soon as possible. It is as late as possible. 3) Using 'faceplate splitters' at every phone reduces but does not eliminate each POTS phone 'eating' a DSL signal. Each filter still 'leaks' DSL signal to a POTS phone. Each filter 'slightly reduces' signal to the DSL modem (otherwise phone would massively reduce that signal). Using one filter for all POTS devices means more DSL signal from CO goes to the modem; less leaked through other filters. 4) And finally, 'faceplate splitters' means many 'stubs'. Each stub or branch line to a POTS phone is where DSL signal goes up that wire, reflects back, and creates what also creates ghosting on a TV. These stubs create reflections that distort a DSL signal because modem receives reflected copies of that same signal. Putting the filter at the service entrance means no 'stub' to reflect DSL signals.

If one neither has nor appreciates why dBs on a status screen are important, then one also would not learn experimentally of these minor losses and distortions of DSL signal. Therefore one could not appreciate the concepts; learn why one filter and why no DSL signals on other branch 'stubs' results in better DSL reception.

Putting a filter (a DSL signal blocking device) closer to a DSL source means putting the filter closer to telco CO and DSLAM. If wires that carry DSL signals are wires only going to a DSL modem, then DSL signal will be cleaner and have less noise.

'Faceplate splitters' mean wires to each filter reflect DSL signals (distortion) and act as antennas to pickup more DSL frequency noise. Connecting all POTS devices through one filter at service entrance, instead, removes these sources of noise and distortion. How much? Again, this is why a dB number on the status page is so useful.

[w_tom]

Those induced spikes are made redundant by protection already inside all modem and other appliances. We even elininated surge 'noise' on a long wire antenna with an NE-2 neon glow lamp (same glow lamp sometimes inside switches). Induced transients are that trivial.

A protector is installed for a massive surge (direct lightning strikes) that might overwhelm protection inside all appliances. If such protection did not work, then BT would shutdown and disconnect all phone service whenever a thunderstorm approaches. Direct strikes - common mode surges - occur often to a network of overhead and underground wires; and made irrelevant because of earthed protectors.

[Andrew Gabriel]

[snip]

Hum -- you misunderstand what a faceplate splitter is [in the UK]. See:

[tony sayer]

Tom!, Your slipping.. not a word about earthing for lightning;-))....

[Huge]

Ah, I think we have a terminological inexactitude. I meant have one filter where the line enters the house. God knows what you meant, because after your posting arrogant patronising cack like this;

You're off to the killfile with all the other shitheads.

*plonk*

[Huge]

[Huge]

Jesus, what an arse. So, you reckon some wanky little surge protector is going to handle a kiloampere lightning strike?