Continues from: /forum-replies.cfm?t=2437721&p=-1#bottom
On the topic of supported modems, whilst I don't support the NBN stance on not releasing the list, it's not really their place to be doing it. In reality it is up to the RSP to test the modem/s they're going to support in the NBN test labs (which exist only in Melbourne) to get a tick against that particular model of modem/router for 'supported' on NBN.
Once the RSP has successfully tested a modem on NBN FTTN, then it would be up to the RSP to publish the information somewhere for it's customers to use. Different RSP's can/will have different "supported" devices due to their own internal support procedures, just the same as they would right now for ADSL ("you've got a fritzbox, what the hell is that, some kind of hair dryer ? Sorry, we only provide support for xyz").
In this way if each RSP provides a list of modems that they support, then collating the list from the different RSP should provide a list of models that have been tested in the NBN kitty litter (sandbox).
Remember, just because your device of choice is "not on the list", doesn't mean you can't use it, just means that at some point in the troubleshooting progress NBNCo will spit the dummy and say you're using a non-supported modem and that they won't check anything else until you rectify this.
NBN Co encourages but does not require Customer to register VDSL2 Equipment to be used with the NEBS supplied by means of the NBN Co FTTB Network or the NBN Co FTTN Network. The NBN Co Operations Manual sets out when NBN Co will, and will not, investigate and repair any faults associated with services that utilise non-registered VDSL2 Equipment, and whether any fees will apply to the investigation and repair of such faults.
The registration process consists of a self-certification, executed by Customer, followed by a registration of the self-certification by Customer to NBN Co. To register VDSL2 Equipment, Customer must supply the VDSL2 Equipment Vendor ID, System ID and Version Number as described in ITU-T G.993.2 section 11.2.3.6, as well as a clear-text name uniquely identifying the combination of hardware and firmware to be entered into the registration database.
As per the above, as soon as the firmware on a device is upgraded it becomes "non-complaint" until verified again, so hold off on those firmware upgrades to fix issues !
In reality it is up to the RSP to test the modem/s
It would also seem to make sense for Australian distributors to get their VDSL2 modems validated, so they could be sold as "NBN compliant".
It would also seem to make sense for Australian distributors to get their VDSL2 modems validated, so they could be sold as "NBN compliant".
Assuming this is something that can be done, yes, I'd agree that would make sense. Depends if NBN allows this to happen I guess.
doesn't mean you can't use it, just means that at some point in the troubleshooting progress NBNCo will spit
That implies that any modem "will" actually work just that you'll have problems when it comes time to trouble shoot with NBN which doesn't seem to be the case as it is possible a non-compliant modem "may" not work at all.
/forum-replies.cfm?t=2478560&p=3#r56
"If your modem isn't compatible with one or more of those features, it may not sync and your DSLAM port will shut down until it is manually reset by our Fibre team. This is enforced by nbn and not something we can control."
Draytek seem to be claiming their VDSL capable modems are NBN compatible, it'll be interesting to see how they go.
/forum-replies.cfm?t=2478560&p=3#r56
"If your modem isn't compatible with one or more of those features, it may not sync and your DSLAM port will shut down until it is manually reset by our Fibre team. This is enforced by nbn and not something we can control."
The bit I quoted about customers self-registering equipment in the DB is from the same part of that same document. It does go on to say;
where specific VDSL2 Equipment or a certain model and/or firmware of VDSL2 Equipment is causing (or NBN Co reasonably considers that it is likely to cause) detriment to other services, NBN Co may:
- remove the VDSL2 Equipment from the VDSL2 Equipment registration list;
- place an Ordered Product using that VDSL2 Equipment into a Repair Profile; and/or
- Suspend an Ordered Product using that VDSL2 Equipment in accordance with the Head Terms.
Which is what the iiNet guy was saying.
You CAN use non-NBNCo-certified VDSL modems, just that if they don't meet the NBNCo specs, then they'll shut your port down. Doesn't mean you can't use your own modem, just that there could be consequences if you do. If you read the whole thread in that iiNet post, this appears to be what happened to the OP, used a modem that didn't support something, his port got shutdown, ordered an NBN compliant modem from iiBorg and then got his port enabled again and the new modem worked (it was a slightly longer process, but thats the quick version).
Draytek seem to be claiming their VDSL capable modems are NBN compatible, it'll be interesting to see how they go.
http://www.draytek.com.au/p
Not inspiring me with confidence when under the section for "FTTP" they have a drawing of FTTN (it's got a DSLAM on it, so it's not just a typo in the image).
On the topic of supported modems, whilst I don't support the NBN stance on not releasing the list, it's not really their place to be doing it.
Sure, as long as you are happy with your RSP dictating how you can access the government owned access network and are willing to hand over all responsibility to your RSP for choosing what features you will be able to use in your access device.
I also assume you are comfortable with giving your RSP a remote backdoor to your network due to their braindead implementation of remote management of "your" modem.
This does not sound like a sane plan. It sounds like the outcome of NBN Co wanting to shift the costs of a NTD onto the end user.
Draytek seem to be claiming their VDSL capable modems are NBN compatible, it'll be interesting to see how they go.
The problem is they are not "VDSL capable modems" they are VDSL capable modem/routers.
So are they saying the VDSL modem is NBN compliant or are they only refering to the router having a WAN port to connect to the NTD thus the router being NBN ready should you be getting it for an FTTP connection and ignoring the VDSL part?
You would hope that saying NBN ready on a VDSL modem/router would mean both the modem and router would be compliant but it doesn't mean they both are if it doesn't indicate that.
What they need to do is say "NBN FTTN/B ready" or otherwise indicate somehow that the VDSL modem is NBN ready and not just the router in it.
Yeah, they don't explicitly state that their modems will work with nbntm's FTTN/B rollout, they just state that they are able to do VDSL2 with vectoring while pointing out that is the tech that the rollout is based on.
Maybe I'm getting a bit cynical after listening to Mr Turnbull's views on the NBN for so long, but I take things with a grain of salt when it isn't -explicitly- stated. Hence the "seem to be claiming".
It would also seem to make sense for Australian distributors to get their VDSL2 modems validated, so they could be sold as "NBN compliant".
