Kitz ADSL Broadband Information
adsl spacer  
Support this site
Home Broadband ISPs Tech Routers Wiki Forum
 
     
   Compare ISP   Rate your ISP
   Glossary   Glossary
 
Please login or register.

Login with username, password and session length
Advanced search  

News:

Pages: 1 ... 3 4 [5] 6 7 ... 44

Author Topic: FTTC - Distance from the Cabinet - Realistic speed estimate  (Read 254031 times)

Bald_Eagle1

  • Helpful
  • Kitizen
  • *
  • Posts: 2721
Re: FTTC - Distance from the Cabinet - Realistic speed estimate
« Reply #60 on: August 09, 2011, 10:33:24 PM »

Suggestion -- When the engineer plugs in his/her JDSU, EXFO or what-not, why not have your camera to hand and ask if you could photograph the relevant screens, so that you have a record of the current statistics?

Cheers burakkucat,

That had not even crossed my mind.
Sometimes the simplest ideas are the best ones.

Paul.
Logged

waltergmw

  • Kitizen
  • ****
  • Posts: 2776
Re: FTTC - Distance from the Cabinet - Realistic speed estimate
« Reply #61 on: August 10, 2011, 12:00:00 AM »

Hi Paul,

I have used my mobile to take pictures of the JDSU pages which can help. HOWEVER I suggest it's important that you get the engineer to unlock the actual VDSL2 modem that you use. The reason I say that is that on standard ADSL1 lines the JDSU always seems to connect at lower sync speeds than the 2700 HGV Business Hub. I must stress that these observations are on very long and noisy lines but I have known I think 3 now where the JDSU hasn't managed to sync even after 20 minutes or so but the 2700 might take say 5 minutes but always manages the minimum speed sync.

Unless you can tell me otherwise, I think the jury is still out on this latter replacement modem. I.e. it has never been as fast as the original "furnace" one ? I don't suppose we'll be told but I would be very keen to know what design modifications were made to reduce the overheating. If those changes have reduced the potential speed, then all subsequent swap-outs will probably behave similarly.

If this were a little cheaper I'd be tempted to get the Zyxel modem asbokid described here

http://forum.kitz.co.uk/index.php/topic,9707.msg196200.html#msg196200


Kind regards,
Walter
Logged

Bald_Eagle1

  • Helpful
  • Kitizen
  • *
  • Posts: 2721
Re: FTTC - Distance from the Cabinet - Realistic speed estimate
« Reply #62 on: August 10, 2011, 12:26:32 AM »

Hi Walter,

The "furnace" modem returned speed test results as low as 7.97 Mb down & 0.66 Mb up.
My profile was getting lower & lower, no doubt due to the disconnections

When my profile was reset by BT at 24 Mb Thursday 4th August, the new modem was connected, & for a few hours I achieved downloads up to 22.7 Mb & uploads at 1.67 Mb.

By Friday morning, downloads had dropped to around 14.5 Mb with uploads still at 1.66 Mb.
I have suffered no disconnections at all since Friday & although I haven't saved the results into a spreadsheet yet, they are still around 14.5 down & 1.66 up.

Some dated results attached.

BTW, both modems are firmware version SP10 (I believe that is the most recent version until BT's version 'B' modem is released, whatever it happens to be).

To be honest, because I still had some disconnections & crackly phone line after the modem had been replaced, I'm not not now 100% convinced it was an overheating modem issue at all, but rather an intermittent line issue all along.

Hopefully, all will be revealed in the morning.

Paul.



