Packet Errors

[#1]

---

@Fat_Leonard wrote:

I have tried a Quietline test and there is a faint, consistent background hiss (nothing like the crackling and fizzing I've heard during previous faults) but it did drop out to complete silence for a fraction of a second several times while I was on the line. It was quite similar to listening to the runout groove on an LP, but irregular.

---

That might suggest that there is some interference which could lead to line instability ... and thus an elevated SNRM with a consequential lower speed.

Which router are you using?  Do you have access to a Plusnet Hub One?  The Hub One is a BTHH5a for which there is a customised version of router stats.  There is also a full version and lite version of the product which runs with a significant number of routers.  See the link below.

It would be interesting to see what the SNRM plot looks like.  Such can tell you a great deal of what is happening on an unstable line.

[#2]

Hi all, I've just had a look into this for you and can see that the downstream speed is as low as 5mbs at present but I'm not seeing any obvious causes. Sorry if you've already done this and I've missed it, could you try a speed test whilst connected to the master socket? 

https://community.plus.net/t5/Library/Testing-From-The-Master-Socket/ba-p/1322242

- Adam 

[#3]

---

@Townman wrote:

---

---

That might suggest that there is some interference which could lead to line instability ... and thus an elevated SNRM with a consequential lower speed.

Which router are you using?  Do you have access to a Plusnet Hub One?  The Hub One is a BTHH5a for which there is a customised version of router stats.  There is also a full version and lite version of the product which runs with a significant number of routers.  See the link below.

It would be interesting to see what the SNRM plot looks like.  Such can tell you a great deal of what is happening on an unstable line.

---

The router is a TP-Link Archer VR2800 v1. The original little square Plusnet router we were sent didn't have enough wireless oomph to go through our stone walls, and not enough ability to control advanced settings like port forwarding or static IPs. The TP-Link also has a Guest Network function which is very useful.

We have now been sent a new Plusnet router which looks like a HomeHub from when we were on BT, so I guess that's the Hub One you're talking about. I can plug that in and try your tests, but we need the guest network today so it will have to be tomorrow. Looking at the RouterStats page the VR2800 isn't listed.

[#4]

Thanks Adam.

The router is plugged into the master socket with a dangling microfilter, as the filter faceplate appeared to be linked to the problems I was experiencing before.

Do you mean 5Mb/s? If so, that's about the maximum we have ever got so not an indication of a problem.

[#5]

Hey @Fat_Leonard I did yes, so it does look like you're getting the maximum available downstream speed up to the router at least.

[#6]

Use there links to check which router you got.

Check you router lights

Your router lights can tell you a great deal about what is happening when things are not as they should be. The colour of the lights and their state will tell you all you need to know ... if only you could find the help sheet which came with the router. That information can also be found in the depths of the on-line help & support guides. The following links make the material easier to find.

• Hub Two
• Hub One
• Hub Zero (2704n)
• TG582n

Little square one was probably the Hub Zero - not a great router on several fronts and does not work with Router Stats.

[#7]

I'm now using the Hub One. The connection is very poor today. Heavily dropping packets in VOIP, BT Wholesale Speedtest doesn't even load, Think Broadband's test shows 300ms average ping with 0.5Mbps down and 0.2Mbps up. Hub light is solid blue.

[#8]

Can you please post the router stats - line speed, SNRM, attenuation and errors.

Also can you consider installing and running router stats for the BTHH5a (aka hub one). See the link below.

[#9]

Thanks, Townman. Routerstats is connected and starting to run.

Current connection has improved, with no intervention from our end. The initial RouterStats box is showing the attached, which I hope will give you the numbers you requested.

[#10]

On revisiting this, I am reminded that you are on a rather long line to the exchange and on ADSL.  The US looks banded.  The DS has improved significantly, the question is how long will that last?

Can you set the SNRM graph capture running please.  That will be the more interesting metric.

[#11]

I set it running last night, but the PC restarted for some updates. Here's today's capture set.

[#12]

This is great stuff and tells a lot!!

There is no interest in any graphs other than SNRM (all of the time) and connection speeds (only at the time of a resync).

This line has become worse since your previous line stats.  DS SNRM is much higher and the speed is slower.

The SNRM plot is ugly!  the basically square wave suggests that there is something giving rise to electrical noise, which switches on for around 8 minutes and then goes off for 15 or so minutes.  That might be a refrigerator.

The line does though twitch all of the time, which suggests that there are other sources of interference.  These are going to take some effort to track down.

[#13]

---

@Fat_Leonard wrote:

We only have the master socket and no other extensions.

---

Well that rules out one possible culprit for the ingress of REIN - extension wiring and the presence of the ring circuit.  Just to be crystal clear on this - there is only the inbound BT line pair connected to the master socket back plate?

Do you have an AM/MW radio?  If you do have a wander around with it tuned to 612kHz - see ::. Kitz - REIN, SHINE & RFI .::

[#14]

Thank you for the suggestions. I've been wandering around with a radio causing much household mirth and invitations to wear a tinfoil hat. The master socket is at the end of a long wire that was passing close to an area the radio suggested was 'hot', so I rerouted it and that caused a minor improvement. Nothing else in the house seems to be particularly radiative, with no change in the plot despite turning off most of the house in turn at the fusebox. Of course I couldn't turn off the ring the router's on, but did my best to turn off everything else on it except the PC running routerstats.

The result that has me scratching my head is that unplugging the cordless telephones (BT 4500 two station and answerphone) and turning off the handsets caused a jump to nearly 15dB.

Currently the number is hovering around 10.5dB after cable rearrangement and changing the order of things plugged into an extension lead. Using a mains filter box had no effect, nor did removing the Powerline adaptors that feed a second PC.

[#15]

This is really excellent diagnostics - great to see you have really engaged with the method.

There is now far less jitter on the SNRM - though it does bounce around a little, the line is much smoother than the previous graph.

The SNRM jumping to 15dB (at 18:35 ?) with the DECT phone unplugged somewhat suggests that its PSU is generating 3dB of noise - that alone is worth one "notch" towards improvement on the DLM.

"The master socket is at the end of a long wire that was passing close to an area the radio suggested was 'hot', so I rerouted it and that caused a minor improvement."

Can you please explain that a little more clearly please?  It implies...

• The circuit to the master socket is moveable (not nailed down)
• Might not be the first socket in the property - some older installations had branch boxes - as standards change, one extension get converted to a master socket ... which is fine for voice but utter dread for broad band

Can you describe the set up in some detail - ideally can you look inside the master socket and get a picture of the wiring.  If it is not twisted pairs it will be a good receptor aerial for REIN!!

Are those nasty spikes you flicking fuse box switches?  Did the router lose the connection?  If it did, without the matching sync speed plot, it is not possible to make a meaningful comparison of the SNRM before and after the event.  The SNRM figures at the end of the plot are the best seen, with speeds slightly improved from the previous plot: DS 4384 (3936) / US 896 (864).

Small steps, but steps in the right direction.