It'd be interesting to see if there was a registration fee as I doubt Australian distributors would be interested in burning through their profit margin. Imagine paying out each time a new firmware was released on something you make little out of in the first place.
This echoes what is seen with the VAST (freeview satellite) service. Freeview via Optus imposes something like a, IIRC, $30,000 certification fee. Change the firmware, and the box needs to be recertified.
Imagine paying out each time a new firmware was released on something you make little out of in the first place.
Even if the option is not being able to tell customers that it will work on the NBN FTTN?
It'd be interesting to see if there was a registration fee as I doubt Australian distributors would be interested in burning through their profit margin. Imagine paying out each time a new firmware was released on something you make little out of in the first place.
it is even worse than that, the NEBS document says nbn� can change the spec as they choose and thus a previously certified modem may no longer have certification and nbn� can deem it now causes problems and shut down the end user
so many twists and turns to remove the active "ntd" from nbn�'s books and offset a minor cost according to Mal "less than $50" he claimed.
I bet the procedures that have now been instigated have a greater than a $50 cost per FTTN/B connection, when you take into account the certification, the wholesale costs, the transport, the RSPs time trying to manage them on top of nbn� managing the VDSL ports
MTM makes the committee that wanted a horse and ended up with a camel seem a raging success
Yeah, they don't explicitly state that their modems will work with nbntm's FTTN/B rollout, they just state that they are able to do VDSL2 with vectoring while pointing out that is the tech that the rollout is based on.
At least Netgear to their credit don't have a big "NBN READY" sticker plastered on their VDSL routers. They do though say on the bottom "NBN&UFB Ready" (and under it say Dedicated Gigabit Port) so average joe could still easily interpret that as the modem rather than the router being NBN Ready and capable of 1Gbps instead of only 100Mbps.
so many twists and turns to remove the active "ntd" from nbn�'s books and offset a minor cost according to Mal "less than $50" he claimed.
I've still yet to see any VDSL2 modem/router (or even just a dumb modem) for $50 let alone less than $50, even ones on ebay located in hongkong.
less than $50
Yep, another one of those remarks that will come back to haunt MT as being out of touch with the real world and not realising that for some people having to spend $150 (or even $50) extra is a very real cost.
In this way if each RSP provides a list of modems that they support, then collating the list from the different RSP should provide a list of models that have been tested in the NBN kitty litter (sandbox).
That's not very efficient nor going to give consumers choice at all. Why would any RSP expend hard earned on testingcand publishing so their competitors can lap up the cream?
Belong offers their Sagemcom 4315 for $60 on no contract (versus free on 12-mth contract) so I'm guessing that as an approximate price of it (albeit not publicly available AFAIK)
so many twists and turns to remove the active "ntd" from nbn�'s books and offset a minor cost according to Mal "less than $50" he claimed.
I bet the procedures that have now been instigated have a greater than a $50 cost per FTTN/B connection, when you take into account the certification, the wholesale costs, the transport, the RSPs time trying to manage them on top of nbn� managing the VDSL ports
MTM makes the committee that wanted a horse and ended up with a camel seem a raging success
At least someone understands the problem, I can't believe the tech media aren't paying attention to this.
blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah
I want a VDSL2 modem that will just bridge into my proper router's ethernet port. Are these available?
So I have another question.
About 200 cable meters away from me, they have installed a node. Unfortunately I don't think I will be connected to this one as I am connected to a DA that is 610 cable meters away.
I've managed to find this out by using multiple "dial before you dig" requests and compared the location of the DA pillars to where I am.
Question is: Can I get moved to the node that is physically closer to my home, or will I be connected to the node that is replacing my current DA pillar? I'd rather be 200m away from the node than 600m away!
Question is: Can I get moved
No
or will I be connected to the node that is replacing my current DA pillar?
You will be connected to a pillar, no matter what. Pillars arent going away. it goes Home---->Pillar----->Node.
not in all cases. ;-)
but no, you cannot move from one node to another, and more than you can from one exchange to another.
Looks like half the country is down � as are the businesses that rely on these networks who it seems are not covered for losses
Perhaps a gross overstatement of the situation. The affected areas won't necessarily have lost service, it's just that Openreach can't guarantee their usual timescales for rectification if they do. Besides, if your local FTTC cabinet has flooded, likely so has your own property, maybe even your local telephone exchange, so you've probably got bigger things to worry about than your internet connection
(indeed, the lack of national news regarding "half the country" losing their internet connections would seem to prove that)
The same would be true if there was a freak accident that endangered an FTTP network, say an exchange had burnt down or flattened (which, since the BT FTT* network is built out from the largest exchanges in a given area, would affect a lot of people spread out over a wide area)
So show me an ISP supplied modem which doesn't work.
Can't comment on the Australian network, but over here in the UK we've had some interesting whoppers regarding ISP modem compatibility.
Openreach/BT enabled G.INP on both upstream and downstream across their VDSL network a few months ago. It revealed that a number of ISP supplied modems (in particular those with a chip from a specific manufacturer) were having weird problems. Most notably, one model of modem/router supplied by BT themselves! I think upstream has been disabled again for the time being
Meanwhile the original batch of VDSL modems (made by Huawei, issued by Openreach themselves) continue to work fine.
About 200 cable meters away from me, they have installed a node.
You should be able to buy FTTP for pretty cheap in that case probably just a few thousand dollars.
They will just run a fibre cable down the Telstra pit back to the other node 200 meters away or the nearest distribution cable (possibly even less than 200 meters in your case) and splice you straight in no problems.
You should be able to buy FTTP for pretty cheap in that case probably just a few thousand dollars.
Not according to NBNCo...
They will just run a fibre cable down the Telstra pit back to the other node 200 meters away or the nearest distribution cable (possibly even less than 200 meters in your case) and splice you straight in no problems.
Sorry, did you say "splice"? Fibre doesn't work that way...
Sorry, did you say "splice"? Fibre doesn't work that way...
That isn't adding connections, that is patching a broken connection
At the moment I can commercially purchase OM3 pre-terminated SC connector fibre (3pr) for about $800 for 250 mtrs. Purchased 2 lots last year for clients. Not to bad actually.
