Print

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

[havelock]

Hi

In respect to Soms queries about where the interference comes from, I have theory as I've never been able to get a BT engineer to actually tell me exactly how REIN interferes with ADSL (from experience RFI filters seem to be preventing DSL from working completely in most cases BTW);

First let me say i'm not an electrician by any stretch of the imagination, but from what I remember voltage transformers (such as the 240->12v converter you'd get in a high-fi) work by passing a primary coil onto a circular or square magnet, and the output voltage is generated via magnetic force via a secondary coil. I believe in some PSU's this EMF bleeds out and interferes with the data flow required for DSL communications. I've also noticed that Dimmer switches (especially those on 12v lighting) produce the same result. With treadmils I believe the interference is again Electomagnetic in nature but more likely caused by the Motor driving the belt.

Interestingly though I've found a use for this...My faulty Nokia charger will without fail knock out ADSL on both sides of my house when my nice neigbours refuse to turn the volume down during all-night gaming marathons

Rgds
Havelock

[kitz]

 

Youre baaaaaaaad

[havelock]

To quote Jessica Rabbit:

"I'm not bad, I'm just drawn that way"

[roseway]

Nice one

[mr_chris]

600m of cable should still not give you an attenuation of 33dB. I'm about 3km from the exchange and my attenuation is 37dB.

Before BT removed the line length test from public consumption, Kitz's line was measured at about 800m (correct me if i'm wrong kitz), and her attenuation is 7dB!

So there's obviously something going on... getting BT to believe it is another matter though.

[kitz]

>> Kitz's line was measured at about 800m (correct me if i'm wrong kitz), and her attenuation is 7dB!

Yep your right, although Im about 120m straight line -  the line length is 783 meters with 7dB attenuation.

[notfes53]

Another call to India resulted in Line Faults calling me back and offering a further engineer visit. They are keeping the file open for at least another three days.

It appears that the latest genetleman at the BT call centre line faults department spoke to BT in the UK and they think I should be getting up to 6.5mbps, as opposed to between 3.0 and 3.5.

  - I'll believe that when I see it!!

My estimate of 600m does follow the route that the engineer detailed. Its effectively going almost all of the long way around a square with lengths of 160m on each side, when I live just over half way along a side that's attached to the exchange at a corner of the square, and the short route would have been 85m! (sorry that this explanation is so wordy) - it just adds to the frustration!

Even at 600m, what sort of attenuation range should we see if we've got relatively good copper all the way?

Is the max d/l synch rate determined partly by attenuation, and partly by signal to noise? Is it possible to construct a simple matrix showing attenuation bands from low to high on one axis and a rising scale of S/N values on the other axis, and predict synch rates in the boxes - even if it's only theoretical?

Finally is 'interferece' just going to affect S/N or can it affect another measurement too? or does that depend on how 'interference' is defined?

p.s. just a thought: I have a separate business line supplied by my employer, and I've requested to have Broadband on that line too, to save dialling in and then tieing up the voice line. If/when it is supplied it will be interesting to see whether the stats are indistinguishable or not. The line has a much more recently installed master socket and it's very own cable from the top of our local telegraph pole!

[mr_chris]

> Even at 600m, what sort of attenuation range should we see if we've got relatively good copper all the way?

Probably about 5 - 6 dB, and a lot of envious people.

Max download sync rate is determined solely by the SNR margin. Attenuation is line loss, and remains the same regardless of what speed you sync at. SNR margin is affected more by interference, quality of the copper, etc..

Because higher sync speeds use higher frequencies, they are more prone to interference, and SNR isn't as great. Max DSL works by setting a target SNR margin, e.g. 6dB, that means your line will sync at the highest rate it can achieve whilst keeping the SNR margin above 6dB.

So to try and answer your question, an attenuation vs SNR graph would be difficult, you'd have to define a variable as a constant i.e. sync speed.

What someone has tried to do is a graph of attenuation vs achievable sync speed.. I would imagine target SNR margin is defined at 6dB. Note it's very rough, and assumes good quality cable along the whole route, but you can see, even with 33dB attenuation you should be getting close enough to 8Mb, which is consistent with what I would have expected anyway.


Image from Wikipedia

[soms]

from experience RFI filters seem to be preventing DSL from working completely in most cases BTW

There are different versions of the RF filters. hardwired BT ones come in the form of a block 80 varient (screw terminals to IDC connectors) with the addition of a set of inductors to filter RFI on the A and B terminals.

