Speed of a fibre strand?

Hello All

Just wondering what the actual speed of an individual strand of fibre was.

This is for our FTTP install, where we are now just waiting for a fibre joiner to join our other end to where the faulty cable connected.

Basically BTOR has just put in a thin fibre cable (at a glance it seems the thickness of RG213 Coax) with about 96 or so strands instead of the 40 odd due to not having any, so I am aware depending on what that connects to would determine how many it will use leaving the rest for spare in case of issues.

But say that it only uses like 48 strands for all the homes on our phone cabinet which is about 394 (including phone boxes, bank tills and TFL) so say 350, that means 8 homes could/will be sharing the same strand, I did read some place that a fibre stand can handle several Gbits to Tbits or more depending where you read, so I was wondering what the average max speed would be per strand.

Thanks

Paul

All depends on what electronics you use to light the strand

A single strand can carry 1.4Tbps if you use DWM and the latest kit

In day to day world usually 1 Gbps or 10 Gbps.

If talking Openreach GPON then you are looking at GPON standard (which without checking I believe has 2.5 Gbps to share between users on the strand, that is then passively split, between 32 or 64 users or 128 users.

So probably use just 5 to 8 strands to serve your area, exact figure depends on the geographic layout.

Well I know they said we have 40 odd stands for our area (i.e phone lines on our cabinet) which due to not having that cable they used a 90 odd strand one.

But if say we go by what you said, in the worst cases if everyone hammers their connection we could see speeds as slow as 56Mbits or so from our soon 330Mbit connection.

I was in the thinking it would be a few strands per pole resulting in about 8 lines per strand.

Lets hope not everyone orders fibre LOL.

Anyhow I guess I will find out when it goes live, TBH anything would be better than what we are on now

Paul

I believe 10gbps XG_PON has already been trialled in Cornwall, and I'm sure it won't stop there.

As far as back haul from future g.fast nodes is concerned, I saw an online OR presentation that was talking of many WDM channels down the same fibre giving the potential for several 10s, if not a 100 or more Gbps.

For all the talk of gigabit broadband, they will all have over-subscription somewhere in the connection back to the peering points. The secret is to have a enough users over a fat enough pipe that the load evens out. I seem to recall some of the early contended ADSL services were horrendous like that as just a couple of maxed-out users would wreck the performance for everybody on a contended backhaul link.

In Japan someone is selling 2 Gbps across standard GPON.

Well I just come across this page, granted its probably not correct.
https://en.wikipedia.org/wiki/Fiber-optic_communicat...

And I read that some company has a fibre cable that is 99.97% the speed of light and can support way higher than 10s of Gbits.
http://rt.com/news/200151-internet-speed-fiberoptic-...

But I guess BT are assuming that not everyone will want fibre and if they do they may only require the lower speeds.

Like I said, I won't really know until its live.

Paul

In reply to a post by MrSaffron:

In Japan someone is selling 2 Gbps across standard GPON.

Ooooh Nice

Paul

The reality is that broadband NO MATTER HOW ITS DELIVERED is contended in the consumer area, that is why its a lot lower cost per month than a leased line.

Take-up of 200 Mbps and 300 Mbps Infinity 3 and 4 where available is pretty low, I can find speed tests as I know which areas have it available, but the majority still buy Infinity 1 and 2. Hyperoptic are not faster in the speed tables because they have users buying their 20 and 50 Mbps tiers and similar with Virgin Media.

Until 90% of homes own a UHD TV and buy UHD content then 20 to 30 Mbps is more than the majority of homes need to cope when everyone is at home.

The UHD market for video may push things towards the 100 Mbps being a more popular choice, i.e. run 3 streams at the same time and still have spare capacity.

My own playing around on a 20 Mbps long line is I can do 5 or 6 HD streams with no stuttering.

There's a news story around about a 1Pbps test that's been done on a single strand, so it's perfectly believable. It's more a matter of the opt-electronic interfaces. Not just the speed of switching, but the number of different wavelengths of light that are used.

However, I've got rather more doubts about a fibre capable of running at 99.97% (in a vacuum) for two reasons. First, optical fibre is made of refractive glass (or the light won't actually stay in the fibre). A the speed of light is inversely proportional to the refractive index, all current mono mode fibre of which I'm aware runs at around two-thirds of the speed of light in a vacuum.

The second reason is that the speed of propagation of light in fibre has little to do with the carrying capacity. Yes, it has something to do with latency (round-trip time is about 1ms per 100km of fibre), but not the actual bandwidth. The latter depends on things like the speed of optoelectronic switching and dispersion which tends to "smear out" the waveform over a long distance.