Dave
The same would be true if there was a freak accident that endangered an FTTP network, say an exchange had burnt down or flattened (which, since the BT FTT* network is built out from the largest exchanges in a given area, would affect a lot of people spread out over a wide area)
There is a big difference between the physical security of an exchange vs. that of a node cabinet. Exchanges are built in strategic locations with vastly superior physical security, fire protection etc, whereas nodes are typically built less than a metre from a road with high speed vehicular traffic without any impact protection like bollards and are often in locations close to a roadside culvert with lower elevation deliberately designed to pool water.
At the moment I can commercially purchase OM3 pre-terminated SC connector fibre (3pr) for about $800 for 250 mtrs
Purchasing the material is a small cost compared with the labour to install it. Even for above ground indoor installs the cost of materials plus labour is around ~$1500 per 100 metres plus another $1000 to field terminate the ends and test for correct operation. External underground fibre cable work needs far more labour depending on the obstructions so the cost is far greater.
Purchasing the material is a small cost compared with the labour to install it.
Thats true but if hes really only 200 meters from a fibre connection the labour would not be very much at all. Thats walking distance away.
After they pull it through the pit all they're going to do is splice his cable into either the node or the fibre distribution cable and install the NTD and he'll be right. Few hundred for the fibre and maybe a thousand for the labour.
Heres the apply link:
https://www2.nbnco.com.au/te
After they pull it through the pit all they're going to do is splice his cable into either the node or the fibre distribution cable
I don't think thats how fibre works... You dont just "splice" an additional cable into an existing cable.
I don't think thats how fibre works... You dont just "splice" an additional cable into an existing cable
Exactly right...and because the nodes do not have room for the FTTP cards in them, they would need to run a fibre connection all the way back to the FAN or the exchange.
The cost would probably be between $10k and $200k...
Exactly right...and because the nodes do not have room for the FTTP cards in them, they would need to run a fibre connection all the way back to the FAN or the exchange.
it seems that for FOD they will be running the fibre to the multiport upstream of the node, that is what is indicated in the network design rules.
But until an actual detailed quote is visible with a full detail breakdown of what is being done.
How it actually is being done is speculation by ourselves.
I guess this is just a part of the total transparency of nbn� they do not seem to have an actual public available guide as to HOW a FOD will be done, not sure if they actually have an internal one either.
Given some of the contradictory statements given to the few on Whirlpool who have had some attempts to get at least to the detailed quote stage
edited for clarity at one point
it seems that for FOD they will be running the fibre to the multiport upstream of the node
That makes sense...it just wasn't in their press releases...
http://www.nbnco.
I had forgotten that they were moving to the riskier star topology...
One has to wonder about what kind of access the multiport will have...
I had forgotten that they were moving to the riskier star topology...
One has to wonder about what kind of access the multiport will have...
if you have a look here jxeeno has an extract image of the topology
http://blog.jxeeno.c
you can use google to find and download the full document from nbn� for a light read
they would need to run a fibre connection all the way back to the FAN or the exchange.
Indeed. And for a FTTN area, there may just be fibre-port ethernet switches in the exchange. So an FTTP GPON fibre coming to the exchange would require a new GPON OLT shelf and port card which is a big chunk of money (unless they use some kind of non-standard product designed for small installations which they would be reluctant to do as it complicates inventory, training etc). The installation and commissioning costs for what might be only 1 customer in whole area need to be added as well. How much of all this would charged back to the FOD customer is not known either.
There is a big difference between the physical security of an exchange vs. that of a node cabinet.
Sorry to burst your bubble, but that is not true. My old primary school now has more physical security (an 8 foot high fence surrounding the building and parts of the play grounds just to protect against vandalism) than all of the Telephone exchanges that I've been into, which includes some major CBD ones.
You're comparing telephone exchanges with node cabinets as though they perform exactly the same purpose. They don't compare. There is a big difference in the consequences of failure or destruction and the time to repair replace.
Consequences of an exchange failure or destruction and the time to repair of an exchange are far greater than those of a street side node. The reason street side cabinets don't have the amount of physical and fire protection that exchanges have is because street side cabinets don't need it.
Failure or destruction of a street side cabinet will only impact a few hundred services, no more than 1000, and a street side cabinet could be replaced in the order of a week in the case of destruction. This is because as they're much cheaper (and smaller) than telephone exchanges, they can be kept in a warehouse, partially pre-built, ready to be deployed when necessary. You can have spares of street side cabinets, you cannot have "spare" exchanges.
Failure or destruction of an exchange, on the other hand, will impact 10s of 1000s of services, and will take months if not more than a year to replace in the case of destruction.
"Warrnambool Exchange Fire � Resilience and Emergency Management"
http://telsoc.org/ajtde/2014-12-v2-n4/a72
Warrnambool isn't a once off either:
"Telstra techs face second fire-damaged exchange"
http://www.itnews.c
Smaller units of failure, such as street side cabinets, are far better because their failure has less impact and recovery is easier and quicker.
But until an actual detailed quote is visible with a full detail breakdown of what is being done.
How it actually is being done is speculation by ourselves.
I guess this is just a part of the total transparency of nbn�
There was a person from Bundaberg that claimed to be getting FOD (I think you might have read it).
The conversation started here:
whrl.pl/Revqhl
He started discussing the build but shortly afterwards deleted all his posts (and the pictures).
I guess transparency from nbn� is out of the question so we'll never know.
You know, its things like this that should be reported in MSM....
Interesting read. Mine is still connecting at 50/13, yet now my speedtests (and real world tests too) are coming out at 44/6.
What I find interesting and that I'm collating data on is that the FTTN speeds look to be failing on the 5:2 ratio of download/upload. At 50mbps down, you should be achieving 20mbps up. Instead I'm seeing more often that it is closer to a 5:1 ratio.
It will be interesting once collated to see if it's possible that upload is being shorted to bump up the download speed.
http://www.computerworld.co
In that article, Bill Morrow is quoted as saying:
�I will point out that we looked at every complaint that has been received on fibre to the node, and not one of them was actually a speed issue that was related specifically to the fibre to the node technology,� Morrow said.
which if he can be believed says that the problem is NOT related to lack of backhaul capacity from an individual node. Perhaps it is just a case of the POI CVC upgrade interval/process and RSP's being caught wanting in this area. Will be interesting to see that once RSP's have upgraded CVC whether that might put more stress on the node backhaul component. Right now it might not be a congestion point because there isn't enough traffic on each one due to lack of RSP CVC being the mitigating factor.
