Print

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

[NewtronStar]

The recovery phase in the US SNRM looks to short for a wet joint as evaporation would take quite a few hours in the wee small hours of the morning.

[GigabitEthernet]

I don't think those SNRM drops look right at all - I'm glad they aren't worrying you I just fear the line may see more of an effect in the future 

[roseway]

NS is right - it's very unlikely to be a water effect (and certainly isn't in my case because it's only a brief blip).

[GigabitEthernet]

Could it be lightning in your case, Roseway?

[roseway]

No, these occasional blips occur at random times, and there's no correlation with local thunderstorms.

[GigabitEthernet]

Crosstalk?

[Weaver]

@Alecr - rain could affect DSL performance without water ingress in a number of ways. If cables are lying in the water without any leaks into the cable, the dielectric constant will be altered because the electrical environment around the cable will be different. This would affect magnetic fields and electric fields surrounding would be altered.

Also, in heavy rain, while the rain is falling, could screen off radio emissions from distant disturbers which would improve DSL.

There are probably more mechanisms but I can't think of any at the moment.

[marjohn56]

@Alecr - rain could affect DSL perfoRmance performance without water ingress in a number of ways. If cables are lying in the water without any leaks into the cable, the dielectric constant will be altered because the electrical environment around the cable will be different. This would affect magnetic fields and electric fields surrounding would be altered.

Also, in heavy rain, while the rain is falling, could screen off radio emissions from distant disturbers which would improve DSL.

There are probably more mechanisms but I can't think of any at the moment.

My thought that the trunk from the Dist box to the pole has water that drains away quickly would fit with your point #1; that could only be proven by looking at everyone elses stats that are in that pole!

The DSL does not improve with heavy rain, at least not in my case otherwise I'd be praying for rain constantly.

An interesting aside on this is that I replaced my Draytek 2860 (Lanteq) with the D7000 (Broadcom) because of this phenomena. The 2860 would go completely bonkers and go off into a loop of connect/disconnect until the SNR recovered suffuciently. It would always recover, sync back normal speed and then you could watch the SNR fall away, and the cycle would reoeat. The D7000 holds on stoically though. I will get concerned as the winter draws in, but until the connection is lost completely or the speed falls below the guaranteed minimum I see no chance of getting anything done about it.

[GigabitEthernet]

It doesn't sound normal that a cable should by lying in water?

[marjohn56]

I agree, but because of the speed of the recovery, it's the only thing that sort of makes any sense. As was stated by another poster, if it was water sitting externally I cannot see it evaporating that quickly in the small hours. Whatever it is, I would really like it to stop.

I am not the only one seeing this sort of effect so I am probably miles wide of the mark. What I need is someone to drive into the pole and snap the cables; my cable would be replaced and we would know then if the problem was/is pole > house or pole > cabinet... any offers?

[renluop]

It doesn't sound normal that a cable should by lying in water?

But what of the like of my estate, built in the early 70s, when often the cables were laid directly in ground? My son's,, built mid 60s are within conduit.

[Weaver]

@AlecR - right outside my house the main cable is, or at least was, lying in the stream. Up on the moor, it used to be in a roadside ditch or is buried in wet peat.

[guest]

@AlecR - right outside my house the main cable is, or at least was, lying in the stream. Up on the moor, it used to be in a roadside ditch or is buried in wet peat.

They stopped that on Lewis decades ago because when the locals (council stopped doing it in 1986 IIRC) cleared the roadside ditches with JCBs then the cables broke more than if they were on poles and subject to winter winds.

For anyone curious, said "ditches" are often 1-2m deep and when you're looking at annual rainfall exceeding 3000mm/year & winds regularly in excess of 80 knots then you're between a rock & a hard place in terms of metallic connections.

[Black Sheep]

It doesn't sound normal that a cable should by lying in water?

But what of the like of my estate, built in the early 70s, when often the cables were laid directly in ground? My son's,, built mid 60s are within conduit.

Alec ...... lots of cables are led in water ........ how do you think trans-Atlantic cables are lain ?? There isn't a massive duct running under the pond.
Also, when the water-table is saturated our ducts (along with other services), will see a massive ingress of water. That is why E-side cables are pressurised and D-side cables are grease-filled. No, it doesn't render them totally water-proof, but it's the required specs. 

[krypton]

My line also reacts to rain. The SNRM decreases within seconds after the first raindrops (picture below). Sometimes the US SNRM drops from above 20 dB to zero followed by resyncs. Interestingly it could recover almost as fast as it dropped, even when the rain goes on. Maybe the "TX Power" gets increased?

I had already 4 guys from my provider here to find the cause but none of them found something. Finally they simply decreased the sync speed.

The last 150 metres are overhead ending in a waterproof junction box at the last pole, from there underground to the DSLAM. I don't see any possibility for water ingress. Is it possible to get such a massive impact only because of changing the dielectric constant? I can't believe this.