Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[burakkucat]

Please don't read too much into my musings regarding "a physical examination" . . . If, for example, Walter, I or another techno-kitizen were nearby one of us might do the (unofficial) deed. 

Am I correct in saying that at one time you had two separate "lines" installed? Two separate telephone numbers? Two separate NTE5/As? Both would have been supplied via that drop cable, as it contains two pairs. That is how I understand the situation, based on your mention of a "line being cancelled".

[OldTimer]

Absolutely right Bcat. Two lines, two numbers, one cancelled when we had our smart tv and needed to re-position the then BT HH so as to connect to the TV by ethernet cable. Also, at that time, when investigating what I needed to do, I was advised by a Beattie sales person that I would also benefit by getting increased DS/US speeds if I re-positioned the HH. It seemed to be a no-brainer so I went ahead and did the deed. DS certainly improved, but US went south, in fact half of what the actual throughput was before. Before on an actual Audio/Video file upload I was getting 106 K/Sec, now it is mostly 45.6 K/Sec.

I have spent hours interacting with Beattie, and to be fair I have had two Openreach engineers vist. The first one fitted me the SSFP Master Socket, but was unable to do a full line test because of a fault at Beattie's end, the second renewed the line from the 'sausage' type joint on my house to where it connects to the Master Socket. It seems evident to me that Wholesale have given up on me and said there is nothing more they can do. The rest as they say is history!

You're always welcome to have a cup of tea if you are ever in this neck of the woods. 

Regards,

OldTimer

[burakkucat]

Thank you for clarifying the situation and confirming my suspicions. 

There is another option available to you. That is to change to a CP (ISP) who will take note of your current problem and then chase Openreach, in the correct fashion, until the defect is fixed. I won't make any recommendations . . . because such a CP will undoubtedly be more expensive than BT Retail. 

[OldTimer]

Just by way of interest, I have just uploaded a 42 minute audio file to the destination server in the USA, and monitoring the SRNM levels throughout showed no discernable variation in the US level for the 6+ minutes that it took to upload. Predictably, the throughput speed was around 46.6 K/Sec.

So far, since I installed DSLstats, the only noticable variation in the US levels have been caused by making and receiving phone calls.

Regards,

OldTimer

[roseway]

There seems little doubt where the problem lies. If your ISP won't organise a proper thorough investigation, and changing ISP isn't an option, then it comes down to waiting and hoping that the problem soon gets much worse, so it can't be ignored.

[OldTimer]

Earlier today I posted a message on the BT Care Forum. I sought to explain that after three solid days of DSL uptime I noticed this morning that I had lost the DSL connection. There were several other observations I mentioned at the same time:

1: ........... after several days of the US speed being 444 Kbps, it seems to have dropped to 168 Kbps. The US SNRM has dropped from just over 21 dB to the present level of 18.6 dB.

2: The DS SNRM dropped from the usual daytime SNRM of around 6.1 dB, to the then level of 3.8 dB.

3: The DS speed had gone up from 4717 Kbps to 5323 Kbps.

The reply I received explained that during the day, for most people, 6dB is normal, but (it wasn't said when) BT have apparently introduced a new normal daytime standard of 3 dB (for good connections). The thinking behind this is that it gives a little more connection speed. Apparently, due to the fact that some lines are unable to maintain the 3 dB, large numbers of errors are sometimes produced during the night, which has the effect of reducing the throughput speed. The possibility exists therefore that such large error-producing lines need to revert back to the previous norm of 6 dB.

Like yourselves the guys on the BT forum are volunteers, so they have my respect, but since receiving that message I have observed something I did not expect on the DSLstats graphs.

Earlier in the day the usual spikes occurred in the US SNRM line when I received a couple of phone calls. They could be seen quite clearly on the snapshot graph.

Then...around 3.15pm (ish) following an outgoing call the expected 'downward' spike didn't happen, instead it went slightly UP from around 18.9 dB to just below 20 dB.

This can be clearly seen in the graph. Then after an unwanted 'nuisance call' (which I did not answer) the SNRM dropped back to around 18.9 dB where it has stayed until now.

I am sorry to weary you again with all this, but I'm just trying to understand a little more about the apparent results of BT's change in SNRM standards. Do you have any thoughts on this?

Apart from the SNRM and Bitloading graphs would any of the others be useful in throwing any more light on what may be happening with my line/connection?

Regards,

OldTimer

[burakkucat]

The changes you are observing in the SNRM graphs (upwards and downwards steps, sudden peaks or valleys) are very clear indicators of one or more marginal joints (HR or semi-conducting) in your metallic pathway. 