[attachment deleted by admin]
« Last Edit: August 10, 2011, 12:54:21 AM by Bald_Eagle1 »
Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: FTTC - Distance from the Cabinet - Realistic speed estimate
« Reply #63 on: August 10, 2011, 01:04:13 AM »

If this were a little cheaper I'd be tempted to get the Zyxel modem asbokid described here

http://forum.kitz.co.uk/index.php/topic,9707.msg196200.html#msg196200

@ Walter -- If only Arthur Daley was still in the Import/Export business, he could have a "nice little earner" by importing the Dare Tech DB120-B2+ direct from China . . .  ::)
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

razpag

  • Reg Member
  • ***
  • Posts: 622
Re: FTTC - Distance from the Cabinet - Realistic speed estimate
« Reply #64 on: August 10, 2011, 11:09:35 AM »

Walter, not sure 100% what you mean buy, ' get the engineer to unlock the VDSL modem', but if it's anything more than taking it out of a box and plugging it in, we don't do ANYTHING more with the VDSL modem.  ;D
Logged

waltergmw

  • Kitizen
  • ****
  • Posts: 2776
Re: FTTC - Distance from the Cabinet - Realistic speed estimate
« Reply #65 on: August 10, 2011, 11:23:38 AM »

Hi RP,

This is perhaps the difficulty as (I believe) the EU can't observe the modem statistics as we can with most ADSL 1 & 2 modems. However there must be some department within the BT Group that is able to observe the modem statistics. I wonder if you'd be allowed to enquire who might have such access and whether the end users could also please be given such access. At present there might be a suspicion that parts of BT are trying to hide the performance of the D side lines.

@ Azzaka, As a CP are your able to access the standard BT VDSL2 modems ?

Kind regards,
Walter
Logged

Bald_Eagle1

  • Helpful
  • Kitizen
  • *
  • Posts: 2721
Re: FTTC - Distance from the Cabinet - Realistic speed estimate
« Reply #66 on: August 10, 2011, 11:31:24 AM »

Mornin' All,

Well, the "special" engineer turned up this morning & everything went really well.........NOT

He was carrying another shiny new Huawei Echolife HG612 modem under his arm as he walked down to the house.
My suspicions regarding his intentions for today's visit were immediately raised.

After the usual pleasantries were over. he scratched his head for a while, looking at the shielded, twisted pair, high quality, high speed "data" cable going into the back of the FTTC socket next to my desk. Eventually he asked "Where is your master socket then, & where did that cable come from".

I told him it was my own cable that the original FTTC installation engineer had hard-wired from the upstairs socket when he relocated my master socket from upstairs to the one he was now looking at. I showed him the upstairs socket that now just has a single telephone only outlet.
 
I aked him if he was aware that the modem had only been replaced last Thursday & I told him that I had been assured that today's visit had been arranged for an investigation into the intermittent disconnections & crackly phone line, & to especially investigate high resistance issues.
He responded with "No, I'm here to see that you are getting at least a 15Mb connection, because that's all we have to guarantee, & to replace the modem.

He would not really listen to a summary of the ongoing issues, or look at my "evidence" of decent speeds.
His only comment when I told him I had recently suffered a complete lack of phone & FTTC service for 4 days was "Ah, so the line's been fixed then".
He certainly did not accept that I still have a suspected intermittent "fault".

Anyway, he unplugged the 6 day old modem & plugged his JDSU HST 3000 meter into the master socket.

I asked him if he minded if I photographed the various screen displays because that was the only way I could obtain a record of the line stats due to the BT modems being locked against end-user access.
After he looked at me as though I was from Mars, he shrugged his shoulders & said, "Well, I suppose so".

After a while he said "There you go, your Max Rate is 22 Meg". See the attached photo.

I asked if he would be requesting a profile/line reset to get it back up to around 30 Mb.
The answer was that no, he wouldn't be doing that because 22 Meg is all my line can provide.

I asked if he would be doing any "special" high resistance or other line testing to investigate the intermittent "faults".
The answer was again no as the line tests OK (LTOK).
When I asked if the display actually showed any high resistance or other issues, again he said no, it is all O.K., it would either pass or fail.
As I had misplaced my reading glasses (what an absolute idiot!!!!) I couldn't see the display & will have to just rely on the single (not very clear) photo which again, I couldn't see clearly on my camera.
I asked him what the attenuation level was then. The response was that it was O.K.
I asked him if it was high.
He said it was O.K.
I asked if it showed a value.
He said it was O.K. & that it would simply either pass or fail.
I asked about the low max rate of 22 Mb.
He said it was OK.
I asked why it wasn't 30+ Mb as on the day of installation.
He said it was O.K. & that's all I can expect my line to provide.
I asked if he knew the distance from the cabinet.
He said it wasn't too far.