The other thing I take away from that article is that people should keep complaining to their relevant local member about problems.
which if he can be believed says that the problem is NOT related to lack of backhaul capacity from an individual node.
No, actually what he is saying is the Sync speeds from the Node to premises are not the problem.
actually what he is saying is the Sync speeds from the Node to premises are not the problem.
It's hard to say either way. It is a bit vague in that he's saying it's "not the fault of FTTN technology", so what is he including in that, just the POTS tail or the backhaul from node to POI as well ?
It's hard to say either way. It is a bit vague in that he's saying it's "not the fault of FTTN technology", so what is he including in that, just the POTS tail or the backhaul from node to POI as well ?
I would take it as only the node itself
But it is so ambiguous as to not be a proper accurate answer.
He said he doesn't know if they monitor the node fibre links, so that would rule them out of his answer
He said he doesn't know if they monitor the node fibre links
Was that in the same article, or somewhere/when else ?
that would rule them out of his answer
Not categorically. Just because they don't monitor them all the time, doesn't mean that in each of these instances of poor performance that they've investigated they haven't checked the utlisation on the node backhaul links.
I'd tend to agree though, my above answer is clutching a little bit and it would certainly be easier to believe that they are not monitored and so hence could be the reason for the congestion shown by users. Unfortunately there is no way of knowing until you get some inside knowledge either from an RSP (that knows what their CVC usage is to the POI and hence whether that is the congestion point) or from someone at NBN who knows a bit more.
Was that in the same article, or somewhere/when else ?
last night in Senate Estimate's
It will be interesting once collated to see if it's possible that upload is being shorted to bump up the download speed.
Nice pick up there MrMac, look forward to seeing the compilation of results and your determinations of that!
It's hard to say either way. It is a bit vague in that he's saying it's "not the fault of FTTN technology",
Actually imo he pretty much said that the faults and issue's that those people who were not part of his PR survey who actually have bitched about their piss poor service � wasn't due to anything associated with NBN Co and its equipment to start with- and was pretty much washing his hands of the issue, putting the blame entirely into the RSP court.....
Will get the exact words when Hansard is published!
And then he Bill � had one of those � whoopsadaisy moments where he just realised he was blaming the RSP's entirely and then backtracked � stating, that they! needed to work collectively together to attempt to resolve why this is occurring!
Gee Bill! � one would have thought these reasons were now well known now in which we continue to add to the growing list! eg
FEXT, AXT, NEXT,
FM Transmitters
Environmental noise
Water Ingress
The gauge of the copper
High Open Faults
Oxidisation
AVC/CVC Congestion in Unicast services
The quality of the copper being used from the Node to your property boundary
The copper pair line length
The state of the copper inside your premise
Impulse noise
Industrial Noise
The state of existing shielding
MBORC (service Interruptions both planned and unplanned)
Bridged Taps both internal and external
Microwave ovens
Cordless Telephones
Spurious Radiation
Non compliant CPE equipment
Type of power back off being applied online both up and down"
Non compliant Central Splitters
Time of Day
Heat
Type of applications people are using (FTTN is a shared resource with limited spectrum)
And with some new ones with a yet to be determined and substantive impact from :-
SPGS" � Small Pair Gain Systems � which were originally designed to and the LNP will love this word "Economically" overcome copper shortfalls by combining several telephone services over a single pair!!!!!!
and
Telstra's "ISS" "incompatible" services" that actually "block" VDSL signals and have to be removed from the network!
wasn't due to anything associated with NBN Co and its equipment to start with- and was pretty much washing his hands of the issue, putting the blame entirely into the RSP court
And then he Bill � had one of those � whoopsadaisy moments where he just realised he was blaming the RSP's entirely and then backtracked � stating, that they! needed to work collectively together to attempt to resolve why this is occurring!
Translation � he just remembered that he's got the RSPs bent over a barrel on CVC pricing (most royally so !) and that if he pushes too hard about it being the RSPs fault they will point the finger of blame back at NBN for charging too high CVC that make it uneconomical for the RSP to provide a service of the quality that people expect (and had with ADSL) at the price point that people will pay. At the moment the RSP's are trying to bear this out as it's mostly a teething thing to get the number of customers and CVC balanced somewhat, but if NBN pushes too hard, then the RSPs will collectively spit the dummy on CVC pricing (as they rightly should). The RSP will also point the finger at the fact that they can't upgrade the CVC easily or frequently like they might if it was more in their own control.
Cant be bothered complaining about it, when I used to be on 20/1 ADSL2+.
That's NOT the point! If enough users complain early in the process there's hope for the rest of us down the track. If they think they can get away with glossy promos & blaming someone else, they will � & all Australia (including those on FW or SkyMuster) will be the poorer for it.
It's an election year which gives us a little more leverage . . .
which if he can be believed says that the problem is NOT related to lack of backhaul capacity from an individual node.
Depends how cynical you are, the fibre network links from the DSLAM to the AAS could be classed as not MTM related ie Not xDSL or HFC. I didn't trust Morrow's answer when I first heard it, it was more what was not said rather than what was said.
I look at it this way, Morrow defends the MTM policy at all costs normally. If he was able to say so, why wouldn't he have simply said 'of all the investigations we have undertaken on FTTN, it has been found that the FTTN backhaul has never been the problem'? Because he can't maybe?
Because he can't maybe?
NO "maybe" about it, if he doesnt answer confidently and promptly, he's lying or hiding the truth or avoiding firm answers so that he can lie about them and claim said answers weren't firm later down the track.
Anyone debating his answers as being even remotely genuine or optimistic, is a fool.
Has anyone been keeping track of which POIs people are connected to that suffer from severe congestion on FTTN?
Could it be NBN Co cutting corners on their portion of the network due to those areas not being connected to their transit network? ie They are using a temporary or interim solution and have provisioned to a cheap and nasty standard.
[edit] Maybe this might explain NBN Co not provisioning a 10G link/backhaul for FTTN from the beginning?
Can someone explain why its taking so long to roll out VDSL if all they are going to do is just change the hardware at the exchange?
