General Discussion
  >> Fibre Broadband


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


Pages in this thread: 1 | 2 | 3 | (show all)   Print Thread
Standard User Nightglow
(regular) Mon 18-Jan-16 08:34:16
Print Post

G Fast Routers/Modems


[link to this post]
 
A brief article over at ISPreview about four new router/ modem for G Fast being tested.

http://www.ispreview.co.uk/index.php/2016/01/a-quick...

Edited by Nightglow (Mon 18-Jan-16 08:36:00)

Standard User RobertoS
(elder) Mon 18-Jan-16 10:24:55
Print Post

Re: G Fast Routers/Modems


[re: Nightglow] [link to this post]
 
Interesting stuff. Thanks smile.

Some modem/routers that supply power to the node!!! That'll cause some heated discussion!

The indispensable man or woman passes from the scene, and what happens next is more or less the same thing as was happening before.
My broadband basic info/help site - www.robertos.me.uk. Domains, site and mail hosting - Tsohost.
Connection - AAISP Home::1 80/20. Sync 59997/15142kbps @ 600m. - BQM
Standard User Nightglow
(regular) Mon 18-Jan-16 10:42:42
Print Post

Re: G Fast Routers/Modems


[re: RobertoS] [link to this post]
 
In reply to a post by RobertoS:
Interesting stuff. Thanks smile.

Some modem/routers that supply power to the node!!! That'll cause some heated discussion!


Ofcom wouldn't allow it, or would they?

I would expect a good discount off my bill for using my electric to power a BT node.

Edited by Nightglow (Mon 18-Jan-16 10:43:50)


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

Standard User TheEulerID
(committed) Mon 18-Jan-16 11:01:58
Print Post

Re: G Fast Routers/Modems


[re: Nightglow] [link to this post]
 
It would be up to the NICC which controls technical standards on the copper network. In any event. why would Ofcom not allow it? As far as the electricity bill goes, firstly this is surely an issue of consideration when looking at the entire retail package cost (in principle, people should look at the power consumption of ISP router/modems - I doubt many do).

Also some sense of perspective is required here. I did a lot of calculations on this and looked at how much power might reasonably delivered on standard copper pairs over distances of up to about 250m using 50v. A reasonable estimate was that the absolute maximum that could be relied on might be around 15W delivered to the node by any one line drawing around 20W (you could get considerably more on shorter lines, but that can't be relied upon). However, with proper power balancing over multiple lines, a reasonable average would be a draw of perhaps 5W (delivering around 4W making some assumptions about average line lengths). A consumption of 5W is about 50p per month at current UK retail electricity prices (even 10W is only about £2).

The simplest way of balancing power requirements is to draw equal current from all lines (which doesn't deliver equal power, but would demand equal power from all the users connected to a line).

This requires the designers to work to an incremental power budget of, about 2W per line with a baseline load of around 15W (the latter is tough). G.fast has very fast training and low power modes when lines are idle, so consumption over time could be lower. The biggest problem with reverse powe (RP)r will be to have sufficient uptake to reach a critical level to provide the baseline power load after that, the more users, the lower the average demand. I'd anticipate that nodes would contain a modest size battery which, in the manner of a UPS, would be trickle charged and would be able to provide sufficient power to keep the baseline load going for perhaps an hour or so.

As far as other costs go, if the alternative is to spend £10k+ on providing mains power (and it can be a lot more depending on location), then that would add much more to the wholesale price than delivering a few watts using RP.
Administrator MrSaffron
(staff) Mon 18-Jan-16 11:08:05
Print Post

Re: G Fast Routers/Modems


[re: TheEulerID] [link to this post]
 
Mains power - they are in the trials in some areas powering from a power cab next to existing VDSL2 cab and that feed could be up to 1km long over the copper. Longer if more pairs are available, and guess what with SOGEA coming they may be more spare pairs for power supply use.