Seems to be a bit of a pattern devoloping here.

I asked if the distance (either physical or potentially mis-represented distance due to potential high resistance) could be the cause of the now low max rate of 22Mb, or could it be likely that the regular disconnections etc. had forced a temporarily lower rate via BT's automated DLM at the cabinet.
He said it was O.K. & that's all my line can provide. He opened up a bit then & said it is probably because of the aluminium in the cables.
When asked if that's the case then, how come I solidly achieved higher speeds for the first couple of weeks, the response was that that's all my line can provide & I would have got the full 40 Mb on installation that has now been trained down to 22Mb.
I asked if he knew that the cables were aluminium, or just suspected it.
He said they are probably aluminium.
Another shrug of the shoulders came when I said the original installation engineer had told me it had NEVER achieved the full 40 Mb, but it was around 35 Mb when first installed. He also commented that it MUST have been at 40Mb for a while.

He did add however that he didn't fully undertand these things & that anyway, the British Army has now been drafted in to install FTTC, not just the cabling, but the installations in the home too.
He further added that he only gets the crappy jobs now.

Just before he left, the engineer told me that today's new modem is the new improved version 2B that will sort everything (phot attached).

1) Did I actually get the right sort of engineer this morning?
2) Is the very limited information that he gave me actually correct?
3) Has any of what the engineer done this morning actually addressed my potential intermittent line "fault"?
4) I have asked Plusnet to provide both BT & me with proof that my line was solidly achieving 32Mb for the first couple of weeks & surely it should now have "trained" itself to that level but for the intermittent disconnections/crackly line  "fault".
5) What the hell, if anything, can I do to get this sorted once & for all. Virgin cable is not available in my area. BT provide the infrastructure, so any ISP would be in the same position as Plusnet.
I was promised this would be escalated if not resolved. But how?

The saga continues.


Paul.

[attachment deleted by admin]
« Last Edit: August 10, 2011, 11:37:10 AM by Bald_Eagle1 »
Logged

Bald_Eagle1

  • Helpful
  • Kitizen
  • *
  • Posts: 2721
Re: FTTC - Distance from the Cabinet - Realistic speed estimate
« Reply #67 on: August 10, 2011, 12:03:54 PM »

@ razpag

The photo from the JDSU is not too clear, but what can be deduced from the values:-

                             UP                    DOWN                 
Actual Rate             1996K                22648K
Max Rate                5212K                22662K
Capacity                   38%                  100%
SNR                      12.9 dB                 5.9 dB
Line Atten              52.9 dB               30.2 dB

a) Does this "appear" to be a very long line? If so, could you give a rough estimate in metres?
The Kitz calculator suggests 2.2km downstream (30.2 dB Line Atten) & 3.8km upstream (52.9 dB Line Atten)
This is for ADSL though - so I don't know if it is the right estimator & how relevant that would be for FTTC anyway.

In reality I live around 0.4 miles (roughly 0.65 km) from the cabinet by road, not in a straight line (I do know the line detours a bit though).

b) Is 22662K indeed the correct Max Rate?
 
c) Is it usual to see such a difference between the SNR & Line Atten for upstream & downstream?
I now recall that I did have an issue about 6 months ago on ADSL when my upstream dropped to around 96K for a few days, even though my downstream was at its highest level of 1250K. When the upstream reverted to its usual level (just a guess at 348K, my downstream went back down to 1000K).

d) Could this morning's engineer have actually told me the reported line length from the cabinet?

e) We didn't even get round to discussing Interleaving. Could the engineer have told me whether it was on or not?

f) Even if interleaving is set at HIGH (as Plusnet told me it was a while ago), would it affect the Max Rate, the Actual Rate, or both?

g) Do any of the values suggest there is an actual line "fault"?

h) Are you aware of any improvements gained from the version 2B modem?

i) Am I just asking for too much i.e. do I now have the maximum stable speed that my line will ever provide & is it time for me to put up & shut up?

