Attenuation is only part of the story. Good SNR is equally as important for ADSL Max, if not more so.

Attenuation and a mismatch termination is part of the story. Signal to noise ratio obviosly drops, sometimes by 20dB is you rationalise customer cabling. A 2 wire presentation to incoming Broadband is vital. Standard BT configurations give an unbalanced 3 wire.

warm wishes

WH

In reply to:

---

Signal to noise ratio obviosly drops, sometimes by 20dB is you rationalise customer cabling.

---

[Cough]. If SNR _drops_ by 20dB, you must be doing it wrong.

In reply to:

---

Standard BT configurations give an unbalanced 3 wire.

---

Not surprising, since they were never installed with the intention of carrying RF signals. I'm afraid most of us are already well aware of this - one of the forum's members (yarwell) documented this over a year ago, and has been advising people about this, whenever he thought it would help.

Sorry, early Sunday morning. What I meant to say was, when you fit a Filter 2000 at the ADSL take off point and back feed the rest of the customer wiring you can save 20dB or more, depending on number of sockets and length of cabling. If as in some houses they have stranded brglar alarm cabling, this can make matters a lot worse.

Sorry for any misleading information.

Rather like saying that if you connect your TV aerial to the TV with figure-8 bell wire, it will give you a poor picture. It's a "given".

The best that I ever achieved was on a . 512M 'out in the sticks'. When I managed to synch with The Exchange I had attenuation of 65dB and an SNR of 3.5dB. after fitting a filter 2000 it increased to 20.5 a gain of nearly 1000times. Failure can also be caused by 'precise' internal cable lengths. This appears as a short circuit to ADSL. By extending the length by 2 metres or so this can be alleiviated.

In reply to:

---

Failure can also be caused by 'precise' internal cable lengths. This appears as a short circuit to ADSL. By extending the length by 2 metres or so this can be alleiviated.

---

Your reasoning is way up the creek here. Since ADSL consists of a large number of 4kHz channels in the 25kHz to 1.1 MHz range (in ADSL's plain vanilla form), any "tuning effects" caused by cable length should only affect one or two tones. This is why ADSL is so resiliant. What you are probably seeing is what happens to the extension wire when it is working as an aerial - the addition/subtraction of a few 10's of picofarads of capacitance, and a microhenry or two of inductance could tune or detune the wire and affect its effectiveness as an antenna. So your changing of the cable length will be affecting the SNR, not the attenuation. Not only should the ADSL signal be filtered at the entry point, unfiltered ADSL wiring should be kept as short as possible.

Edited by deleted (Sun 11-Feb-07 13:47:59)

I'm afraid that your explanation does not cover the situations that occur when noise interference is added to the circuit on BT's line (outside the customer's property) both over and underground. In certain locations this can be the limiting factor on sync speed. The source could be radio signals in the MF band or just cross talk in the multicore or at the exchange. Attenuation is not the only factor (although it is the most important).

In reply to:

---

Attenuation is not the only factor (although it is the most important).

---

Err, no it's not.