From Newsgroup: uk.d-i-y

On Tue, 17 Feb 2026 13:40:59 +0000, Andrew wrote:

Does anyone know if the faceplate of the latest NTE5C socket has a
filter ?.

There are a bunch of different faceplates, some of which contain a filter
and some which don't.
The basic faceplate with a single phone socket does not contain a filter.
The chunky faceplate with both phone and RJ11 sockets does contain a
filter (on the phone port, and the passthrough to the extensions).
The faceplate with a single RJ11 does not contain a filter.

There's other versions too - some designed for back-feeding voice from a router (with a green socket), which don't contain filters. These don't
seem to have been popular and may be discontinued.

Mike

--- Synchronet 3.21b-Linux NewsLink 1.2

From Newsgroup: uk.d-i-y

On 17/02/2026 15:02, The Natural Philosopher wrote:

On 17/02/2026 14:37, Andrew wrote:

On 17/02/2026 14:06, The Natural Philosopher wrote:

If it doesnt have a phone socket it doesn't need a filter

It has a 'phone' socket but that only plugs into the BT Smart Hub.

Is that a big flat BT phone socket or an RJ11 ?

The usual socket that a landline phone would plug into. Since
it is intended for modem use only, it would have made more
sense to use the same style of socket that is on the back of the
phone (or modem) and supply a suitable double-ended cable

What I can glean about the latest BT pile-o-crap is that it has RJ11 for
DSL and an VOIP interface to it's-a phone port

-aSo you wall plate will have an unfiltered RJ 11 unless its *also* got
a BT style phone socket. If BT supplied that when you went VOIP.
I would advise leaving it as is since that's what BT expect to find when doing support

The phone plugs into another (green) phone socket on the back of the
router.

Interesting. VOIP? Or just a filtered DSL socket?

No, just the original landline phone plugged into the router.

It looks a bit naff.? Do you still have a landline?

Yes, (for now) plugged into the back of the router so now IP connected

Ah! VOIP equipped router, Nice!

I would advise some contact cleaner and leave things as they are

I'm surprised that it had such an effect. I'll check more regularly.

Trust me. RF and oxide is a recipe for total unreliability.
Back when we flew model planes-a on 35Mhz the 'rusty fence' syndrome
would cause issues at the edges of flying fields

But the contacts are supposed to be gold-flashed, unless they
came from China where faking Gold and Silver (in addition to lots
of other stuff) and cheating their fellow citizens, seems to be
the accepted way of doing things.

https://www.youtube.com/watch?v=iZxwQ69ax1g

PS When I enter 192.168.1.254 a message now insists that I use
HTTPS but that doesn't work because the https is in red with a line
through it, prefixed with an exclamation mark in a red triangle.

This is the bit that is relevant...

"If you receive a security warning when accessing 192.168.1.254, it is
generally safe to click "Advanced" and proceed, as the warning
indicates a self-signed certificate, not that your connection is
actively being compromised."

The router is self certified. So encryption still works, but any
certificate is deemed 'illegal' by the browser
Just tell the browser to naff off and connect anyway

I do click on advanced, and ignore the warning.
I don't know any other way of proceeding to enter the password.

That's how you do it

You may be able to globally disable that in the browser.

--- Synchronet 3.21b-Linux NewsLink 1.2

From Newsgroup: uk.d-i-y

The Natural Philosopher <tnp@invalid.invalid> wrote:

On 17/02/2026 14:03, Theo wrote:

Presumably if it's single socket now copper voice is dead, it's now just carrying a broadband connection and there's no filter. The filter was only ever to prevent DSL signals from interfering with the voice - you didn't need a filter for a data-only connection.

Other way around. It's to prevent voice stuff degrading the
broadband...it is essentially a low pass filter to the phone.

Well, it does both. The voice on the line is low pass filtered at the
exchange at 3.4kHz. The broadband doesn't come anywhere near that low - the original ADSL Annex A started at 25kHz[1], and I think VDSL does too[2].

The broadband modems will happily ignore voice frequencies, but there's no guarantees that the user's random handset also has a sufficient low pass
filter to block the DSL carriers. Hence the microfilter drops them before reaching the handset.

Some people can hear 25kHz directly, but also it could cause resonances and other effects in handsets which were never designed to accept it.

[1] https://en.wikipedia.org/wiki/ADSL#ADSL_standards
[2] https://www.ofcom.org.uk/siteassets/resources/documents/research-and-data/tv-radio-and-on-demand-research/radio-research/analysis-of-bt-openreach-vdsl/rsgb-vdsl-report.pdf?v=324383

In theory, the extra connectors add a bit of insertion loss so even in copper voice days it was better to have the DSL plugged into the wall socket
directly if you never used the phone (or used it at another extension).

What actually happens is that the contacts slowly oxidise and create a crappy diode junction which is just another bad joint on the uplink.

That starts mixing the RF all together and generating side bands and the
DSL takes one look and drops half its RF buckets...

That's what can demodulate (or rather downconvert) DSL carriers into the
voice band, where the filter won't block it and you can hear it.

The attenuation is unnoticeable compared with the copper uplink to the cabinet.

Your theory doesn't have actual numbers.

It's minimal insertion loss in resistance terms, but if the microfilter is dirty or capacitive then it adds a little reactance, which could degrade the SNR a little.

DSL 'works' even in the presence of an airgap, but it doesn't mean you get
the best performance. You want the fewest impedance discontinuities for the best speed. I agree I don't have numbers on that.

In practice it's the oxide that is the problem. It is in fact the grain
of truth behind gold plated connectors rather than lead/tin or silver plated

Yep.

Theo
--- Synchronet 3.21b-Linux NewsLink 1.2

From Newsgroup: uk.d-i-y

On 17/02/2026 14:39, The Natural Philosopher wrote:

On 17/02/2026 14:34, Nick Finnigan wrote:

On 17/02/2026 13:40, Andrew wrote:

As per subject

The whole front of the new NTE5C master sockets unclips a bit like the
lower part of the earlier NTE version, with a inner socket on the
backplate fixed to the wall.

-a-aMaster Socket 5C MK4 has a Pressac VDSL filer inside the chunky front >> section. Voice seems to have pins 2 and 5 direct connection.

No matter what it seems, it is the broadband that has the direct connection if there is in fact a broadband socket (RJ11?) at all.

There is a an RJ11, which also has direct connections, to incoming pins 2 and 5 (unlike voice, after more investigation). Presumably to allow for different configurations.

--- Synchronet 3.21b-Linux NewsLink 1.2

From Newsgroup: uk.d-i-y

On 17/02/2026 22:44, Theo wrote:

Your theory doesn't have actual numbers.

It's minimal insertion loss in resistance terms, but if the microfilter is dirty or capacitive then it adds a little reactance, which could degrade the SNR a little.

DSL 'works' even in the presence of an airgap, but it doesn't mean you get the best performance. You want the fewest impedance discontinuities for the best speed. I agree I don't have numbers on that.

Non linear contacts - crude diodes - are what makes DSL cause 'hiss'
into the audio band.

But the actual loss of signal is unmeasurable.
--
It is the folly of too many to mistake the echo of a London coffee-house
for the voice of the kingdom.

Jonathan Swift


--- Synchronet 3.21b-Linux NewsLink 1.2