Can someone explain why its taking so long to roll out VDSL if all they are going to do is just change the hardware at the exchange?
they are not putting VDSL in the exchanges
it takes time to build nodes
they are not putting VDSL in the exchanges
it takes time to build nodes
so they are upgrading nodes to be capable of VDSL and that is the reason its taking time
That's NOT the point! If enough users complain early in the process there's hope for the rest of us down the track.
You mean like this!
I'm writing because I believe I have story that is/will effect a lot of Australians and I'm hoping iTWire will help raise awareness to this issue and in doing so, help get this problem fixed.
When the service first went live, I thought I'd been wrong in decrying the axing of the FTTP approach in favour of the FTTN �solution�. I was seeing ping (latency) times of 12-13ms, download speeds of around 85 mbps and upload speeds of around 30 mbps, life was good,
Until I had the audacity to try and use my connection when I wanted i.e. when I was at home after work or on the weekend.
During peak times pings are on average sitting between 130 ms to 160 ms � with download speeds on average being less than ....4 mbps,.... while upload speeds hover around 10 mbps.
From my own observations peak times are from 4pm to 1am (9hrs) on weekdays and 9am to 1 am (16 hrs) on weekends!
Lots more here!
Whirlpool even gets a mention later on
so they are upgrading nodes to be capable of VDSL and that is the reason its taking time
no, they are installing nodes, no nodes exist.
Most copper goes back to the exchange (some goes to RIMS)
Most nodes are closer to the premises than nodes
Most nodes are closer to the premises than rims
There fixed that for you Dazed.
I fear that with the Summer over for the most part it will be another 10-12 months before that particular aspect of the nodes gets a thorough working trial.
next hot day I may, if in an English mood, go out in the midday sun with a sound meter and again at about 5pm after everything is well and truely heat saturated.
The fans were not anywhere near an obnoxious level
Fibre to the node is like getting gas lighting when you were promised electric lights
More like being promised flushing toilets connected to the sewer but being given a bucket.
Fibre to the node is like getting gas lighting when you were promised electric lights
It's the dialup network of the 21st century!
Some LNP NBN marketing phrases:
It's the dialup network of the 21st century!
Who needs fast internet, smoke signals are good enough!
It's all labors fault for starting this!
Near enough is good enough!
It's cheaper, faster and sooner*!
It's just as good*!
Trust us*!
*some conditions may apply
Just because every other country does something, doesn't mean we should. Would you jump off a bridge if every other country did?
next hot day I may, if in an English mood, go out in the midday sun with a sound meter
Maybe take a spray-bottle of water to cool them down....
Near enough is good enough!
It's cheaper, faster and sooner*!
It's just as good*!
Trust us*!
*some conditions may apply
I like that one �
Maybe take a spray-bottle of water to cool them down....
Actually I would be more concerned with latent heat problems due to humidity rather than sensible heat. So mists of water near the vents are not going to help.
However, It will be interesting to see if you can fry an egg on the top of the cabinets during a very hot day.
~mods please delete
Maybe take a spray-bottle of water to cool them down....
or an Irish Wolf Hound or 2 or 3 claiming territory
However, It will be interesting to see if you can fry an egg on the top of the cabinets during a very hot day.
most of the cabinets are in full sun and are currently extremely lightly loaded.
The ones I was at yesterday would have been around 50C going by feel and comparing them to hot water.
might also take a thermometer as well as a sound meter
I'm all connected now in the WOYY area.
Stats here � Sync @ ~61mbps down / 28mbps up � Guessing around 400-500m from the node.
http://i.imgur.com/tkJxbaG.jpg
Currently using a standard rj12 phone cable ~2meters long from the wall to the modem.
Is VDSL sensitive enough that I would actually see a difference by using a shorter cable? I'm thinking of comparing the stats with a <10cm cable and also using cat6 twisted cables but with RJ12 ends to see if there is a difference.
Anyone have experience with sync changes due to minor changes like that? (perfectly happy with the speed though tbh)
Currently using a standard rj12 phone cable ~2meters long from the wall to the modem.
Is VDSL sensitive enough that I would actually see a difference by using a shorter cable? I'm thinking of comparing the stats with a <10cm cable and also using cat6 twisted cables but with RJ12 ends to see if there is a difference.
Anyone have experience with sync changes due to minor changes like that? (perfectly happy with the speed though tbh)
shorter cables should give better speed, as most phone cables are usually "flat ribbon" not twisted pairs, maybe not much, but every bit counts.
Not sure if trying cat 5 or 6 with 2 pairs terminated into an RJ12 would make any real difference over a short phone cable.
Can't recall if the copper cores in phone are smaller than CAT 5 or CAT 6 so they may not terminate properly
Internode have confirmed they will be charging $299 (IIRC) for a "spare pair" installation
Yet another reason why Internode will probably lose me as a customer � and before anyone thinks I'm just being petty over $29; first the $270 NBN charge for a spare pair install is crap, it would cost Internode nothing-zip-nada in time/effort to accommodate their customer's wish to install on a spare pair and to then pile on with an additional $30 sucks. Be interesting to see what Exetel do.
$270 NBN charge
+ 10% GST = $297, which Internode are charging customers.
/forum-replies.cfm
http:
Subsequent FTTB/FTTN Install Fee: $297
Where you have an existing copper phone service at your premises but elect to have your NBN FTTB or FTTN service installed on a separate copper pair instead of converting an existing service, we may charge you a fee of $297 for installation. You may be charged additionally for any extra materials or labour required to complete the installation.
+ 10% GST = $297, which Internode are charging customers.
nbn� are charging the Telcos this
$270.00 + Labour Rate+ Materials
over and above Initial Standard Installation
to which Internode have to add 10% GST which comes to $297 PLUS any materials and extra labour
edit:- added nbn� document with list of charges from nbn� below
shorter cables should give better speed, as most phone cables are usually "flat ribbon" not twisted pairs, maybe not much, but every bit counts.
Not sure if trying cat 5 or 6 with 2 pairs terminated into an RJ12 would make any real difference over a short phone cable.
Can't recall if the copper cores in phone are smaller than CAT 5 or CAT 6 so they may not terminate properly
Up to a point yeah, gotta clarify to the person you're replying to that it's still hard limited by the internet speed. If you're already getting 100Mbps+ over the current cabling you have there is no point in changing it at the moment.