I would not expect to see those hyper-high speeds used in the distribution network as they require seriously expensive optoelectronics. They will be confined to back haul where it makes sense. In the distribution network cost and power consumption matter, so I think it will be a long time, if ever, before we see peta-bit/s capable fibre run into people's homes. After all, most domestic LANs tend to top out at 1gbps, let alone what WiFi delivers.

Oh I am aware of it all being contended I was just assuming it was a bit better LOL.

But yeah, I am sure it would be fine.

I have even seen phone poles that have the FTTP and have done for a few years now and cannot see any fibres going to the homes, so maybe they either are not aware of fibre down their road even though BTOR have stapled a leaflet on each pole saying fibre is here etc.

Lets hope everyone else orders option 1 or 2 LOL.

I am probably just being over paranoid and worrying over nothing.

As for UHD TV's I was tempted in buying one a few months ago, but decided on waiting until content becomes more available etc.

I will be mostly using our connection for game development for downloading and uploading our content files that go into around ~3GB which I cannot do over our ADSL2+ connection, so sending loads of burnt DVD's LOL, where when our FTTP finely goes live I would be able to upload it instead.

Paul

http://www.theregister.co.uk/2014/01/23/alcatellucen...

The 1.4 Tbps and it ran over live fibre with live traffic between Martlesham and BT Tower. Achieved by squeezing the multiple wavelengths closer together in the fibre.

Have seen it running (geek exciting but otherwise boring) but was at less than maximum speed on the day we visited, which was mainly to see the G.Fast kit http://blog.thinkbroadband.com/2014/10/g-fast-shows-...

The scare monger story recently of the UK running out of fibre capacity was really that, plenty of headroom for expansion and upgrades and adding new fibres if needed, or simply adding more diverse routes, plus unless Net Neutrality kills it off, caching video content at key points through out the network. There is a risk that Net Neutrality laws could outlaw for example the multi-cast live streaming BT TV uses for sport content all depends on how the laws are written and interpreted.

https://www.youtube.com/watch?v=0OOmSyaoAt0

In reply to a post by MrSaffron:

My own playing around on a 20 Mbps long line is I can do 5 or 6 HD streams with no stuttering.

What's HD in this case? 720p? 1080p?

My 11 Meg connection is usually ok with the latter, but my package limits kick in pretty quick if I overdo it.

A mix of 720p and 1080p depending on output device

In reply to a post by MrSaffron:

A mix of 720p and 1080p depending on output device

Sounds about right, we can get 1080p on our 4 to 5Mbit ADSL2+ as long as nobody else is using it LOL.

Paul

Nice to hear they're doing something!

In reply to a post by PaulKirby:

Well I know they said we have 40 odd stands for our area (i.e phone lines on our cabinet) which due to not having that cable they used a 90 odd strand one.

It depends a little on *where* in your area they are talking about.

Between the Aggregation Node and the splitter node, the fibre will be GPON-shared (at densities of up to 32x, 64x or even 128x, but I suspect less than 32x will be more common), so fewer strands will be needed. However, the cable could be looped-through a splitter node, with most of the strands running through (without splicing) to the next splitter in the path.

A splitter node itself can have 4 splitters, each with 32 outputs. It will barely need one group of fibres (almost all the cables seem to group the strands of fibre into groups of 12) to feed the 4 splitter inputs, even accounting for spares.

Between the splitter node and the fibre distribution points, everyone's connection will be on an individual strand of fibre, so this part needs a higher density of strands per cable. Each fibre DP can cope with (IIRC) 24 premises, and I can imagine they'd feed 48-strand cable from the splitter output to one of the DPs (with plenty of spares); possibly more strands if the cable loops-through to head towards further fibre DP's.

Between the fibre distribution points and the manifolds, I'd expect 7-core or 12-core BFT as empty tube with no fibre whatsoever. Only when a connection is ordered will BT blow fibre through this part ... probably using 4-strand cable for each property.

Some info here:
http://forums.thinkbroadband.com/fibre/4375161-fibre...

But if say we go by what you said, in the worst cases if everyone hammers their connection we could see speeds as slow as 56Mbits or so from our soon 330Mbit connection.

It is ironic that the first generation of GPON creates about the same density of capacity in the access network in our streets as FTTC.

For example, one FTTC cabinet of 288 lines, with an average of 50Mbps, will spread over 14 Gbits of capacity over the streets it serves. Adding vectoring might make the capacity 20 Gbits.

