Frequent dropouts 25th Jan 2025

[#1]

@outcast  I disagree with you, only Stability profile and of course DLM reset are available to Plusnet all the others are DLM controlled. Care to give me some examples of Interleaving, G.Inp and Banding being adjusted by Plusnet, they can't do it.

[#2]

@Dan_the_Van

This thread has been busy today!!

With all the information provided here, I would be looking for sources of RFI. There’s at least three PSUs close to the master socket and there’s a loop in the xDSL cable. If they are current switching PSUs they can be notoriously bad for RFI and the cable is in an ideal place to pick it all up.

Later evidence suggests that prior to the change, the line was clamped (one way or another) at around 12dB.

Now the line is negotiated at 6dB yet the observed SNRM is a little over 3dB which suggests an increase in RFI subsequent to the initial sync speed negotiation. There’s an environment problem to be found and eliminated, which has been masked by the historically high SNRM (and associated lower speed).

With that thought with an upwardly adjusted SNRM the resultant sync speed might be little different to what the line was before.

[#3]

@Townman 

If you refer to message 8 you can see since the change to SoGEA the connection has become less stable, the current equipment setup has been in place for sometime. The OP has had issues with the Master Socket there is wood bar across the front to protect it (master socket)  from being kicked by accident.

Something happened 25th Jan which prompted this thread, hopefully changing the connection to 'Stable' will help

[#4]

From the log information provided (see message 16), the target SNRM appears to have changed from 11dB to 6dB, however the observed SNRM is circa 3dB. That infers the possibility of RFI.

Prior to the change the high SNRM will have masked the impact of the RFI. Unfortunately the hub two does not support dynamic monitoring.

[#5]

Don't get me wrong I am not disagreeing with you, just presenting more of the story.

RFI is not always obvious as a cause to issues, in some cases "fast as you can go" is not for the best.

I have a TP-Link powerline adapter plugged into a power extension strip along with four other powerpacks including the Hub Two, the powerline adapter is less than 30cm from the Hub, the DSL cable runs in the same conduit as the mains cable to the strip. The Master Socket is next to the mains socket for the strip, I have no obvious RFI issues.

[#6]

---

@Townman wrote:

... however the observed SNRM is circa 3dB. That infers the possibility of RFI.

That depends whether the DLM target SNR is 6dB or 3dB.

If the target SNR is 6dB then yes the line might be struggling with noise, which could be RFI, but if that was ongoing wouldn't you expect the SNR margin to get bigger.

If the target SNR is 3dB then it might be a sign that the DLM algorithm is trying too hard to maximise speed at the expense of stability.

.

[#7]

The target SNRM is in the connection log record - 6dB

We need to see the last connection record and the current SNRM.

[#8]

@Dan_the_Van

I do not get you wrong her (or elsewhere). There’s a lot going on here and the usual culprits are not looking like the villain here. Observed SNRM being markedly different to that when the xDSL session was negotiated is disconcerting.

Now you come to mention it, dLAN devices can play havoc with VDSL modulation. My Devolos have a configuration option to mitigate such.

It would be interesting here to note the current SNRM and then switch off all of the non-essential devices plugged into that power board … and see if the SNRM changes by anything significant.

[#9]

Your discussions are too technical for me to understand hence I can't contribute. I am simply waiting to see how the change to 'Standard' plays out but only time will tell.

I can only reiterate that before my new contract on 31st October 2024 the broadband was stable. It had been 127 days since the previous dropout and nothing has changed with the setup at my end.

I'll update when I have some more facts.

Thanks for all input.

[#10]

"I can only reiterate that before my new contract on 31st October 2024 the broadband was stable"

That is readily explained by the fact that whatever line instability measures were present on the service have been removed.  That will be what has contributed to the increase in line speed ... but at the cost of stability.  There is every possibility that your line has been performing well below par for a good while due to things you can see within your own environment.

Given that there is no fault detected on the line, the next logical consideration is that you have electrical interference on the line (RFI).  You have a lot of electronics next to the master socket - one of those PSUs (I see three) could be the source of the interference.

You might well find that, as is, you line will not sun faster with stability than it did previously.  If you do have environmental interference, you will have to live with it ... or get a little technical to identify and eliminate the source.  It is not too difficult, there is lots of guidance readily available on the internet.

Recently one user on here eventually realised a correlation of disconnections with the use of their kitchen lights - they were GU10 LEDs ... totally took out his connection.  A service configured for high stability (lower speed) might well be tolerant of such ... until the line configuration is modified for optimal speed (over stability).

[#11]

---

@Madeleyite wrote:

Your discussions are too technical for me to understand hence I can't contribute

The techies are trying to express concern that your incoming phone line can be sensitive to electrical noise, that can cause the symptoms you are seeing, and consistent with what your modem stats are reporting.

For example you appear to have a laptop power supply resting on the incoming telephone wire -

Laptop power supplies are known to be noisy electrical devices that happen to use an oscillator that runs at a frequency within the frequency range that your incoming VDSL broadband uses.  So simply moving the laptop PSU perhaps nearer the desk surface, might help your modem stats.

Likewise, a shredder uses electric motors (that cause all sorts of noise), so having that cable so close the the phone line might be worth reconsidering.

.

[#12]

And there appears to be a loop (stuck up with tape) in the modem cable.

The two other PSUs should not be ignored either.

It is possible (out of sight) that the printer has an external PSU too.

There are all potential sources of RFI.

We are not so much 'guilty' of being "techies" but like it or not, internet transmission is technically complex and sometimes we just have to learn what is required for optimal performance if that is what is desired.  Not everything is in the realm or gift of the ISP or BT Openreach to fix; sometimes the best that they can do is mitigate matters, by slowing a connection to improve connection stability.

If the disconnections are nothing more than inquisitiveness (that is they have no material impact on the use of the internet) for the benefit of improved speed, one might well just leave this as it is, if investigating the possible presence of the RFI is thought to be too much trouble.

[#13]

@Townman @outcast 

Looking at the pictures of under the desk there may not be much scope for @Madeleyite to change the layout of the equipment and associated cabling.

I think a more programmatic approach is needed, if setting the line to 'Stable' works then no further action is needed as the EU will be happy. But if it doesn't then further investigation would be advised.

[#14]

Thanks. I have rearranged the cabling as best as possible (without a major relocation of the Belkin 8-way surge protected adaptor and subsequent new longer length cables) to move PSUs away from DSL wires and master socket. See photo.

The DELL laptop PSU is now 30 to 40cm away from the Master socket.

The shredder cable is now 9cm away from the DSL incomer and master socket.

The printer cable (no PSU) is 10cm away from the master socket.

I am unable to move the Speaker PSU.

All I can do now is to wait and monitor the situation.

[#15]

This might be beneficial... 1M Cat6a Shielded Modem router cable VDSL RJ11/RJ11 vdsl BT Infinity, Sky, adsl | eBay