It also seems that nbn�'s installers are so inefficient that it takes them a minimum of two hours (or three hours if after initial installation) to install a splitter...
shorter cables should give better speed, as most phone cables are usually "flat ribbon" not twisted pairs, maybe not much, but every bit counts.
Not sure if trying cat 5 or 6 with 2 pairs terminated into an RJ12 would make any real difference over a short phone cable.
So I did a bunch of tests.
Normal phone cable 2m from wall to modem
CAT6 with 1 pair terminated into RJ12 2m from wall to modem
Normal phone cable 10cm from wall to modem
CAT6 with 1 pair terminated into RJ12 10cm from wall to modem
The difference is minimal. What I determined after averaging it all out was that the CAT6 with 1 pair terminated into RJ12 2m from wall to modem gave the same results as using a 10cm cable. No difference between 10cm normal phone cable vs CAT6 style. The worst stats are when using the normal phone cable 2m from wall.
So I'm leaving the 2m CAT6 with 1 pair and RJ12 ends in use.
Difference is :
0.4dB lower downstream attenuation
0.2dB lower upstream attenuation
~3mbit downstream sync speed increase
~300kbit upstream sync speed increase
Downstream FEC are still high. 1022781 after 50mins (with heavy connection use). No speed drops or latency spikes or packet loss happening though. I'm wondering if the connection will drop out once this counter hits a certain point though.
$270.00 + Labour Rate+ Materials
over and above Initial Standard Installation
to which Internode have to add 10% GST which comes to $297 PLUS any materials and extra labour
Screw 'em. I have a service (10 pair) existing with my entry socket wired for at least two lines so maybe I'll order a 2nd line which will then become an existing service and no $270 fee. And btw wtf is the open ended "+ Labour Rate + Materials" bs.
Telstra "Connection and Moving Fees":
Tech Visit � no cabling r'qd $75.50
(If cabling $179)
Telstra Budget line rental � 24.95 x 3 = $75
(need three months because there's a $100 "temporay" fee otherwise)
So either $150.50 or $254 vs $297. Not that Telstra deserves the money but NBN Co (Turnbull) surely don't.
Is VDSL sensitive enough that I would actually see a difference by using a shorter cable? I'm thinking of comparing the stats with a <10cm cable and also using cat6 twisted cables but with RJ12 ends to see if there is a difference.
The length of cable would make bugger all difference. It's the impedance change from twisted pair cat3/5 cable up to your wall socket having an abrupt change to a different impedance using flat ribbon cable that will reflect some signal back to the source. With any transmission line to get maximum signal to the destination you need to minimise any impedance mismatches (that typically occur at joints or interfaces or different cable). Maintaining the twist ratio of the cable at each joint is critical at ~17 MHz, but it makes almost no difference whatsoever at voice PSTN ~1kHz frequencies. The Telstra copper network was mostly engineered only for voice frequencies, never for 17+MHz bandwidth.
Only use flat ribbon cable for POTS voice services if you have a choice.
Screw 'em. I have a service (10 pair) existing with my entry socket wired for at least two lines so maybe I'll order a 2nd line which will then become an existing service and no $270 fee.
When NBN services are released for sale you can't order any old Telstra copper services, so you have to commit to this well in advance.
can any one on the FTTN with TPG or any ISP prefably around the 100/40 Mbps Speed pack send me some speed tests of what they are getting, I think I am 400 metres from a pillar. And the node is like 50 metres from the pillar.
There is some other thread that has people posting modem stats and distance from the node, suggest you go look for info there. I'm sure I could find it for you, but I'm too lazy to do a search for it right now.
So why is the $56 billion NBN taking so long to get connected to?
It's not the NBN, it's FTTN � they apparently have virtually no people working on connecting customers, and have to deal with the knackered copper that FTTN uses.
they apparently have virtually no people working on connecting customers
It's like they underestimated the amount of work involved again or something...
why two weeks?
an NBN technician needs to physically migrate your copper pair over to the node. future activations/migrations on your line will be able to be done instantly.
It's like they underestimated the amount of work involved again
Who could possibly have predicted that!
like they underestimated the amount of work involved again
I recall seeing something about the design (aka the planning of which premise connects to which node etc) phase suffering from poor quality reviews ... mainly it seems where it has been outsourced to the sub-continent no doubt to 'save money' i.e. deliver on the 'cheapa' meme.
The downside of that is the sacrificing of 'fasta' and 'bettera' goals.
Basically its a classic cluster where the KISS principle is nowhere to be seen and dumb decisions abound.
/Hanrahan.
25Mbps is fine if you believe the Strategic Review (it claims the average download speed requirement is only 19Mbits in 2023!) but I think most people by now understand this was very much low-balling it.
Definitely obsolete before it's finished then.
The current 3-year plan says that they're going to rollout about 4mil premises in the 3 year period (mostly MTM stuff). That will bring them to a (planned) toal of 5.6m premises by 2018. There are about 8m households in AU (from stats data, not sure whether households == premises), but either way that gives them at least another 3 years to complete the remainder, given that they are doing all the easy ones first and the other ones will likely take a bit longer. Add on an extra year or two for delays and the whole shebang could be done by around 2022, which is pretty much the year where the strategic review suggests the average download requirement will be 19mbps ?
'cheapa'
'fasta'
'bettera'
It's hard to know without completing both builds how it would/could have turned out (ie. how bad would the funding & rollout times for FTTH blown out), but it's looking likely that they will potentially fail on at least two of those three goals.
Also can any one on the FTTN with TPG or any ISP prefably around the 100/40 Mbps Speed pack send me some speed tests of what they are getting, I think I am 400 metres from a pillar. And the node is like 50 metres from the pillar.
Another thing, how has the HUAWEI HG658 VDSL Modem Router performed for people? Stable connection? I plan to use it only to connect to my good N600 TP-LINK Modem Router and use mostly the TP LINK modem for my Gigabit local LAN network. And just use the HUAWEI VDSL Modem just for the VDSL and VOIP connection as my partner wants a home phone and its part of TPG's unlimted plan, so why not I guess...
I am ~500meters from the pillar. I sync at ~64mbps down / ~30mbps up. Speedtests are basically matching the sync speeds minus a few mbit. I've tried offpeak and onpeak many times, always maxes, no congestion.
