CRC and HEADER errors when phone rings

Now, after having a load of issues which never got resolved [disconnects, DLM raising SNR, lower speeds] (BT engineers out, line tests, new face plate, swapped twin cables etc, et al - nothing wrong!) I still only get 2.5mbish [plus 4mb cap] sync even though I am 1.5km from the exchange. In the old days pre this ADSL2+ thingy I was always syncing around 5~6 mb.

Well, while decorating, and running a script on my router lots of times, I noticed that when the phone rings, CRC and HEADER errors go up a lot, like 2 (for each) every 5 seconds. If the phone doesn't ring, they sit still and stable (no errors) - I don't have to answer the phone either for this to happen, just let it ring!

So I unplugged the phone. Rang my number again, and, yep, CRC and HEADER errors go up again in the same fashion.

So what is going on here? Is my line truly that bad? Why can a simple phone call affect a ADSL signal if I have no phone connected? What can I do next after BT engineers say all is hunky dory*?

Nick
*I guess as the Government's aim is to get everybody at least a 2mb connection, then this is acceptable.

Edited by deleted (Mon 27-May-13 17:49:21)

Can you run RouterStats-Lite and check the noise margin graph when the phone rings?

http://www.vwlowen.co.uk/internet/files.htm#routerst...

It doesn't change - plus it happens when the phone is unplugged!

Nick

plus it happens when the phone is unplugged!

The phone has nothing to do with it. It'll be the 75 volts a/c coming down the line when rung in to that is causing the issue.

But surely here will be no voltage/current if the phone is not plugged in to ring? I mean, it would be like a battery that isn't connected to anything - there is nowhere for the voltage/current to flow?

Nick

In reply to a post by Lethe:

But surely here will be no voltage/current if the phone is not plugged in to ring? I mean, it would be like a battery that isn't connected to anything - there is nowhere for the voltage/current to flow?

Nick

There is plenty ...

For a ringing to appear on the line there does not need to be a bell circuit - the potential of the line will be raised to 75v (AC). With the phone disconnected there is the normal line termination of a series capacitor and resistor in parallel with a spark gap - that will allow an AC current to flow.

Then there is the filter and that will have a path between the A and B legs - exactly what will depend on the filter design but both AC and DC will pass.

Even taking out the termination circuit and filter there would still be 75V present on the line and current will flow through the front end of the modem.

it would be like a battery that isn't connected to anything

which still has the voltage present across its terminals - stick your tongue on it to see (the battery, not the phone line).

You might have a line fault that is reacting to the ring voltage, or a filtering issue at one end or the other.

From: http://www.amazon.com/Implementation-Applications-Te...

6.2.1.4 -- Rectified Loop Faults

Corrosion (or verdigris) across PCB tracks or between adjacent pins on telephone sockets are prime examples of this type of fault condition. Whilst stable linear loop faults do not give much cause for concern to DSL systems, rectified loop faults are more of a problem because they are nonlinear and introduce cross modulation (a sort of self interference which cannot be equalized away). They are also often very dynamic in nature and can stay in an "inactive" state until they become activated, e.g., by a DC line polarity reversal which may be associated with the application of ringing current to the associated analog interface. Essentially, these types of faults are "turned-on" when an incoming call is received at the ADSL customer's analog interface - the customer has no problem making outgoing POTS calls but may experience problems with premature ring trip (false answer) and high ADSL bit errors on incoming calls.

cheers, a