Technical Discussion
  >> Technical Issues


Register (or login) on our website and you will not see this ad.


These posts have been archived and can no longer be replied to or modified.
Pages in this thread: 1 | 2 | 3 | [4] | 5 | (show all)   Print Thread
Standard User pitnicker
(knowledge is power) Mon 12-Feb-07 07:40:14
Print Post

Re: Surpisingly bad line stats....


[re: deleted] [link to this post]
 
In reply to:

{cahaddras}Attenuation is only relevant indirectly - in its contribution to the SNR. It's the SNR that determines line speed and reliability. Attenuation is very easy to compensate for - you just amplify the signal again. ADSL is all about differentiating signal from noise.
{ETEE}That doesn't really make sense. If you amplify the signal you also amplify the noise. My point is that attenuation is a limiting factor in that it defines the highest speed attainable on a line.



If you amplify, SNR remains constant [apart from noise from the amplifier]. So with a good SNR, but high attenuation, amplification gives you the same SNR and less attenuation

In reply to:

{ETEE}you are misrepresenting what I said which is really the same as you are saying in that noise on a line determines the actual speed.



So you are saying attenuation determines highest attainable speed, but SNR determines actual speed? Pls explain

=======================================
Everything below the line above is my sig
[Notice: Private mail is turned on, but anything received by these means can and will be published]
Standard User deleted
(deleted) Mon 12-Feb-07 08:40:13
Print Post

Re: Surpisingly bad line stats....


[re: deleted] [link to this post]
 
"But the attenuation of line 2 means that it would never achieve over about 5Meg anyway. "

Line 1 has 4dB higher attenuation, and has already achieved it !

Standard User deleted
(deleted) Mon 12-Feb-07 08:49:58
Print Post

Re: Surpisingly bad line stats....


[re: deleted] [link to this post]
 
{ETEE} "My point is that attenuation is a limiting factor in that it defines the highest speed attainable on a line."

That's NOT correct ! If that were the case, then BT could give us all very high speed broadband, just by whacking in huge amplifiers into every line socket. It's a very complex subject. It's all about the ratio between the signal and the noise - if the wanted signal is only 2dB above the noise level, amplifying it will do nothing - the signal will still only be 2dB above the noise.

Higher noise levels are synonymous with long lines because the longer the line, the more places the line can pick up noise. High attenuation is also synonymous with long lines, but shorter lines can have higher attenuation for other reasons - thinner copper, aluminium cables, high number of crimps, etc. In these cases, you can have high attenuation with low noise, and in many cases these lines can perform much better.


Register (or login) on our website and you will not see this ad.

Standard User deleted
(deleted) Mon 12-Feb-07 14:19:56
Print Post

Re: Surpisingly bad line stats....


[re: deleted] [link to this post]
 
In reply to:

Higher noise levels are synonymous with long lines because the longer the line, the more places the line can pick up noise.



I believe the maximum output power is just over 19 dB and my downstream output power (obtained from my router stats) never rises above 18 dB on my 54 dB attenuation line.

Is the reason for this because if the DSLAM tried to boost the broadband signal, nothing would be gained because the noise would also increase making no real difference?
Standard User deleted
(deleted) Mon 12-Feb-07 14:53:53
Print Post

Re: Surpisingly bad line stats....


[re: deleted] [link to this post]
 
BT could boost the power of all ADSL signals (it would have to be both ways), but the net result of that would be higher crosstalk from adjacent pairs, so higher noise.
Standard User deleted
(deleted) Mon 12-Feb-07 14:57:11
Print Post

Re: Surpisingly bad line stats....


[re: deleted] [link to this post]
 
Thanks zebedeee
Standard User deleted
(deleted) Mon 12-Feb-07 22:48:45
Print Post

Re: Surpisingly bad line stats....


[re: pitnicker] [link to this post]
 
In reply to:

So you are saying attenuation determines highest attainable speed, but SNR determines actual speed? Pls explain


Yes indeed. With say 32dB of attenuation on downstream using ADSL (not ADSL2+) the maximum attainable sync rate would be about 8128kbps on an average line. Add in extra noise than usual due to say crosstalk in the exchange and the sync rate drops.

With 45dB attenuation the figures are lower. ~5Megs would be the highest normal attainable sync rate on ADSL. Lower sync rates are achievable. Note that I am not talking about a specific line but lines in general.

Is that sufficient explanation?
Standard User deleted
(deleted) Mon 12-Feb-07 22:52:33
Print Post

Re: Surpisingly bad line stats....


[re: deleted] [link to this post]
 
Alright then, we are on different wavelengths. Consider what would happen to your line 2 if the noise on it was suddenly reduced miraculously by 12dB but the attenuation remained the same. What sync rate would you expect on ADSL (not ADSL2+)? Not 8Meg for sure!
Standard User deleted
(deleted) Tue 13-Feb-07 09:06:44
Print Post

Re: Surpisingly bad line stats....


[re: deleted] [link to this post]
 
I don't think you can talk about "lines in general", because no two lines are identical, especially when you include the maze of extensions that many have in their houses.
Standard User deleted
(deleted) Tue 13-Feb-07 09:13:26
Print Post

Re: Surpisingly bad line stats....


[re: deleted] [link to this post]
 
You are only looking at half the equation. Attenuation reduces the amount of signal that arrives at the other end of the cable. A cable can have an attenuation of 40dB, but if it picks up a lot of noise on the way, you could end up with a sync speed of 2M or lower. Similarly, you can have a line with a 55dB attenuation with a low noise pickup, which can sync at 6M or even more. I've seen people report this sort of performance every day in these forums. SNR is a RATIO of wanted signal against unwanted signal, so the SNR is a bigger problem because noise pickup in a cable can be variable, wheras the attenuation of a cable stays quite stable if the cable has no defects.

Pages in this thread: 1 | 2 | 3 | [4] | 5 | (show all)   Print Thread

Jump to