P.S. My upload speeds are capped at 2Mb with the "Value Fibre" package that I am on, so I am not querying upload speeds at all.

Paul.
« Last Edit: August 10, 2011, 01:25:43 PM by Bald_Eagle1 »
Logged

waltergmw

  • Kitizen
  • ****
  • Posts: 2776
Re: FTTC - Distance from the Cabinet - Realistic speed estimate
« Reply #68 on: August 10, 2011, 12:59:56 PM »

@ Paul,

Comparing the JDSU VDSL figure to those of ADSL the attenuations "look" transposed. Note also that the attenuation seems only to be measured properly from the far end. I.e. the downstream observation from your home should be reasonably accurate whereas the upstream is usually an approximation of about half the downstream. My guess is that JDSU Tests are done with the entire line connected (as your service is a SMPF one - Shared Metallic Path Facility. I.e. your telephone service is still provided over the whole length of line back to the exchange. (Note FMPF {Full MPF}, which is available to some independent Communication Providers, has the D side line isolated at the PCP so any telephony services provided by them use a form of VOIP telephony.)

You might ask the next visiting engineer if he can do a TDR test if he's allowed to. http://en.wikipedia.org/wiki/Time-domain_reflectometer

However we have had discussions before here suggesting that the JDSU instrument is possibly not the best device for this test.

Here's a JDSU example on a long line as you can see.

Kind regards,
Walter

« Last Edit: August 10, 2011, 01:49:38 PM by waltergmw »
Logged

Bald_Eagle1

  • Helpful
  • Kitizen
  • *
  • Posts: 2721
Re: FTTC - Distance from the Cabinet - Realistic speed estimate
« Reply #69 on: August 10, 2011, 01:12:28 PM »

Hi Walter,

The "special" engineer was supposed carry out a CIDT test this morning that identifies the "faults" such as high resistance that basic tests can't identify. (please see the attached .pdf, although I suspect that personally you will have read this previously.

I don't know whether his JDSU meter did that this morning or not. Any idea razpag?
You may have gathered that the engineer wasn't particularly helpful or forthcoming with his responses this morning.

Paul.

[attachment deleted by admin]
Logged

Bald_Eagle1

  • Helpful
  • Kitizen
  • *
  • Posts: 2721
Re: FTTC - Distance from the Cabinet - Realistic speed estimate
« Reply #70 on: August 10, 2011, 01:17:22 PM »


Here's a JDSU example on a long line as you can see.


Hi Walter,

Did you mean to attach an example from another JDSU, or were you talking about my JDSU example?

Paul.
Logged

waltergmw

  • Kitizen
  • ****
  • Posts: 2776
Re: FTTC - Distance from the Cabinet - Realistic speed estimate
« Reply #71 on: August 10, 2011, 01:56:25 PM »

Hi Paul,

It was supposed to be another one which was too large but I'll have another go.

K R
Walter



[attachment deleted by admin]
Logged

razpag

  • Reg Member
  • ***
  • Posts: 622
Re: FTTC - Distance from the Cabinet - Realistic speed estimate
« Reply #72 on: August 10, 2011, 06:44:25 PM »

Hi Walter,

The "special" engineer was supposed carry out a CIDT test this morning that identifies the "faults" such as high resistance that basic tests can't identify. (please see the attached .pdf, although I suspect that personally you will have read this previously.

I don't know whether his JDSU meter did that this morning or not. Any idea razpag?
You may have gathered that the engineer wasn't particularly helpful or forthcoming with his responses this morning.

Paul.

Hi BE

Firstly, I'm not going to comment on the engineer other than, it sounds like he's basically doing what's been asked of him via our training courses. I'll have to leave that one there i'm afraid.

