Getting a New Year's speed boost again!

No idea why (possibly reduced cross-talk from business lines) but every year, since I've been on a rate adaptive product, I have seen a sync speed boost in the week between Christmas and New Year.

This year is no different;
I normally get a maximum sync speed (if I re-sync during a low noise period) of around 4.8Mbps.
Since Christmas I've been sync'd at 5033kbps (not bad for a 58dB line on ADSL2+) and held sync.

Anybody else noticed this?

The opposite for my line, a drop from 15Mbps to 8Mbps on New Years Day.

Would have thought it to be very odd to get a speed boost every year, unless you get a speed drop in the summer.

Edited by deleted (Sun 02-Jan-11 20:59:30)

In reply to a post by 5km:

Would have thought it to be very odd to get a speed boost every year, unless you get a speed drop in the summer.

It's not a speed drop in the summer because I can be pretty sure my sync speed will be back to normal (4.5 to 4.8Mbps) within a week, and will stay there until Christmas day next year (when it, once again, goes up for just over a week).
It's done this every year since ADSLmax was introduced. It can only really be caused by reduced cross-talk on the incoming phone lines.

Oh I see. That is very strange. You would never work out what is causing that! Maybe someone goes on holiday for a week at this time every year and switches off an interfering device/modem.

On long lines it is not uncommon to see an increase in connection speed during the colder winter months followed by a drop as it gets warmer again. That is what I was thinking you might have been experiencing.

Edited by deleted (Sun 02-Jan-11 21:23:39)

I don't know about new year speed boost, but the ThinkBroadband Speedchecker is still rubbish with my ISP.


http://www.speedtest.net/result/1094424063.png

Edited by deleted (Sun 02-Jan-11 21:30:33)

That's coz you have a rubbish ISP, nothing to do with TB.

Select port 80 and you might get a better (higher) result but will be representative of web/http traffic not other types.

It can only really be caused by reduced cross-talk on the incoming phone lines.

Interesting idea, have a lot of people in your area gone away for X'Mas and New Year then?


Harry

In reply to a post by 5km:

That's coz you have a rubbish ISP, nothing to do with TB.

Select port 80 and you might get a better (higher) result but will be representative of web/http traffic not other types.

Firstly, Port 80 makes no difference to the result, and secondly, you seem the usual forum oik that dismisses everything out of hand, and who would not recognise a good ISP if you fell [censored]-over-head over one.

Lol

In reply to a post by Anonymous:

It can only really be caused by reduced cross-talk on the incoming phone lines.

Interesting idea, have a lot of people in your area gone away for X'Mas and New Year then?

Not the immediate neighbours but if it's reduced cross-talk it's more likely to be caused by those closer to the exchange (and the phone lines around here are likely to go past an industrial estate plus two large housing estates, before they get here), so it could be cross-talk, or a lack of it.

I normally get around 0.5-1mbit boost accompanied by noise bursts stopping for a week.

It's Mr Hacker/Professor, et al is it not?

Not the immediate neighbours but if it's reduced cross-talk it's more likely to be caused by those closer to the exchange (and the phone lines around here are likely to go past an industrial estate plus two large housing estates, before they get here), so it could be cross-talk, or a lack of it.

I used to get the same effect whenever a small business working from home four houses away, in fact on a different street and pole, but fed from the same cabinet, packed up for holidays. He had two or three staff at his house every day so I assume they had permanent internet connection in the working day.

In reply to a post by 5km:

That's coz you have a rubbish ISP, nothing to do with TB.

Select port 80 and you might get a better (higher) result but will be representative of web/http traffic not other types.

I used to think that until seb corrected me. Selecting port 80 on the tbb speed test does not invoke http protocol.

Ah! Thanks for the correction. I had always though it was there to see if there is a difference between web (http, port 80) traffic and other protocols/ports speed on ISP's with traffic shaping.

In reply to a post by 5km:

Ah! Thanks for the correction. I had always though it was there to see if there is a difference between web (http, port 80) traffic and other protocols/ports speed on ISP's with traffic shaping.

So did I. From what seb said it looks as though it just by-passes any possible throttle of 8095, but uses the same protocol as through that one, whatever that is.

'On long lines it is not uncommon to see an increase in connection speed during the colder winter months followed by a drop as it gets warmer again. That is what I was thinking you might have been experiencing.'

The copper expanding and contracting.....

hate to join in with the hate, but im a random that sometimes checks on forum topics. and m0aur your isp is in general '[censored]'

dont take offence and the service might be ok for u, but overall and in general is pants lol

Post deleted by billford

Deliberately evading the language filter is never a good idea.

In reply to a post by billford:

Deliberately evading the language filter is never a good idea.

Well it gives the little 'Tin Hitlers' a chance to throw their weight around.