The block 80 RF and RF2 filters do stop ADSL from working properly, however the more recent RF3 filters are fully ADSL compatible.

[havelock]

Cheers Soms...something I didn't know. I Should really go look at PSTN equipment myself. In most cases BTwholesale don't tell ISP's (or at least their helpdesks) when there are new introductions of PSTN hardware even if it is Broadband affecting...

Havelock

[notfes53]

Well, it's been a week now. No more contact with BT/engineers, but my speeds have slowly climbed:

Today's speed test and stats:
IP profile for your line is - 4000 kbps
DSL connection rate: 448 kbps(UP-STREAM)  4832 kbps(DOWN-STREAM)
Actual IP throughput achieved during the test was - 3684 kbps

I'm now syncing at 4832 = almost 4000 up on my low point a week ago

Uptime: 2 days, 6:26:09
 
Modulation: G.992.1 annex A
 
Bandwidth (Up/Down) [kbps/kbps]: 448 / 4,832
 
Data Transferred (Sent/Received) [GB/GB]: 8.20 / 10.01
 
Output Power (Up/Down) [dBm]: 11.5 / 15.5
 
Line Attenuation (Up/Down) [dB]: 7.0 / 22.0
 
SN Margin (Up/Down) [dB]: 16.0 / 7.0
 
Vendor ID (Local/Remote): TMMB / ALCB
 
Loss of Framing (Local/Remote): 0 / 0
 
Loss of Signal (Local/Remote): 3 / 0
 
Loss of Power (Local/Remote): 0 / 0
 
Loss of Link (Remote): 0
 
Error Seconds (Local/Remote): 422 / 0
 
FEC Errors (Up/Down): 0 / 510,514
 
CRC Errors (Up/Down): 0 / 0
 
HEC Errors (Up/Down): 0 / 0

However - the biggest change has to be the Attenuation. It has suddenly improved from 33dB to 22dB - without me making any changes in the house at all. Perhaps BTw or BT Openreach have done something to my line without my knowledge. And the S/N Margin is even showing a shade better on the above stats. It was always either 6.0 or 6.5 previously.

I don't know if there's any margin for further improvement. From the graph posted by Mr Chris it does seem as though I should be able to get even higher speeds - but noise may be the limiting factor.

I wonder what can have happened, and whether there is anyone that I can call to find out how they've made the improvements? I also wonder if any further 'complaint' will even get a response?

Thanks for all the support to date - Steve

[roseway]

That big reduction in attenuation means that either a line fault has been repaired or they've connected you to a different line pair which follows a shorter route. Whatever the reason, it's definitely good news!

The small change in noise margin when you took the reading doesn't mean anything - the target noise margin (6 dB by default) is the starting point when the router syncs with the exchange, and it will sync at the best speed it can consistent with that noise margin. After that the margin will drift up and down a bit as interference conditions change.

You probably won't get BT to admit that they've done anything - it's not unusual for connections to improve after a complaint, even though BT say that they haven't changed anything. I think I would just keep quiet and see how it goes.

However you should really be getting a faster sync speed for that attenuation. I haven't checked back over all the messages in the thread, but have you done all you can locally to improve your internal wiring and filtering arrangements?

Eric

[notfes53]

Hi Eric, et al

Spoke too soon!!

Attenuation gone back from 22dB to 31dB
IP profile dropped from 4000kbps to 1500kbps
Sync dropped from 4832kbps to 1856kbps
D/s throughput dropped from 3715kbps to 1392kbps

Back on the phone to BT - going through to Line Faults.

Will update asap - Steve

UPDATE:
Lovely lady in BTBB Line Faults has promised to get another engineer around to my house.
This morning Royal Mail delivered me a new CISCO 837 ADSL router for my business line, courtesy of my employer. Once I have that up and running and the line activated I cannot wait to see if there are any differences in the statistics!

Any experience of/views on that router as compared to the BT Home Hub??

[kitz]

>> Attenuation gone back from 22dB to 31dB


Wonder if this perhaps ties in with some heavy rain? - water/underground cables then it gets better as the weather clears? 

>>  new CISCO 837 ADSL

very nice   cant give you a personal opinion since I think they are normally out of most peoples price range.  From the comments I have seen, users seem to think they are the dogs danglies, but Ive heard they are a bit more complicated to configure.