The HG658 is stable so far. Connection is been up for ~5 days straight at this point. Wifi is fine, however maybe my house is small. Can easily saturate the link speed over wifi.
Unfortunately you won't be able to bridge the Huawei and keep the VOIP ports. If you bridge it to your N600, the VOIP ports won't work on the Huawei afaik.
I myself am waiting for when TP-LINK release a VDSL modem/router that is verified as working with NBN FTTN. The Huawei is not terrible, but I would like some misc features in higher end products.
Hi guys,
I'm confused. :gasp:
So many questions ... that have no doubt been asked before.
1.Are we still in the 1,000 node trial window? http://www.nbnco.com.au/conte
2. I'm shortly getting a service connected in Gorokan NSW; part of the nbn� 1,000 node trial.
My key question is ... what actually happens 'at FTTN connection'?
I understand that the copper pair direct from the exchange is swung over at the pillar adjacent to the FTTN cabinet to the fibre side of things.
So is the FTTN connection effectively the same in concept as naked ADSL? ie the phone line loses dial tone and has a data only connection or is the exchange connection maintained just like ADSL? If I had a current phone service (which I don't), how is that dealt with? Is it 100% VoIP once the FTTN switch is triggered or same same via a filter.
What about the second pair that enters the house? Where does it terminate?
Under what circumstances does a Telstra tech needs to visit the home?
Is there a current nbn customer advisory (How To) for FTTN?
I wonder if the LNP will at least consider FTTdp
1.Are we still in the 1,000 node trial window?
no, we are in full deployment now
2. I'm shortly getting a service connected in Gorokan NSW; part of the nbn� 1,000 node trial.
My key question is ... what actually happens 'at FTTN connection'?
I understand that the copper pair direct from the exchange is swung over at the pillar adjacent to the FTTN cabinet to the fibre side of things.
yes this is basically what happens with a few provisos.
a) you may be able to migrate your current landline number to your chosen RSP if you want, but some cannot do it so will have to issue you a new local number. If you require it of course
b) is your second pair an active service?
i) if not, nothing will happen with it
ii) if yes, you will need to apply for that to be transferred as a second service and nbn� are now charging for this
on the scheduled day of cut over make sure you unplug your adsl modem as it may cause the port in the node to be locked out
I understand that the copper pair direct from the exchange is swung over at the pillar adjacent to the FTTN cabinet to the fibre side of things.
Your line will already go to the pillar. It will then be routed via the node rather than directly to the exchange as it does now.
So is the FTTN connection effectively the same in concept as naked ADSL? ie the phone line loses dial tone and has a data only connection or is the exchange connection maintained just like ADSL?
Yes basically. If you choose so you can keep your exchange based phone service for up to 18 months though.
If I had a current phone service (which I don't), how is that dealt with? Is it 100% VoIP once the FTTN switch is triggered or same same via a filter.
You basically need to port your landline number to VoIP at some stage. How that happens exactly I'm not so sure yet.
What about the second pair that enters the house? Where does it terminate?
It should terminate to the pillar as it does now. If it is connected to the exchange then I presume it will get disconnected at the end of the 18 month cutover period.
Under what circumstances does a Telstra tech needs to visit the home?
Never. I believe an NBN tech may visit though if there are issues getting the service running.
Never. I believe an NBN tech may visit though if there are issues getting the service running.
funny though, in some areas the "nbn" tech is driving a Telstra vehicle and wearing Telstra PPE
Thanks gents,
I was a bit hazy and lazy. :)
It is a minefield as you know.
Just sped read "How will voice be delivered over FTTN?" :gasp:
The mind does boggle that we could be this far down the FTTN roll-out track and still dealing with so many basic concept misunderstandings by RSP's and customers (and nbn it would appear).
Where is the nbn FTTN fact sheet documentation?
IMHO it is an extraordinarily unbelievable debacle. :(
So one other question, do we absolutely know if both a phone and VDSL service can be provisioned on pair 1?
Once the RIM gets turned off in about 16 months time, and the node being the only equipment running, will the vectoring then kick in and my sync speed will therefore increase?
It was my understanding that vectoring does not increase speed at all; it just cleans up interference so more packets are accepted.
This has the effect that you will get more throughput at the current speed with fewer dropped packets from interference.
Calculations become more complex as line length goes up however, so vectoring becomes ineffective after a certain distance. How long that is and how many Australians will be beyond that mark is unknown at this time.
So one other question, do we absolutely know if both a phone and VDSL service can be provisioned on pair 1?
nbn� do provide the option of a "voice continuity" option where the voice is still provided from the exchange, but this will only be the case for the 18 month transition period, after which voice will be by VOIP over NBN fixed line
Some providers may be offering to let you use "voice continuity" but some either are not or the CSRs do not have it in their sales scripts
and then there is the separate problem with number porting.............
Where is the nbn FTTN fact sheet documentation?
Sorry, that is Commercial in Confidence (at least till after the election :-)).
It was my understanding that vectoring does not increase speed at all; it just cleans up interference so more packets are accepted.
You've just negated yourself in the one line. Vectoring most certainly does increase the attainable rate and sync speed by co-ordinating the spectrum amongst the different line services so that each service has optimal signal to noise ratio on each tone. Without vectoring each service is isolated and the dynamic link adaptation to noise changes across the spectrum is much slower resulting in a more conservative sync speed.
Calculations become more complex as line length goes up however, so vectoring becomes ineffective after a certain distance. How long that is and how many Australians will be beyond that mark is unknown at this time.
Actually vectoring makes the disparity between short and long lines far more efficient. Longer lines can't utilise higher frequencies due to greater attenuation as you increase frequency, so vectoring can step in and co-ordinate the different line services so that longer lines are prioritised on the lower frequency tones and shorter lines are prioritised to use higher frequency tones first.
Actually vectoring makes the disparity between short and long lines far more efficient.
I suspect you mean 'more equitable'?
Longer lines can't utilise higher frequencies due to greater attenuation as you increase frequency, so vectoring can step in and co-ordinate the different line services so that longer lines are prioritised on the lower frequency tones and shorter lines are prioritised to use higher frequency tones first
Does it bond 'tones' on discrete lines so that that the lower frequencies, which naturally have lower signalling/carrier bit rates produce in aggregate what the closer high frequency services get?