In reply to a post by m0aur:

Well it gives the little 'Tin Hitlers' a chance to throw their weight around.

And an opportunity for those who feel they have the right to break agreements they entered in to.

In reply to a post by Zarjaz:

'On long lines it is not uncommon to see an increase in connection speed during the colder winter months followed by a drop as it gets warmer again. That is what I was thinking you might have been experiencing.'

The copper expanding and contracting.....

Hmmmm. So that shortens it, but as electrical current flows along the circumference of cable the resistance increases when cold. Discuss.

getting a constant speed day and night over christmas, no real change in sync.

In reply to a post by krazykizza:

getting a constant speed day and night over christmas, no real change in sync.

You have to resync to see a change. Maybe SNR has risen?

In reply to a post by billford:

In reply to a post by m0aur:

Well it gives the little 'Tin Hitlers' a chance to throw their weight around.

And an opportunity for those who feel they have the right to break agreements they entered in to.

Well Billy boy, I simply will not be talked down to or abused by ANYBODY, agreement or no agreement. I initially treat others as I would wish to be treated, from then on as they deserve. The problem with forums, is that they are predominately frequented with snotty-nosed know-it-all anoraks, each convinced that their view is the only correct one.

In reply to a post by m0aur:

The problem with forums, is that they are predominately frequented with snotty-nosed know-it-all anoraks, each convinced that their view is the only correct one.

True.

In reply to a post by RobertoS:

In reply to a post by 5km:

That's coz you have a rubbish ISP, nothing to do with TB.

Select port 80 and you might get a better (higher) result but will be representative of web/http traffic not other types.

I used to think that until seb corrected me. Selecting port 80 on the tbb speed test does not invoke http protocol.

Actually it isn't necessarily as simple as that even.

It all depends on how shaping/throttling/prioritisation is being done. If they do it by protocol then selecting port 80 will make no difference. If they do it by port then it could make a big difference (irrespective of the protocol actually used).

It depends on what equipment is in use and how it is configured.

In reply to a post by RobertoS:

In reply to a post by Zarjaz:

'On long lines it is not uncommon to see an increase in connection speed during the colder winter months followed by a drop as it gets warmer again. That is what I was thinking you might have been experiencing.'

The copper expanding and contracting.....

Hmmmm. So that shortens it, but as electrical current flows along the circumference of cable the resistance increases when cold. Discuss.

I always thought that it was just RF currents that flow on the surface hence copper wire is silver plated to improve conductivity at high frequencies.

In the case under discussion, the diameter of the copper will reduce when cold along with the circumference hence the signal will not have so far to travel and get to its destination faster

It's also plain wrong as resistance increases with temperature rather than decreasing with most materials (including copper).

http://en.wikipedia.org/wiki/Temperature_coefficient

A negative temperature coefficient (NTC) occurs when the thermal conductivity of a material rises with increasing temperature, typically in a defined temperature range. For most materials, the thermal conductivity will decrease with increasing temperature.

In reply to a post by ian72:

In reply to a post by RobertoS:

I used to think that until seb corrected me. Selecting port 80 on the tbb speed test does not invoke http protocol.

Actually it isn't necessarily as simple as that even.

It all depends on how shaping/throttling/prioritisation is being done. If they do it by protocol then selecting port 80 will make no difference. If they do it by port then it could make a big difference (irrespective of the protocol actually used).

It depends on what equipment is in use and how it is configured.

[cough]
Isn't the important bit of that what I say here?

Can't argue with that! btw: When I referred to RF currents I had in mind frequencies in excess of 30 MHz.

However, for whatever reason, my BB connection speed has been higher and more consistent during this cold spell.

In reply to a post by RobertoS:

In reply to a post by ian72:

In reply to a post by RobertoS:

I used to think that until seb corrected me. Selecting port 80 on the tbb speed test does not invoke http protocol.

Actually it isn't necessarily as simple as that even.

It all depends on how shaping/throttling/prioritisation is being done. If they do it by protocol then selecting port 80 will make no difference. If they do it by port then it could make a big difference (irrespective of the protocol actually used).

It depends on what equipment is in use and how it is configured.

[cough]
Isn't the important bit of that what I say here?

Hmm, sort of. Not as clearly as I stated though and you didn't explain how a difference between protocol and port may make a difference. But, it's a new year so let's not argue

noise bursts on schedule yesterday for first full working day after new year. Thankfully they didnt seem too heavy and my router handled them ok. (as no SRA again).

I've resyned, lots of times

And yes, SNR is on the move around lately, don't know why, but things are happening at my exchange so I'm not fussed as it won't be for long.

Reliability for me does tend to be better in the week between Christmas and the New Year

Until the DNS server crashed on the BeBox, I'd managed 40+ days uptime which is a lifetime record for me and syncs over 8500kbps without the usual re-syncs I get on a daily basis