Slower sync; possible noise?

Morning all,

After having my FTTC connection installed last week; I'm not gaining full sync speed when my line is of quite a short length.

Stats are:
Maximum data rate:
20000 / 60787
Noise margin:
6.9 / 6.3
Line attenuation:
15.5

Where is live is all 70's underground cabling (no overhead here); and the master socket has been moved internally over about 10m of Cat5e avoiding power cables. The router is connected directly to the master socket using a Mk3 Openreach filter.

Am I suffering due to noise of other connections or is something else going on? I have a short 1M RJ45 to RJ11 on order to try and improve the situation.

Cheers,

Dave

What kind of cat5e have you used? Most domestic cat5e is stranded. I've no idea why people use cat5e for such jobs. I like to use the same cable all the way. Should really be using at least 0.5mm solid core copper for carrying the dsl signal, such as cw1308.
By all means use stranded cat5e for extensions, or anything after the modem has received the xDSL signal. Personal preference I guess.

Do you know the distance to the PCP? Most lines achieve better with an attenuation as low as yours, but not everyones can or will.

Edited by j0hn83 (Mon 09-Oct-17 08:32:51)

I have used copper solid core Cat5e external grade (as it travels outside for 5m or so); you can't terminate stranded into an IDC correctly. Looking at the specs it has 24AWG conductors (0.51mm diameter). I have used it because I have been told it's the best cable to use by various sources; and BT even use the Cat5e spec in their 2 pair CW1724.

Distance is 115m in a straight line; or 240m following the road.

In reply to a post by dsgoody:

Distance is 115m in a straight line; or 240m following the road.

Your attenuation of 15.5 suggests it is significantly longer than that and potentially around 400m although that will vary slightly depending on the actual gauge of copper.

The SNR figures being 6.x do suggest that the line is running at or around the best potential speed and that is being dictated by noise levels from all sources including cross talk.

A 16dB line could normally have a sync any where from low 60s to 80 Mbps and yours is probably one of those at the lower end.

Which exchange are you on?

What modem are you using?

Your attenuation of 15.5 suggests it is significantly longer than that and potentially around 400m although that will vary slightly depending on the actual gauge of copper.

Agreed ..... and with mention of it being a 70's house in the OP's post, what's the betting some or all of the D side is ali ?

I've possibly only come across one or two lines with aluminium and never checked the stats.

As it is possible to get both Line Attenuation and Signal Attenuation, visible in DSLstats, I wonder if there is a way to identify Ali being present?

In reply to a post by MHC:

I've possibly only come across one or two lines with aluminium and never checked the stats.

As it is possible to get both Line Attenuation and Signal Attenuation, visible in DSLstats, I wonder if there is a way to identify Ali being present?

Why, do you think there's some sort of difference on aluminium lines?

In reply to a post by BatBoy:

Why, do you think there's some sort of difference on aluminium lines?

There is, but how much and whether it would be obvious, I'm not sure of.

At higher frequencies the conductivity of some metals is reduced, one reason is the skin effect and that they form oxide layers which do not conduct. Silver for example is great at DC and low frequencies but as the frequency increases the advantages are lost. Aluminium does oxidise on the outer skin with potentially a similar effect.

So if the attenuation per tone group - D0, D1, D2, D3 or per tone could be plotted, would the gradient of Aluminium compared to copper be noticeably different?

In reply to a post by MHC:

So if the attenuation per tone group - D0, D1, D2, D3 or per tone could be plotted, would the gradient of Aluminium compared to copper be noticeably different?

Available on the pbParams but maybe not on MDWS. We just need a user who knows they're on aluminium, and maybe runs DSLstats

Thanks for all the input.

The exchange is Chipping Sodbury; Cabinet 13 (ECI). The modem/router is a BT Smart Hub. I have had similar results with a BT HH5A.

As for line length; it's possible it could go all around the houses or could be Ali. I know nothing about old Ali lines but I have heard of them; where they popular in the 70's? I think from memory my house was built 1975.

Popular because the material was cheaper, so yes. At the time, there was no such requirement as broadband etc and it provided a phone service which functioned adequately.

Ah, on an ECI cab so no G.INP or Xdb at the moment to help.

What about local noise, do you have any phone extensions? Is the master socket the first connection on your line? Is it an NTE5?

No phone extensions; the master socket is the only connection (although itself on the end on 10m of Cat5e routed away from power cables as in my original post). It's an NTE5A with an Openreach Mk3 Filter.

The only question mark is the quality of the Openreach connection coming from the drop cable (or in this case; the underground cable). The original pairs are cut very short; and their are some very old jelly crimps connecting to two jumper cables before some new jelly crimps onto the Cat5e. Probably not a problem; but the only internal fault I can pick at.

I will see if I can get a photo.

In reply to a post by dsgoody:

No phone extensions; the master socket is the only connection (although itself on the end on 10m of Cat5e routed away from power cables as in my original post). It's an NTE5A with an Openreach Mk3 Filter.

Thanks. Who did the wiring?

The only question mark is the quality of the Openreach connection coming from the drop cable (or in this case; the underground cable). The original pairs are cut very short; and their are some very old jelly crimps connecting to two jumper cables before some new jelly crimps onto the Cat5e. Probably not a problem; but the only internal fault I can pick at.

I will see if I can get a photo.

A photo of this would be really good

I run the cable; OR terminated it when the line was re-activated.

https://s5.postimg.org/4tpdc95rb/20171009_225803.jpg

That really doesn�t look especially healthy (the crimps on the armoured UG feed end) BUT you cannot always judge by appearances with such things.

It would require a new duct and a lead in to replace, and that would be expensive ... and usually only organised to resolve a �hard� fault eg earth/battery/dis

I agree; I've taken the approach that if it's working to leave it alone.

The *only* fix I can think of is to chisel out underneath the back box and move it down the wall slightly; but I have decided against this mainly due to the risk of causing damage to the OR cable and neighbours property with it being such a deep back box (it's on the boundary wall).

I do have access to jelly crimps if I decide to be naughty (ie. working with the OR cable) in the future......

I suppose the question is; if these crimps have degraded would they cause an increase in noise or attenuation?

If noise; then we can assume all is ok.

Doesn�t look good. How long is the cable to the master?
Who wired in the master and who fixed the master to the wall?
Is it solid or loose?

Edited by deleted (Tue 10-Oct-17 11:03:09)

Openreach; and it's solid.

Are we talking about the cable between the router and master? If so it's about 3 to 5m of ISP supplied goodness. Far too long and is being replaced with a 1m cable on order.

it's ok. I'm just re-reading the info you've already posted and I think you've answered all my questions already.

Is that rust in the back box? If so, I would be more concerned about damp in the wall than the connection.

Don�t be touching them crimps, there�s not enough wire inside to play with.

To be honest, I think the wiring is fine and you're just being affected by aluminium cables as already stated.

The best way forward in my opinion it to use a modem that supports DSLstats and get some graphs on MyDslWebStats for analysis.

Reasonably common, damp tracking back up the exterior of the direct in ground cable. Not a good way to install cables, and that�s why they stopped it.

Update; my 1m RJ45 to RJ11 lead has arrived. It's made with BT CW1724 cable and has improved attenuation and max sync speed slightly.

Maximum data rate:
24817 / 62028
Noise margin:
8.9 / 6.2
Line attenuation:
15.2

If I ever get a separate modem in the future I will be sure to get the stats checked out for Ali.