Stuck on 40/10

[#1]

There's just one more thing you could do, which dispenses with the filter and bypasses the faceplate.

Wire the RJ11 router cable directly to the terminals on the master socket backplate as I have - see photo.

Note that for neatness I have fitted a clip-on standard phone socket faceplate over that as it's smaller, but it does nothing other than cover the hole!

[#2]

@HPsauce 

If the back plate is wired correctly then you must use the unfiltered faceplate to interconnect the otherwise isolated extension output to the test socket, so you are no better off.

[#3]

Unless you resync, the line speed will not change.  If the SNRM fell without a line resync, that could be indicative of the presence of REIN.

[#4]

@JHarland  You did not get a 5Mbps uplift from the miniscule change in the noise margin. I believe all your speed increases are down to DLM reducing the SNRM,  apart from the initial DLM reset which removed the 40Mbps cap, why it was banded at that speed I do not know. Your actual throughput speed is also limited by the BT Wholesale IP profile which  tracks the sync speed but does not change instantly. This could be why your speed test had not improved after the latest SNRM reduction  The DLM reset must also have enabled G.Inp error correction on the downstream as without that you would not get the reduced SNRM figures. Your stats in msg#3 show no error correction.

Your line could perform better, I am not convinced by other suggestions such as poor connections in the master socket and even interference (REIN). There is no particular evidence for these problems.

I would suggest investigating your internal phone cabling as your incoming pair seems to come via what looks like very old extension cable which may not even be twisted pair. It is definitely not the incoming cable feed, so there must be some other junction box in the house where this cable connects to the incomer. It would be useful to locate the incoming cable and perform some tests there with the dubious cable disconnected.

@Baldrick1  I am not sure what point you are trying to make. You own post msg#27 describes what is happening to the line, if there were errors caused by a poor connection DLM would never reduce the SNRM to the minimum value.

[#5]

Thanks for the suggestions, do you think it's likely I will see much more of a speed benefit as I'm not that close to the cabinet? Line attenuation is 11.1/18.3 and signal attenuation 11.1/24.7?

[#6]

@Baldrick1 Sorry I'm not sure what you mean by "interconnect the otherwise isolated extension output to the test socket" as I thought it had been established that there was no extension wiring connected and none was required. 🤔

My suggestion (as I implemented) is only to simplify wiring and reduce connections/components (any of which can fail) where all you have is a broadband router connected directly to the master socket. If that's not the situation here please disregard. 😎

[#7]

@JHarland I think you're probably about as good as you can get. I'm on a similar distance/attenuation, slightly lower at 10.7.

My speed at best is 61/11mbps and I'm in a quiet area - village with spread out properties and few if any sources of interference.

You said your speed increased, what are the actual up/down connection speeds now?

I also saw that  @RealAleMadrid queried your incoming cable which does not look like a normal "drop wire" type. However it does look like standard 6-core BT phone cable (CW1308?) which is twisted pair to an appropriate standard for broadband. It might be worth, as suggested, checking back for a joint somewhere just to make sure it's all "in order".

[#8]

---

@RealAleMadrid wrote:

@Baldrick1  I am not sure what point you are trying to make. You own post msg#27 describes what is happening to the line, if there were errors caused by a poor connection DLM would never reduce the SNRM to the minimum value.

---

To quote @JHarland: our latest router stats show that DLM has reduced the target  downstream SNR margin to 3 dB resulting in a slightly higher sync speed. Nothing to do with the rj11 cable or micro filter. 

For whatever reason, the changing of the cable and removal of the faceplate resulted in a reduced SNR and increased speed.

[#9]

---

@HPsauce wrote:

@Baldrick1 Sorry I'm not sure what you mean by "interconnect the otherwise isolated extension output to the test socket" as I thought it had been established that there was no extension wiring connected and none was required. 

My suggestion (as I implemented) is only to simplify wiring and reduce connections/components (any of which can fail) where all you have is a broadband router connected directly to the master socket.

---

You are suggesting that the extension cable connectors can be used as the output to the router without fitting a faceplate. I was simply pointing out that the faceplate plug is used to connect the extension wiring, either to the filtered output of a filtered faceplate, or direct with a non filtered type. That's how extension wiring is disconnected when you remove the faceplate.

Consequently, unless you fit a faceplate the connection is not made and it will not work. So, for it to work the connection still has to be made via the faceplate connector.

Edit

I see where the misunderstanding is. You are assuming that the dedicated extension connectors that you are using are connected to the incoming cable.

[#10]

@HPsauce
Data rate is 7.964/58.497
Maximum data rate is 7.964/59.695

I'm assuming my speed test results showing 50mbps download is due to overheads?

[#11]

---

@HPsauce wrote:

There's just one more thing you could do, which dispenses with the filter and bypasses the faceplate.

Wire the RJ11 router cable directly to the terminals on the master socket backplate as I have - see photo.

Note that for neatness I have fitted a clip-on standard phone socket faceplate over that as it's smaller, but it does nothing other than cover the hole!

---

Now that's a good idea! Would have been a useful thing for me to have ultimately done back in the days of ADSL2+.

However, since I've been on FTTC (10/40) I mysteriously(?) haven't needed to take any 'special measures'.

Or doesn't removing the faceplate break the connection to the extension connector? I forget, as it is so long since I had to remove it. 😀

[#12]

@Baldrick1  Maybe it's the hot weather but why have you attributed my post msg#43 (slightly edited) to @JHarland.

You seem unable to accept that DLM is reducing the target SNRM. Shall we agree to differ on this, I'm not that bothered.🙄

[#13]

@Baldrick1 I think you're mixing my posts up with someone else.  😣 My wiring is clearly (and I stated this) only for setups with no extensions, no phones, no anything JUST a router connected directly to the master socket. I have no extension wiring, no phones (SOGEA).

But YES, just to be clear, you do (I think) need a faceplate of some sort in there to make those connectors live as it bridges the upper and lower contacts on the latest type of master socket as @pvmb pointed out. As I said, I have a plain POTS-style one for neatness, not the dual outlet one.

[#14]

---

@JHarland wrote:
Thanks for the suggestions, do you think it's likely I will see much more of a speed benefit as I'm not that close to the cabinet? Line attenuation is 11.1/18.3 and signal attenuation 11.1/24.7?

---

My line and signal DS attenuation are higher than yours and the sync is much higher ...

Data rate:13.596 Mbps / 78.006 Mbps

Maximum data rate:10.784 Mbps / 74.019 Mbps

Noise margin:2.7 / 1.7

Line attenuation:9.1 / 19.8

Signal attenuation:9.1 / 26.0

[#15]

@Townman Your noise margin is ludicrously low! How does it get to 1.7, no wonder your line is fast? Mine is Noise margin:6.0 / 5.1

Just realised I didn't quote my attenuation figures correctly, they're quite similar to @JHarland :

Line attenuation:10.7 / 18.9

Signal attenuation:10.7 / 20.6