Actually vectoring makes the disparity between short and long lines far more efficient. Longer lines can't utilise higher frequencies due to greater attenuation as you increase frequency, so vectoring can step in and co-ordinate the different line services so that longer lines are prioritised on the lower frequency tones and shorter lines are prioritised to use higher frequency tones first.
Awesome, it means I will have more to look forward to when the ADSL and POTS gets disconnected, and then everyone on my node will have vectoring doing a better job.
If you have an unregistered modem, you will basically not get support. So now you can't find out which modems are registered, and you can't get any support unless it is registered.
Someone should appeal that decision? ;)
1, thing I dont like is that if you dont have a copper lead in, you will need to still pay a installation fee of $299, but if your in a FTTP area, the line install is free. thats just messed up
And it isn't like ABS either Geo101
My analogy was supposed to high-light the fact that the calculations are performed at a high rate, such as is done by ABS, but as usual analogy's are often misinterpreted, or poor as you suggest.
It does this at a much faster rate and with better control than a driver could manage.
Knowing number of Telstra CTs I keep in touch with, copper customer access network, is lucky to have 60% of O-Pairs ok with ADSL2, gaud knows how VDSL will work better than current ADSL speeds ??
These are going to be a real problem where installed!
PGS" � Small Pair Gain Systems � which were originally designed to and the LNP will love this word "Economically" overcome copper shortfalls by combining several telephone services over a single pair!!!!!!
Not to mention Telstra's "ISS" "incompatible" services" that actually "block" VDSL signals and have to be removed from the network!
One thing however that is a bit of a worry is given the copper was originally designed to carry a 4k signal � just how much additional juice are they going to pump down the lines to enable VDSL to function and what impact does this have on the already 2 minutes to midnight copper in further degrading the service?
It does this at a much faster rate and with better control than a driver could manage.
https://en.wikipedia.org/wi
And my point is that is BS, a skilled driver can and will outperform ABS in nearly every situation. But ABS will outperform the average driver in most cases.
But now we are getting OT. And this is not the case with vectoring, the computational complexity is staggering.
1, thing I dont like is that if you dont have a copper lead in, you will need to still pay a installation fee of $299, but if your in a FTTP area, the line install is free. thats just messed up
You might even have a copper leadin, but if you don't have any active service on any copper pair then your first install will be classified a "subsequent" install and the $270ex (and RSP margin on top) will be charged.
That means if you want to port you landline number away to a VOIP provider so you don't lose it or have to have VOIP from your RSP you will probably have to pay the fee too.
This is because the number porting to VOIP provider will cancel the ADSL on your service, then when you order FTTN there is no active service (voice and ADSL codes not on line) and now you have a subsequent install.
It looks to me like NBN Co are trying to offset the extra costs due to difficulties with poor records and no active service to aid the technician. It is disingenuous from NBN Co IMHO.
1, thing I dont like is that if you dont have a copper lead in, you will need to still pay a installation fee of $299, but if your in a FTTP area, the line install is free. thats just messed up
Add on to the fact that Telstra charges $299 to run copper from the pit to your house, yet NBN have allowed this cost on fibre to blow out to $2500, which is pretty much the same process
And my point is that is BS, a skilled driver can and will outperform ABS in nearly every situation. But ABS will outperform the average driver in most cases
Not true, 4 channel ABS can brake individual wheels up to 15 times a second, I'm not aware of any human that do that, it's why they used it in F1 until all the fun toys got banned for the 94 season.
More validation of the future of copper under XG.FAST (well at least for FTTB and FTTdp) � don't quite think it works for some of the crazy FTTN line lengths.
Real world condition speeds of 8Gbps at 50m using standard copper, 11Gbps using bonded Cat6
Real world condition speeds of 8Gbps at 50m using standard copper
No, not real world conditions � lab conditions, using new unspecified diameter copper... certainly not Telstra standard copper!
8Gbps at 50m
How many nodes are we going to have to build before the people wasting our tax money realise it's more economical to run fibre, where we could get 1000Gbps on technology available now and already used in the real world.
How many nodes are we going to have to build
Just tell Mal that it's one node per premises and he can still call it FTTN.
Just tell Mal that it's one node per premises and he can still call it FTTN.
just call the FTTP ntd a micronode
it supplies up to 4 data connections
and up to 2 voice connections
simples
it supplies up to 4 data connections
and up to 2 voice connections
And we only needed one every second house too, as you easily just re-use all those existing cat6 leadins to connect to your neighbour's NTD :-).
And we only needed one every second house too, as you easily just re-use all those existing cat6 leadins to connect to your neighbour's NTD :-).
you could make it one per every 3 houses that would be a good cost cutting exercise, also with forgetting all those nice "cat 6" leadins, good old "cat 3" will do
:P
and by the way.
clear off, everyone should get their own.
Not true, 4 channel ABS can brake individual wheels up to 15 times a second, I'm not aware of any human that do that, it's why they used it in F1 until all the fun toys got banned for the 94 season.
Dude, I worked with the braking R&D guys at a well known company in AU for 12 months. You can talk about how often the data is sampled etc, but the value of that is measured by outcomes and not sampling rates.
I have seen first hand circumstances where ABS has been completely tricked, resulting in near zero braking force applied where even a regular driver would do better.
I think F1 is an unreasonable measure compared to real world outcomes, F1 is a much more limited in terms of the environment that systems need to operate within.
Real world condition speeds of 8Gbps at 50m using standard copper, 11Gbps using bonded Cat6
Funny how the article you linked literally says All trials were performed using prototype equipment from Bell Labs under laboratory conditions.
Are things really turning that bad for the MTM now that you have to resort to simply lying so you can keep thinking the copper dream is more than a fantasy?
not sampling rates
Yeah I think the "rates" at which vectoring works is far beyond ABS in hindsight to my analogy?
It's really the fact that processing (including digital signal processing, etc) is quite cheap these days, that is why DSL transmission vectoring is even half viable!!
I have seen first hand circumstances where ABS has been completely tricked
I'm guessing it wasn't a 4 channel, 4 sensor system?
Không có nhận xét nào:
Đăng nhận xét