The same lines, split 32x in a GPON, would get 9 strands of backhaul fibre. Each would have 2.5Gbps, giving a total of 22Gbits.

It's a good job we don't all use it in one go...

In reply to a post by WWWombat:

Nice to hear they're doing something!

Tell me about it, my local MP was not very amused LOL, I was told by many BTOR engineers when asked when in the chamber/pit that it had already been done (our end) back in 2011 when the rest of our road was fibred up.

My guesses why we didn't get it was due to blockages up the road which has reciently just been resolved which was why I was about 95% sure that BTOR would be down our road not long after LOL.

In reply to a post by WWWombat:

In reply to a post by PaulKirby:

Well I know they said we have 40 odd stands for our area (i.e phone lines on our cabinet) which due to not having that cable they used a 90 odd strand one.

It depends a little on *where* in your area they are talking about.

Well I was told for our area where I said what the area for our cabinet where I got a small slight nod of the head.

In reply to a post by WWWombat:

Between the Aggregation Node and the splitter node, the fibre will be GPON-shared (at densities of up to 32x, 64x or even 128x, but I suspect less than 32x will be more common), so fewer strands will be needed. However, the cable could be looped-through a splitter node, with most of the strands running through (without splicing) to the next splitter in the path.

A splitter node itself can have 4 splitters, each with 32 outputs. It will barely need one group of fibres (almost all the cables seem to group the strands of fibre into groups of 12) to feed the 4 splitter inputs, even accounting for spares.

Between the splitter node and the fibre distribution points, everyone's connection will be on an individual strand of fibre, so this part needs a higher density of strands per cable. Each fibre DP can cope with (IIRC) 24 premises, and I can imagine they'd feed 48-strand cable from the splitter output to one of the DPs (with plenty of spares); possibly more strands if the cable loops-through to head towards further fibre DP's.

That 48 stands must be the 90 odd strand cable that they put in insted, I was told it might still only use 48 of the 96 depending on what its connected back onto which makes sense.

As far as I know we are at the end of our fibre line due to the next pole is on a different cabinet than us.

In reply to a post by WWWombat:

Between the fibre distribution points and the manifolds, I'd expect 7-core or 12-core BFT as empty tube with no fibre whatsoever. Only when a connection is ordered will BT blow fibre through this part ... probably using 4-strand cable for each property.

Some info here:
http://forums.thinkbroadband.com/fibre/4375161-fibre...

Yeah, I recall one of them saying either 12 or 18, so depending on how many sign up for fibre on our pole, but they might have to add another tubing on there.

BTW, whats this 4 srtand per property?
I thought we only got one, or is it a form of redundancy?

In reply to a post by WWWombat:

But if say we go by what you said, in the worst cases if everyone hammers their connection we could see speeds as slow as 56Mbits or so from our soon 330Mbit connection.

It is ironic that the first generation of GPON creates about the same density of capacity in the access network in our streets as FTTC.

For example, one FTTC cabinet of 288 lines, with an average of 50Mbps, will spread over 14 Gbits of capacity over the streets it serves. Adding vectoring might make the capacity 20 Gbits.

The same lines, split 32x in a GPON, would get 9 strands of backhaul fibre. Each would have 2.5Gbps, giving a total of 22Gbits.

It's a good job we don't all use it in one go...

I didn't think vectoring would do anything for FTTP and was more for the FTTC.

Paul

Vectoring is only for VDSL2/G.Fast solutions.

No RF interference to cause issue on FTTP

In reply to a post by MrSaffron:

Vectoring is only for VDSL2/G.Fast solutions.

No RF interference to cause issue on FTTP

I thought so.

Paul

Also I was assuming that the fibre that daisy chains in and out of the Fibre DP would have a yellow stripe on it, the cable I saw them putting in the chamber was about 10mm thick and black and with no stripe, it seems the same size as the one with the yellow plastic wrapped round each of the two cables.

Maybe this cable also contains the actual fibre due to I was told the fibre doesn't need to be blown in for this short distance.

I know some Black tubing with the yellow strip was being put in yesterday.

Paul

That would just be tubing

http://www.coolwebhome.co.uk/fibre-milton-keynes/ Has pictures of all the bt kit including the very thin fibres

In reply to a post by MrSaffron:

That would just be tubing

http://www.coolwebhome.co.uk/fibre-milton-keynes/ Has pictures of all the bt kit including the very thin fibres

Yeah, the tubing with yellow stripe I saw yesterday was I think to go up the pole from the fibre dp enclosure, however that had already been done, but I think they may of changed some due to the metal guard had been removed and re-nailed back down.

