Router connection dropouts

[#1]

I'm not disagreeing.

---

@Dan_the_Van wrote:

For what it is worth performing a buffer bloat test always gives me a result of D on a good day I might get a C.

---

Yes, and knowing that means that if you see a "D" you can assume that it's normal, whereas an "F" is a red flag.

Clearly there are thousands of people using a Hub2 with a "D" score who are perfectly happy, so no need to comment.

My point about the waveform test, is that it captures a snapshot of many relevant metrics in one test, as a diagnostic starting point.

---

@Dan_the_Van wrote:

Over diagnosis of an issue in the first instance should be avoided as it can add a layer of complication that's not required.

---

Having the most complete set of metrics in a single test, doesn't mean that you have to react to it, or over diagnose anything.

The waveform test simply gives the maximum amount to data in one simple test, without the user avoiding telling you something.  The test itself is no more difficult to run than any others, and provides a convenient and simple URL to reply with - rather than having the user figure out how to do screenshots, upload the image, and post a picture in their reply, etc.

The issue I most often see is something like, someone is complaining that their YouTube is buffering, so you ask them to screenshot a speed test (of both upload and download speeds),  but because they think that they have a download speed issue, they repeatedly don't provide the upload result (because they don't think or accept that it's relevant).  Whereas getting them to provide the waveform result URL, you can see the upload speed and latency - so are able to point out that the speed test you asked for was done while another user of the home network was uploading something, or a mobile was doing a cloud backup etc, and invalidating the entire test.

The problem with doing one test at a time approach, is that network conditions change between tests, and you never know when the person you are trying to help is now doing the next thing you've asked for, or if they've disappeared for the day (or week) - so you are waiting around for a reply that sometimes never comes.  Again, doing the test that provides the most metrics, means that you and I don't waste our time waiting for the next response.

.

[#2]

As this thread has got somewhat sidetracked and off topic, I'm reposting the last relevant post for @82sg  to respond to -

---

@jab1 wrote:

@82sg Thanks for the event log. From that, I think it is safe to say you are having too many DSL drops for it to be claimed you are 'running normally'. Personally, I would go back to PN and ask them to carry out some more checks. I appreciate you have said previously that you have done reboots/checked cabling etc, which obviously will give rise to some of the drops, but very recent behaviour still suggests there is something not quite right.

Before you do go back to PN, though, can you post your 'Technical Log - Information' screen, so we can see your connection status -  obscure your user name in it.

---

.

[#3]

Hi,

Sorry for the late reply, it’s been a hectic day. Thanks to everyone for the suggestions too. Here is a screenshot of the technical log. Timing is great as the router just reconnected again.

[#4]

@82sg Looking at that Technical log, I am convinced there is a fault somewhere, despite PN's denials. Your line attenuation suggests you should be getting around 60/65Mb/s connection speed, your SNR margins are not indicating any issues, and yet you are restricted to about half what I would expect you to be getting.

I appreciate it is a pain, but can you give us a another export of your error log, please?

[#5]

Hi John,

Not a problem, the latest extract is attached and thanks again for your help.  I’ve kicked off the automated help check again so will see what it comes back with this time.

Thanks,

Gary.

[#6]

I'm going to guess it will come back with 'no fault found' - again, but there is one. Wait and see what it says, and we can take it from there.

[#7]

It actually came back this time with a speed fault and an engineer is due out on Thursday. I’ll post back here with their findings once they’ve been out and checked everything. 

[#8]

Miracles do happen. 🤣

[#9]

---

@jab1 wrote:

Your line attenuation suggests you should be getting around 60/65Mb/s connection speed,

I'm intrigued why you would conclude that ?, my estimate would be nearer 42Mb if it wasn't capped at 40Mb.

The latest log file suggests to me that there is an intermittent external line fault somewhere,

as the connection seems to toggle between 32Mb and 40Mb, with no pattern related to time, or SNRM.

Hopefully the engineer will find something !

.

[#10]

OK, I might be a bit optimistic, but we shall see.

[#11]

@jab1 

At the risk of being lectured (not by you) over the years looking at Hub event logs I came up line attenuation against Maximum data rate matrix

Typica Line attenuation to Maximum data rate

0.7 Km from cabinet Maximum Data rate 35,000 Mbps; line attenuation 21.9
0.6 Km from cabinet Maximum Data rate 44,450 Mbps; line attenuation 19.1
0.5 Km from cabinet Maximum Data rate 50,415 Mbps; Line Attenuation 15.7
0.3 Km from cabinet Maximum Data rate 89.472 Mbps; line attenuation 11.6

This is for guidance only, no way set in stone as many factors can change Line Attenuation. 

As we know https://www.broadbandchecker.btwholesale.com/#/ADSL would be the best indicator of expected speeds

@82sg 

Looking at the event log you posted there is a device with MAC Address 80:91:33:E5:13:F7 which is repeatedly trying to connect using your WiFi. I doubt it is the cause of your issue but not helpful for your LAN performance.

MAC Address 80:91:33:E5:13:F7 lookup comes up with AzureWave Technology Inc. If that helps identify the device.

It is likely a wireless password issue.

Do you use your Landline Phone? Are there any redundant phone extensions around the property as internal extension cable can add to the Line attenuation resulting in slow speeds

 

[#12]

@Dan_the_Van Thanks for that data. I can't really comment, as I moved from PN ADSL (with an attenuation of ~39) and speed of around 12Mbps, to FTTC with Zen, whose 7530 Hub, at least in its default configuration, didn't quote attenuation figures, merely estimated copper lengths, and at around 400m, said I should expect a  max of around 60Mbps.

It would possibly be useful if the OP could give us an estimate of his 'last mile' length?

[#13]

@jab1 

Must not forget ADSL line attenuation would be to the exchange, VDSL would be to the cabinet.

Yes it would be helpful to know the last mile length

Me - 0.3 Km from cabinet Maximum Data rate 89.472 Mbps; line attenuation 11.6

Looking at Speed against distance from the cabinet 60Mbps / 400m looks like you received the expected speed.

[#14]

Yep - I know ADSL was to the exchange, which in my case was about 3 miles away.😉.

Actually, I was on the 40/10 VDSL product - there was enough difference in the price of 40/10 vs 80/20 to suggest it wouldn't be worth it for me - I didn't, and still don't, need the speed.

I've only moved from FTTC to FTTP - lowest speed tier - as a bit of 'future-proofing', to be honest.