Electric current flows on the outer layer of copper. So contraction due to low temperaure may shorten it, but the reduction in circumference could increase the resistance. ie increase the attenuation.

But the conductivity of copper changes as te temperature drops ... conductivity improves which is effectively the same as the resistance dropping.

In reply to a post by MHC:

But the conductivity of copper changes as te temperature drops ... conductivity improves which is effectively the same as the resistance dropping.

Hehe. "Super". .

My currently poor connection has also improved (slightly) as the weather has become colder.

I hope that doesn't just mean that it will be completely useless next summer (rather like it was for most of last summer), if you could actually call it summer.

In reply to a post by RobertoS:

Electric current flows on the outer layer of copper. So contraction due to low temperaure may shorten it, but the reduction in circumference could increase the resistance. ie increase the attenuation.

Well i guess i was just one lucky chicken then cos my attenuation was always a good 3db or 4db to the better in the winter depending on how cold it got. Summer attenuation was low 50s, winter attenuation high 40s

I think it unlikely such an effect is anything to do with the temperature of the copper. If it were it would be a well-known phenomen, but in general peoples' attenuations never vary by more than 1dB unless there is a problem somewhere, or they change router - that can affect the reported value by several dB.

In reply to a post by RobertoS:

I think it unlikely such an effect is anything to do with the temperature of the copper. If it were it would be a well-known phenomen, but in general peoples' attenuations never vary by more than 1dB unless there is a problem somewhere, or they change router - that can affect the reported value by several dB.

Well it happend every winter for 8 years, i am not what you would call a noob and i know full well about chaning routers etc. I made sure when a router change came it was to one with a broadcom chipset, these always without fail gave me a lower attenuation and around 700kbps extra sync. Which on a 3.5km+ line length is significant.

Sorry, I could have said more.

What I meant was there are many joints, chambers or whatever in the line between you and the exchange. Temperature could very well be affecting something dodgy in that chain. May even be some aluminium, and I have no idea of the temperature effects on that.

I just don't think that the effect on the copper itself is likely to be significant. I don't dispute you have something bad going on and that temperature seems to play a part. Or at least something to do with the ambient climate.