But the ones I saw today was the one shown in image IMG_4340 on the right with the yellow tape wrapped round it, couldn't see any yellow stripe on it, its basically half the size of the tubing that goes up the pole.

I can see the cable in image IMG_4350 for the in and out daisy chain cable, cannot see a stripe on that so its probably that cable, which has the fibre already in it so I was told.

I did see a thinner off cut version of the tubing with yellow stripe yesterday that was shown to me by the engineer, so that must of been the 7 core (i.e. smaller version) fibre BFT

Paul

In reply to a post by PaulKirby:

BTW, whats this 4 srtand per property?
I thought we only got one, or is it a form of redundancy?

Redundancy, I guess.

You can see the mention of 4 strands in this picture gallery (the pictures show pretty much the same as those Mr Saffron took):
http://www.zdnet.com/pictures/photos-fibre-broadband...

Inside the house, it looks like BT prefers EZ Bend fibre, which comes with 2 strands; one strand from this will be spliced to 1 strand of the 4, and the others laid up as spares.

I didn't think vectoring would do anything for FTTP and was more for the FTTC.

My reference to vectoring was because, in that paragraph, I was talking about the total capacity available across all lines connected to an FTTC cabinet - which could indeed be augmented by vectoring.

I've also seen reference in the BT developer's guide that suggest their GPON intentions will be for no more than a 32x split.

In reply to a post by WWWombat:

In reply to a post by PaulKirby:

BTW, what's this 4 strand per property?
I thought we only got one, or is it a form of redundancy?

Redundancy, I guess.

You can see the mention of 4 strands in this picture gallery (the pictures show pretty much the same as those Mr Saffron took):
http://www.zdnet.com/pictures/photos-fibre-broadband...

Inside the house, it looks like BT prefers EZ Bend fibre, which comes with 2 strands; one strand from this will be spliced to 1 strand of the 4, and the others laid up as spares.

I guess they could use one if you wanted an additional fibre connection to be greedy LOL, but yeah, probably for spare, I will ask when it goes live and they come and install it.

In reply to a post by WWWombat:

I didn't think vectoring would do anything for FTTP and was more for the FTTC.

My reference to vectoring was because, in that paragraph, I was talking about the total capacity available across all lines connected to an FTTC cabinet - which could indeed be augmented by vectoring.

Ah, ok.

In reply to a post by WWWombat:

I've also seen reference in the BT developer's guide that suggest their GPON intentions will be for no more than a 32x split.

Is that 32x split per strand or as whole cable (all 48/96 strands)?

Paul

WWWombat

I think you are wrong about the FTTC capacity. You have added up the Total bandwidth to the users. But the Back haul from the Cabinet to the headend will be a lot less.

I seem to remember the details when first provided where 2 cards of 48 ports to 1 Ge backhaul. with a max of 6 cards and 4 Ge backhaul so nowhere near your 14Gbit

That gives 13.9Mb each if everyone is using at the same time, (unusual with 288 customers). Figure last year were suggesting an average country wide at less than 1Mb.

The GPON kit goes all the way to the head end at 2.5Gb so between 32 customers would be 78.125 Mb per customer before contention kicks in. At 10Gb GPON that would be 312.5Mb. ( That is why the split goes up to 128 bringing it back to 78M)

Backhaul from the Headend is likely to be 10Gb per wavelengh but both Sky and BT have 100Gb systems in the core ( from press releases) and 400Gb per wave length have been tested. Last time I looked 40 Wavelengths seemed to be the max used in a live network but I think those are at 10Gb.

In theory a pair of fibres could carry 40*400=16Tb or even higher if the 1.4Tb on a wavelength was practical in a live deployment.

This is all theorectical as there is not that amount of data in the country in total. (and Caching of video reduces the amount in the core dramatically from that carried at the edge.

In reply to a post by PaulKirby:

In reply to a post by WWWombat:

I've also seen reference in the BT developer's guide that suggest their GPON intentions will be for no more than a 32x split.

Is that 32x split per strand or as whole cable (all 48/96 strands)?

32x per strand.

So one strand is needed from the head-end, out along the spine to the aggregation node, spliced onto one strand to the splitter. There it is split onto 32 strands heading out to fibre DP's, and then onwards to 32 homes.

To then scale these numbers up to town-sized and city-sized solutions...