The normal condition of the pair is a state of being biased in one direction by the exchange battery but when an incoming call is signalled, that nominal 50 VDC has an alternating voltage superimposed upon it. The elevated voltage then present can form "micro-welds" across such a marginal joint, for example. Or it can just be that the alternating voltage is sufficient to break down a semi-conducting joint. The net results are changes in the SNRM graph such as you have shown, above.

[JGO]

"The reply I received explained that during the day, for most people, 6dB is normal, but (it wasn't said when) BT have apparently introduced a new normal daytime standard of 3 dB (for good connections). The thinking behind this is that it gives a little more connection speed. Apparently, due to the fact that some lines are unable to maintain the 3 dB, large numbers of errors are sometimes produced during the night, which has the effect of reducing the throughput speed. The possibility exists therefore that such large error-producing lines need to revert back to the previous norm of 6 dB."

Sometimes my speed is dropped overnight due to slowly failing S/N, (when the system is not in use anyway grrr !) , and  it doesn't recover by itself anything like as fast next morning when the S/N  improves. Me doing a quick reset is quicker. This sounds like a potentially good idea to me.
 
 BUT as your problem is a bad line, it doesn't seem very relevant to you until the bad joint is fixed.
 

[OldTimer]

Good evening one and all. I've not been in touch for quite a long time, but I just thought to update you on an interesting development within the last couple of days regarding my upload problem.

A Beattie Mod rang me regarding an e-mail issue I raised with them, and, to be fair, he was quite helpful. While I was speaking with him I mentioned the upload issue, and he suggested removing the cap that another Mod had put in place. We agreed that I would monitor the situation and get in touch again should that prove necessary. Well, when I logged on this morning I was delighted to see that in theory my upload speed was back to what I thought I should be getting on adsl2. The max attainable was showing as 1,000 Kbps, although the actual throughput when I actually uploaded an audio file was quite naturally somewhat less. Nevertheless I was well pleased.

I activated DSLstats and started recording - you're probably there before me, but later on in the morning when I made a couple of phone calls there was a massive spike on the upload line at around 11.13am (see 1st image attached - use slide bar to view relevant part). Immediately after the call the upload SNRM had gone from around 5.9 dBs right up to around 18 dB where it has remained, more or less since then.

This afternoon my wife made another call, again there was a further recorded spike, downwards this time, to around 8 dB. However, as soon as the phone call finished the upload SNRM went back up to around 18 dB (see 2nd attached image).

The current upload line rate is 370 Kbps. So, I'm just about back to where I was before the Mod suggested removing the cap.

He said that he would ring me tomorrow, it promises to be an interesting call.

Regards,

Alan

[roseway]

That's the classic symptom of a high resistance line fault. If it's repeatable it should be simple enough to diagnose.

[OldTimer]

Well, well, well! Today I had a visit from a really helpful Openreach engineer. He was a young man, only having been in the job for two years, but I take my hat off to him. I had fully explained my problem to him, plus commenting that it had been suggested to me that my 'issue' was very likely the consequence of a High Resistance line fault. Other engineers haven't agreed, but this one listened! I also showed him the DSLstats snapshots (thanks Eric). After his various tests he finally agreed that there was a HR fault on the line.

He went away for a little while, but commented on his return that he had traced the fault 'back to the exchange', which presumably means that it was somewhere between the exchange and the street cabinet. He further added that he had put me on another line, and lo and verily behold although my download speed is very slightly down to it's current speed of 3707 Kbps, the upload speed has risen dramatically to 876 Kbps (from the 147 Kbps it was first thing this morning).

Before I finally cry problem solved, I will monitor the situation for the next couple of days, but with fibre due to be activated on my line on Wednesday at least I do have the comfort in knowing that the HR fault was 'before' the street cabinet, so any issue should be solved anyway.

What can I say to all you good folk? I am so pleased and thankful to all of you who took an interest in what for me was a really taxing problem. The breakthrough is entirely down to your very helpful and knowledgeable advice, without which I wouldn't have been able to point the engineer towards where the problem might lie. Many, many thanks 

Regards,

OT

[burakkucat]

Thank you for sharing the good news.

As you have said, it does appear that the fault was in the E-side (between the PCP and the exchange). To have an upstream speed of around 850 - 900 kbps is what I would expect. (Concerning the abbreviation for the kilo- prefix in SI units. It is, actually, lower case "k". Why? Because the unit of temperature, Kelvin, uses upper case!)

Hoping all goes well on Wednesday.