Which logically leads to the conclusion that filters are not needed.

Which we know is untrue.

Point 1.
You yourself stated that the broadband side is straight through, i.e. the DSL and phone signals are present all the way from the MSAN/DSLAM to your modem. A filter does not and cannot make a difference to the broadband signal attenuation.

Point two.
If a phone is connected to such a circuit without a splitter two things can happen. You may not hear the signals themselves, but it's worth listening to see if you hear anything at the point you switch your modem(-router) on.
More importantly the signal-carrying wires of the phone and any extension cabling easily pick up electromagnetic noise in their environment. Any ring wire present also being a prime factor in noise pickup.

Filters, as you say, filter the high-frequency signals coming down the line. They also prevent any picked up noise on the filter to phone wiring from getting back onto the line to the MSAN/DSLAM.

Good modern filters also make sure noise picked up by ring wires are prevented from getting onto the line. Cheapos may not.

None of the above affect the signal attenuation!. They affect the Signal to Noise Ratio which generally we don't see, and because they affect that they also affect the Signal to Noise Ratio Margin (post-sync), which is what we do see reported by our modem(-router)s.

The filters are in the PHONE side; and are High-Frequency Reject Filters/circuits, to prevent the ADSL/VDSL signals from being attenuated by the PHONE load

is simply complete rubbish and the sort of thing that leads to internet myths.

As for connecting a phone directly to an NTE with no router on the circuit, and not getting any problems, something you stated twice in that post - ?????? What else would anyone expect?

In an earlier post you said "The line is normally very clean on the 17070 QLT". Short of thunderstorms in the area it should be always very clean. If yours isn't, I suggest you have a problem.

The poster seems to have proved to his complete satisfaction that phone and broadband work perfectly with a simple line-splitter, no filter.

The point about the DSL signals being outside hearing range is valid. I shall try sometime to see if anything is audible without a filter when synchronising .

I may be wrong on the audible side, but I am most certainly right that they are nothing to do with preventing the phone load attenuating the broadband signal . Which was in fact my main reason for posting in the first place. Dangerous misinformation.

Point 1
At least we are agreed that the xDSL connections are straight through.

Regarding the filter and possible attenuation from the PHONE side, assume that it is a very poor filter, impedance in the low ohms - would that not place a degree of load on the xDSL side and thus cause some attenuation?

But agreed with the assumed high impedance of such filters across quite a wide band, that attenuation should be minimal.

So theoretically possible; but unlikely in practice.


Point 2 - Modem switch-on.

It would be awkward for me to carry out such a test properly, requiring the co-operation of another person because of the relative locations of the items.

I have tried the following.

With all my POTS and Modem wiring as normal, I did use the 17070 QLT on the adjacent (corded) phone, leaving that running whilst I unplugged, waited about 10 seconds and reconnected power to the modem.

The phone was on its speaker side, at maximum volume.

There were no sounds emitted by the phone over the whole 5 minutes 20 seconds that the operation took, apart from the QLT announcement about every 15 seconds.

The modem was re-powered at about 30 seconds, taking another 3 minutes to establish full contact, followed by my taking the computer on-line for the remaining QLT time (automatic QLT shut-down) and longer.

As I have already stated, that was with the house phone/broadband wiring "as normal" - but it is slightly unusual.

The modem has a direct separate "Ethernet" or similar cable straight down to the NTE area in the cupboard; but originally was connected to the Phone wiring locally via an VDSL filter.

When I obtained that cable, I left all of the xDSL filters scattered around the house in situ, including the local one, just in case I ever had to revert back.

At the NTE in the cupboard, I fitted a new (dangly) VDSL filter directly in to the simple, ancient, unfiltered NTE, with the new direct modem cable in to the straight-through VDSL side, whilst all of the various phone lines were connected in to the PHONE filtered side of it.

Thus all of the phone equipment effectively have two filters in series, the (new) one at the NTE and their own individual ones at the various phone outlets.

It does work, getting typically 37/9 on a 40/10 contract, at about 300 Metres electrically from the FTTC.

The new direct cable to the modem, 10 Metres, brought about a 200 Kbps improvement in Dowload.

------------

Regarding the QLT, I check that in odd moments about once per week; and have never had any problems with such, in xDSL days.

But about 1988 in the good-old dial-up days, 300 Baud and all that, there was a problem with noise on the line - and it was helping the BT Phones lad to track that down, that made me aware of the otherwise invisible peregrinations of the "D-Side". (Crimp joints re-made)

It was also back about then that I had that line attenuation measured, it turning out to be about 23 db over the full length, Exchange to house.

It cost £25.

I think you are incorrect in that belief. It may be that some modern phones have low pass filters built in them, but that wasn't historically the case, and there's a lot of that stuff still around. That's particularly so with "steam age" line-powered phones (which we are all urged to have at one of as they work in power cuts). Line-powered phones tend to present a low impedance across a wide frequency range, which is not a good thing at all for xDSL.

If all the xDSL filters are removed and voice equipment just connected in parallel with the modem, then the symptoms might well include a much lower sync speed or sync being lost when a call is made/received. I have experienced exactly that (with a DECT phone - the ADSL signal would sync until the phone was lifted when a filter wasn't used). Note that splitter is often used as an alternative term for a microfilter (at least according to Kitz). It does not necessarily mean that it's a simple Y connection.

Mostly the low-pass filter is there to protect the xDSL signals. It has a secondary role in preventing xDSL signals getting to the voice equipment (some may be vulnerable, but not those old line-powered phones).

Which belief?

It's eckiedoo that said:-

In reply to a post by eckiedoo:

Since then, I found my simple Line Doubler (ie no filter) and have generally repeated the earlier tests.

The results were the same, although the old corded Test Phone was clearly on-line simultaneously with the the VDSL signals.

I also tested the Test Phone by having it as the only item directly connected in to the old non-filtered NTE.

and

In reply to a post by eckiedoo:

The ADSL and the VDSL sides of the devices, whether you call them splitters or filters; are actually "Straight-through".

The filters are in the PHONE side; and are High-Frequency Reject Filters/circuits, to prevent the ADSL/VDSL signals from being attenuated by the PHONE load.

It seems you either agree with him about the latter, or prefer just to argue with me and leave that unchallenged.

The major misinformation is that the filters are there to prevent attenuation. Funny that the fact the phone signal goes straight to the modem without that attenuation.

An Interesting question regarding this: http://aa.net.uk/broadband-phoneline.html

I guess for fault testing then a filter and phone would be required (if the modem remains sync'd to the exchange) but just for broadband use then a filter would not be required?