To cover a town of 10,000 premises (ie 10,000 lines), you'd need just 312 active strands heading out of the head-end.

If the fibre spines head out of an exchange in 8 different compass directions, it would need 40 active strands in each spine. Plus lots of spares, I imagine.

For this size, you can see where 48 and 96-strand cable comes in handy.

A city with 100,000+ premises is likely to be covered by 10+ exchanges, each with 10,000 - 30,000 lines. Presumably the head-end only exists in some of the larger exchanges, and the spines out from these will be taking the workload away from the smaller exchange. I guess the spines could then be 10x bigger for such places.

It is then easy to see how cables with 288 fibre strands can be used.

In reply to a post by WWWombat:

In reply to a post by PaulKirby:

Is that 32x split per strand or as whole cable (all 48/96 strands)?

32x per strand.

Ok.

In reply to a post by WWWombat:

So one strand is needed from the head-end, out along the spine to the aggregation node, spliced onto one strand to the splitter. There it is split onto 32 strands heading out to fibre DP's, and then onwards to 32 homes.

So, a single strand is split into 32 strands which would be for 32 homes.
So for all homes on our PCP Cabinet (~350 homes) we would need X amount of strands that go off into each of the fibre dp's in the chambers/pits and each of those get split off into 32 strands, and where there isn't 32 homes per pole (my area) we would end up with redundant strands for later use.

In reply to a post by WWWombat:

To then scale these numbers up to town-sized and city-sized solutions...

To cover a town of 10,000 premises (ie 10,000 lines), you'd need just 312 active strands heading out of the head-end.

If the fibre spines head out of an exchange in 8 different compass directions, it would need 40 active strands in each spine. Plus lots of spares, I imagine.

Yeah, nothing worse than having to lay another cable due to needing just 1 more strand LOL.

In reply to a post by WWWombat:

For this size, you can see where 48 and 96-strand cable comes in handy.

Yeah, so if we use the 48 of the 96 strand cable they installed (like I said, only cable they had on them) we would have loads spare LOL.
Well depending on what the other end is connected to.

In reply to a post by WWWombat:

A city with 100,000+ premises is likely to be covered by 10+ exchanges, each with 10,000 - 30,000 lines. Presumably the head-end only exists in some of the larger exchanges, and the spines out from these will be taking the workload away from the smaller exchange. I guess the spines could then be 10x bigger for such places.

Coming from our exchange there are about +25,500 lines, tbh, most if not all the exchanges around my area are on the large side, I guess that's due to being located in east London.

In reply to a post by WWWombat:

It is then easy to see how cables with 288 fibre strands can be used.

Yup.

Paul

In reply to a post by PaulKirby:

Hello All

Just wondering what the actual speed of an individual strand of fibre was.

This is for our FTTP install, where we are now just waiting for a fibre joiner to join our other end to where the faulty cable connected.

Basically BTOR has just put in a thin fibre cable (at a glance it seems the thickness of RG213 Coax) with about 96 or so strands instead of the 40 odd due to not having any, so I am aware depending on what that connects to would determine how many it will use leaving the rest for spare in case of issues.

But say that it only uses like 48 strands for all the homes on our phone cabinet which is about 394 (including phone boxes, bank tills and TFL) so say 350, that means 8 homes could/will be sharing the same strand, I did read some place that a fibre stand can handle several Gbits to Tbits or more depending where you read, so I was wondering what the average max speed would be per strand.

Thanks

Paul

Technically a fibre strand (The cable itself) can support almost unlimited data (There will be a limit somewhere). The limit is how fast the equipment on either end can flash light down it rather than the cable itself.

In reply to a post by kitcat:

I think you are wrong about the FTTC capacity. You have added up the Total bandwidth to the users. But the Back haul from the Cabinet to the headend will be a lot less.

Absolutely agree that the backhaul is insufficient.

I was trying to show the total capacity of the copper spreading out across the streets from the cabinet (rather than the capacity of the fibre into the cabinet). And to show that this, ironically, happens to be very similar to the chosen capacity of the 1st generation of PON that will spread out in the same streets.

I'm well aware that there are limitations on the backhaul currently (I think your 4 Ge is correct), but that doesn't have to stay the limitation forever. 10Ge backhaul could be switched in in the future ... but whatever the backhaul, it won't change the total capacity of the copper still spreading out.

The total capacity of the copper will change if/when G.fast gets deployed, but we won't know anything about the total capacity until BT have worked out just how far they will be putting nodes from homes.

