Myreader.co.uk  
uk news, chat and community
   home   |   control panel login   |   archive   |  
 
misc
announce
answers
consultants
d-i-y
environment
environment.conservation
gov.agency.csa
gov.local
gov.social-security
gov.social-work
misc
philosophy.atheism
philosophy.humanism
philosophy.misc
radio.amateur
railway
sci.astronomy
sci.med.nursing
sci.med.pharmacy
sci.misc
sci.weather
singles
telecom
telecom.broadband
telecom.mobile
telecom.voip
test
transport
transport.air
transport.buses
transport.ferry
transport.london
transport.ride-sharing
  
 
date: Sun, 27 Jul 2008 19:57 +0100 (BST),    group: uk.telecom.broadband        back       
Re: SNR Target vs SNR Margin ?   
On Sun, 27 Jul 2008 11:39:34 +0100 nospam <nospam@please.invalid> wrote:

> *From:* nospam <nospam@please.invalid>
> *Date:* Sun, 27 Jul 2008 11:39:34 +0100
> 
> eddy@127.0.0.1 (Eddy) wrote:
> 
> >Can anyone explain the difference between the SNR Target and the SNR 
> margin?
> >I think these two terms get mixed up sometimes.
> 
> The higher the connect rate the more signals the line must carry. The 
> more
> signals there are the smaller they must be to 'fit' in the line. For 
> the
> same amount of noise the smaller the signals are the lower the ratio 
> of
> signal to noise.
> 
> >I always assumed
> >the higher SNR Margin was better as it gives a better sync rate, at 
> least whilst
> >the line is quiet.
> 
> The reported SNR is the signal to noise ratio at that moment and 
> gives no
> indication of line quality without also specifying the connection 
> rate. For
> the modem to see the signals they must be bigger than the noise and to
> accommodate some variation in noise a margin of 6db is typically 
> required.
> While establishing an ADSL connection the exchange tells your modem 
> the
> target SNR (6db usually) and it is up to the modem to connect at a 
> rate
> which gives that SNR. 
> 
> >How do you find out what the SNR Target is, rather than the SNR 
> Margin as reported
> >by the router?  I cannot see that anywhere on the 2600 router stats.
> 
> Immediately after establishing a connection the reported SNR should 
> be a
> good indication of the target. Thereafter the reported SNR will vary 
> as
> noise on the line varies. The exchange sets the target from 6db 
> upwards in
> steps of 3. 

Ok, thanks for that info. I think I now understand how the SN works, but how do you
get to know the Target SNR? Do some routers report it?
Mine (Draytek 2600) only seems to report the SNR Margin, not the target.

Eddy.
date: Sun, 27 Jul 2008 19:57 +0100 (BST)   author:   (Eddy)

Re: SNR Target vs SNR Margin ?   
> Ok, thanks for that info. I think I now understand how the SN works, but 
> how do you
> get to know the Target SNR? Do some routers report it?
> Mine (Draytek 2600) only seems to report the SNR Margin, not the target.

The target SNR is not reported by any router that I know. Your ISP should be 
able to tell you, or as others have said the SNR immediately after a 
reconnect is a good guide. Don't get too hung up on the figure because it 
means very little on its own. Likewise the sync speed does not tell the 
whole story. The only real criteria to judge how good or bad a connection is 
by measuring it. Unfortunately most means of doing this, that are available 
to users, have weaknesses. In particular a measurement over a minute or so 
only gives a rough snapshot of what happened at that time. You can see the 
errors reported by the Draytek. I have to say that I have always been a big 
fan of Draytek modems but recent experience has made me change my mind if 
the line is noisy. After a number of visits from BT one of their senior 
engineers suggested I try a BT Business Hub 2Wire modem. This is in fact a 
rebadged version of the 2Wire 2700HGV that is optimised for use on rural USA 
lines that are notoriously bad. They can be picked up on eBay for less that 
£30 and mine has improved the sync speed from 4000 to 4700. Initially the 
target SNR was set at 6dB but the error rate was collossal and so it was 
raised to 9dB which dropped the sync speed to 4100 but when the thoughput 
was measured over an hour it was significantly higher at the lower sync/SNR.

Peter Crosland
date: Mon, 28 Jul 2008 21:10:47 +0100   author:   Peter Crosland

Google
 
Web myreader.co.uk


    COPYRIGHT 2007, YARDI TECHNOLOGY LIMITED, ALL RIGHT RESERVE  |   contact us