Reverse power is a no brainer in the FTTB deployments so countries are doing i.e. 1 node per apartment

The author of the above post is a thinkbroadband staff member. It may not constitute an official statement on behalf of thinkbroadband.
Standard User WWWombat
(knowledge is power) Mon 18-Jan-16 11:20:52
Print Post

Re: G Fast Routers/Modems


[re: TheEulerID] [link to this post]
 
I agree - somewhere like 40-50p per month.

Demo of power switching here:
FTTdp Reverse Power Sharing Module RPS8 Demo at B: http://youtu.be/RICL26vLEBk

Last I saw was that the developers were working at making sure the DPU could turn power off per-port, ensuring it ran at the lowest possible power for just a single user. Quite a tricky target.

However, the impression I got from listening to BT demonstrate their g.fast equipment is that reverse power is not even close to their thinking in deployment. It probably isn't worth thinking about for the UK for a good few years.
Standard User WWWombat
(knowledge is power) Mon 18-Jan-16 11:25:02
Print Post

Re: G Fast Routers/Modems


[re: MrSaffron] [link to this post]
 
Yup - forward power over spare copper looks much more likely than reverse power right now.

SOGEA offers an interesting option for "spare" pairs, but I thought that kept the copper bearer back as far as the MDF, even if it wasn't connected to anything else.
Standard User TheEulerID
(committed) Mon 18-Jan-16 11:34:42
Print Post

Re: G Fast Routers/Modems


[re: WWWombat] [link to this post]
 
Pretty well what I was expecting, and I don't expect to see it happen any time soon in the UK. It's elegant in principle, but getting it all to work in harmony and keeping the power budget under control is gong to be tricky (but then so is xDSL).

What I do find amusing is those people who haven't got the foggiest notion how much power might be demanded and how much it would cost, but are determinedly against the whole principle. A few minutes work using a spreadsheet, the resistive properties of copper loops and PSTN DC voltage levels soon puts this in perspective.
Standard User TheEulerID
(committed) Mon 18-Jan-16 11:58:27
Print Post

Re: G Fast Routers/Modems


[re: WWWombat] [link to this post]
 
Easier to organise than reverse power, but it's very distance limited and at 1km wastes a lot of power. A single 1km loop of 0.5mm diameter wire can only deliver about 4W to the load and to do so consumes 8W at the source (half gets wasted in resistance losses). Of course thicker gauge wire will allow more power to be delivered at the same distance (in fact, disproportionately more when compared to cross-sectional area) or the same power to be delivered to greater distances in direct proportion to cross-sectional area).

The RP mechanism is, in principle, much more efficient as it wastes a lot less power in transmission and automatically scales with higher speed standards as, necessarily, that means shorter loop lengths.

So, pragmatically, I can see this happening, but "forward power" has it's own issues but could be done without complex issues regarding customer equipment.

I'm assuming a nominal 50V DC. Higher voltage would, of course, deliver more, but I can't image anybody countenancing that.

nb. if only spare pairs were to be used, then this would surely be a big problem as the number of "spare" loops required over 1km would be roughly similar to the number of serviced lines. Of course it's perfectly possible to use existing pairs in principle (at just means supplying the 50V to the users phone line from the local powered cabinet), but I suspect there would be horrible network rearrangement jobs to do it and, possibly, other diagnostic issues.
Standard User godsell4
(member) Tue 19-Jan-16 17:37:13
Print Post

Re: G Fast Routers/Modems


[re: WWWombat] [link to this post]
 
In reply to a post by WWWombat:
Yup - forward power over spare copper looks much more likely than reverse power right now.


That is a big assumption, in my location there are no spare pairs due to neglect by BTO, the default behaviour of any attempt at fault repair we see in this location is to 'just try a different pair'. Well there are no spare pairs here now for any new lines at all.

This behaviour of BTO is in the kind of locations where G.Fast would make sense, so if the power has to go over spare pairs, that is an issue and BTO maybe well aware of this and so must be looking at the power problem in a different way including use any/all wire pairs in the cable used or not used.

PlusNet Unlimited Fibre 3Mb to 5Mb
Pages in this thread: 1 | 2 | 3 | (show all)   Print Thread

Jump to