Kitz Forum

Broadband Related => Broadband Technology => Topic started by: NewtronStar on August 12, 2015, 10:28:55 PM

Title: Minimum Target SNRM
Post by: NewtronStar on August 12, 2015, 10:28:55 PM
Doe's anyone think OR will reduce the Snrm target margin from 6.0dB to 3.0dB when G.INP and vectoring are working together as this should give users 30% to 40% increase in sync but still keep below there ISP profile for errored seconds per day.

And the cap of 80/20 would need to be removed as users already on 80/20 could get 140/40 as long as the DLM is soften up.
Title: Re: Minimum Target SNRM
Post by: Chrysalis on August 12, 2015, 10:48:52 PM
my view is take it down to 4db, but yeah I think it would be viable.  I say keep the cap tho.  Maybe raise upload to 25.

Also remember hg612's and the eci's have 100mbit ports.
Title: Re: Minimum Target SNRM
Post by: NewtronStar on August 12, 2015, 11:11:58 PM
my view is take it down to 4db, but yeah I think it would be viable.  I say keep the cap tho.  Maybe raise upload to 25.

Also remember hg612's and the eci's have 100mbit ports.

I'm glad you also think so it's my thinking why introduce G.INP and Vectoring to an old FTTC product if your reasoning was to make everyone's line stable surly they want to increase the users sync rate and the reduction of target margin is a good step forward.
Title: Re: Minimum Target SNRM
Post by: kitz on August 12, 2015, 11:22:58 PM
Im not sure about 4dB.   If they ever did decrease it, I would say its more likely to be 3dB.

The industry standards are min 6dB for adsl1,  3dB for adsl2+

Increments usually rise in 3dB stages as 3dB of SNR = 1 bit.
Title: Re: Minimum Target SNRM
Post by: NewtronStar on August 12, 2015, 11:39:27 PM
Sorry Kitz i've had a brain storm my crosstalk distuber lowers my SNRM by 4.0dB so with Vectoring my DS SNRM would be 11.0dB imagine what my snyc rate would be if the target margin was set at 3.0dB ok i may be hitting 200 errored seconds a day on G.INP but i'll still be Green  :)
Title: Re: Minimum Target SNRM
Post by: Chrysalis on August 12, 2015, 11:43:00 PM
I dont know about industry standards, in other countries 7db is more popular, sky also use 7db on their adsl llu.

Whilst bt are more likely to use 3db due to their history it is less logical, as 2 extra tones will take 7db to 1db still a positive number, whilst with 6db its down to 0db and will get errors.  For this reason if they using 6db they may as well use 4db as both can only tolerate one extra bit loading.

so logical best is 1,4,7,10,13 db loading.
not 0,3,6,9,12,15 as BT do.

bear in mind I said what I think they should do, I am aware BT wont use 4db.  I dont think they will even use 3db tho as they seem more paranoid about stability on FTTC because of iptv.
Title: Re: Minimum Target SNRM
Post by: GigabitEthernet on August 13, 2015, 12:20:23 AM
I believe Sky don't actually use 7dB anymore. I believe it goes as low as 3dB nowadays :)
Title: Re: Minimum Target SNRM
Post by: burakkucat on August 13, 2015, 05:55:30 PM
b*cat makes the observation that this thread was started about the topic of "G.INP on ECI cabs" and has gradually mutated into "something completely different" (to use a phrase from Monty Python's Flying Circus).  ::)
Title: Re: Minimum Target SNRM
Post by: Weaver on August 13, 2015, 06:34:03 PM
Split thread? Please?
Title: Re: Minimum Target SNRM
Post by: kitz on August 13, 2015, 11:33:08 PM
I dont know about industry standards, in other countries 7db is more popular, sky also use 7db on their adsl llu.

Skys adsl 2+ can and does do 3dB, 6dB etc.    I think 7dB used to be their manual setting. 
Title: Re: Minimum Target SNRM
Post by: Weaver on August 13, 2015, 11:38:41 PM
@kitz - apol. I don't understand the scaling used in SNRM here

from earlier, 3dB = 1 bit? In what sense, could you explain. [off-topic, i realise :-) ]
Title: Re: Minimum Target SNRM
Post by: kitz on August 14, 2015, 12:29:10 AM
See Bit loading (http://www.kitz.co.uk/adsl/adsl_technology.htm#bit_loading) :)
Title: Re: Minimum Target SNRM
Post by: Weaver on August 14, 2015, 01:16:15 AM
[apol for contd digression..

@kitz - thx Kitz, I was familiar with that section, I should have clarified. I couldnt see why 3dB =1 bit rather than 6dB. But I think the penny may have dropped. Twice the voltage means four times point as many points in the QAM constellation, its two dimensions that make the squaring that gives the missing factor of two in dB that I couldn't find. Does that make any sense? Good guess?]