That gives 13.9Mb each if everyone is using at the same time, (unusual with 288 customers). Figure last year were suggesting an average country wide at less than 1Mb.

I agree.

The peak-hour usage averaged over everyone still comes out less than 200kbps, I think. That includes both FTTC subs and ADSL subs. My current rule-of-thumb is that average FTTC subscribers consume about twice the bandwidth per month as average ADSL consumers.

We're a long way from needing the cabinet backhaul to be substantially upgraded.

For the use an individual home might present, as opposed to the average over many, many users, you might like this:
http://www.bcssouthwest.org.uk/presentations/DougWil...
and
http://www.bcssouthwest.org.uk/server.asp?page=video...

At 10Gb GPON that would be 312.5Mb. ( That is why the split goes up to 128 bringing it back to 78M)

The split *could* be increased, but I suspect an upgrade to a 10G PON would happen because more capacity is needed for the 32, rather than allowing the split to be made larger.

Everything I've seen from BT suggests only a 32-way split - either in a single go, or two layers of 8x and 4x.

But never say never ... remember DACS.

Backhaul from the Headend is likely to be 10Gb per wavelengh but both Sky and BT have 100Gb systems in the core ( from press releases) and 400Gb per wave length have been tested. Last time I looked 40 Wavelengths seemed to be the max used in a live network but I think those are at 10Gb.

You might like these presentations from UKNOF earlier in the year...

Sky core revamp: https://www.youtube.com/watch?v=uQUoO4Wb7s4
BT bringing North-West Scotland closer to their core: https://www.youtube.com/watch?v=5ZU5tvBUwZM

In reply to a post by PaulKirby:

So, a single strand is split into 32 strands which would be for 32 homes.

Yup.

In reply to a post by PaulKirby:

So for all homes on our PCP Cabinet (~350 homes) we would need X amount of strands that go off into each of the fibre dp's in the chambers/pits and each of those get split off into 32 strands, and where there isn't 32 homes per pole (my area) we would end up with redundant strands for later use.

The split happens before the fibre DP (ie closer to the exchange). The fibre DP is the point where your individual strand swaps from being within a cable shared with 47/95 others (for other homes) to being within a cable that is dedicated for your home (and blown to it within BFT).

- A pole might average around 8 homes connected to it, with a manifold at the top or in the chamber
- A fibre DP might have 16-20 homes connected to it, so probably serves 2 poles.
- A splitter node can serve up to 128 properties, so can serve around 6 fibre DP's (ie 12 poles). Perhaps this would be one per street?
- A cabinet would have served 300-500 properties, so probably needs to have 3-6 splitter nodes.

I imagine that *everywhere* probably ends up with 50% redundant strands. I guess copper used to look like that in the sixties and seventies.

In reply to a post by WWWombat:

In reply to a post by PaulKirby:

So, a single strand is split into 32 strands which would be for 32 homes.

Yup.

Cool.

In reply to a post by WWWombat:

In reply to a post by PaulKirby:

So for all homes on our PCP Cabinet (~350 homes) we would need X amount of strands that go off into each of the fibre dp's in the chambers/pits and each of those get split off into 32 strands, and where there isn't 32 homes per pole (my area) we would end up with redundant strands for later use.

The split happens before the fibre DP (ie closer to the exchange). The fibre DP is the point where your individual strand swaps from being within a cable shared with 47/95 others (for other homes) to being within a cable that is dedicated for your home (and blown to it within BFT).

Ok.

In reply to a post by WWWombat:

- A pole might average around 8 homes connected to it, with a manifold at the top or in the chamber
- A fibre DP might have 16-20 homes connected to it, so probably serves 2 poles.
- A splitter node can serve up to 128 properties, so can serve around 6 fibre DP's (ie 12 poles). Perhaps this would be one per street?
- A cabinet would have served 300-500 properties, so probably needs to have 3-6 splitter nodes.

Well we have a fibre dp in all our chambers/pits by our poles (well the several I have seen in down my road and side roads), and also 16 to 20 homes may not be enough due to in some places there is about 24 to 27 homes, so would the homes on that fibre dp not all get fibre, our one has about 20 odd homes and some of those homes have multi lines, but I guess those multi lines if they want to be greedy could use one of the 4 strands I guess.

In reply to a post by WWWombat:

I imagine that *everywhere* probably ends up with 50% redundant strands. I guess copper used to look like that in the sixties and seventies.

Yeah, like I said its better to have spare ones than not have enough.

Paul