Now then, the JDSU info you've posted up does indeed 'show' a good test, and that you are receiving the 'max' expected for your line. What has really made my brain hurt, is Walters comments about the attenuation figure being the full length of the 'copper pair' from exchange to premises. Now I've racked this over in my head for about an hour, and I can't see how the attenuation calculator within the JDSU can measure through the DSLAM, so it's my guess (and that is all it is), that the attenuation reading is from the Cab to the premises. If my guess is right, then you have a helluva high reading if you are positive you only live 0.5 miles away from the Cab ?? This kind of integer would also fit in with a large drop in synch speed, from 40Meg down to what you are receiving presently ??

This truly is puzzling, as you are certain you've had higher speeds of 35Meg which would also fit in with a relatively short distance drop from 40Meg over 0.5 miles ?? It's bizarre. The comment about everyone starts at 40Meg and the DLM drops accordingly is not right either. You get what you get and DLM does minor tweaks, not 12/13 Meg drops from 35 to 22 meg !!!!

The only rationale I can produce, is that you still have an HR fault, this is giving the DLM the impression you are 30.2dB from the exchange and as such, it has given you the synch speed it determines is suitable for this distance and quality of line. The tests carried out by the JDSU that you have shown will not 'see' a HR. Only the TDR function within the JDSU can do that, when testing back to the exchange with voltage applied. Again, i'm not 100% sure about the attenuation figure as to whether it's the whole of the 'pair' or just from premises to Cab. I've not taken much notice when on installation tasks, and as the engineer stated, they've now employed ex-services personnel and the apprentices to cover ALL FTTC/FTTP/Managed Install work. We experienced, multi-skilled engineers don't get a look in these days. It's wrong on so many levels but that's BT politics for you.

Sorry I can't be of much help mate, I really wish I could. :'( 
Logged

waltergmw

  • Kitizen
  • ****
  • Posts: 2776
Re: FTTC - Distance from the Cabinet - Realistic speed estimate
« Reply #73 on: August 10, 2011, 09:10:08 PM »

Hi RP,

I too am working in the dark just on the theoretical approach. As Paul's BT recorded line distance is 5283 m to the exchange and the attenuation is high I'd assumed the opposite that it was measuring the whole copper pair, but possibly with some strange effect due to the DSLAM "tee off" being in the FTTC. It also seems strange that the attenuation figures seem abnormal and perhaps reversed ?

Paul's D side pair is under 600 m in length but neither of us know whether aluminium is involved.

As you haven't mentioned it, I assume that you are not instructed to disconnect the E side pair in the PCP before doing FTTC JDSU tests ?

Kind regards,
Walter
Logged

Bald_Eagle1

  • Helpful
  • Kitizen
  • *
  • Posts: 2721
Re: FTTC - Distance from the Cabinet - Realistic speed estimate
« Reply #74 on: August 10, 2011, 09:15:35 PM »

@ razpag

The way I understand matters is as follows (please correct me wherever I have got it wrong):-

Via ADSL, the DSLAM is located at the exchange.
Via FTTC, the DSLAM is located at the FTTC cabinet, effectively bringing the exchange a lot closer to the end users (for broadband purposes).
The broadband speed at the FTTC cabinet is currently restricted to 40 Mb (for now), & will always be a minimum of 40 Mb
The telephone service still comes via wires (copper or ally, or a mixture) all the way from the exchange to the old cabinet
The broadband fibre & telephone wires only join up at the FTTC cabinet
The broadband & telephone services are then delivered to end users from the FTTC cabinet via wires
The wires between the cabinet & the end user can often take quite a detour, increasing the line length that actually affects broadband speeds substantially in some cases
The overall length of the wires from the cabinet to the end user has a distinct effect upon Line Attenuation
Line attenuation can additionally increase due to the conductive material the wire is made from e.g. copper conducts much better than aluminium, the thickness (poundage) of the wire, the quality of any joints, corrosion, water ingress, temperature.
Line Attenuation directly affects the broadband speed at the end user's master socket i.e. the longer the line, the higher the attenuation & the lower the broadband speed at the master socket
End users very close to the FTTC cabinet (within 100m or so) can expect the full 40 Mb, less any general overheads. The usual full speed is around 39 Mb at the master socket.
Broadband speeds graually decrease over a line length of around 700m - 800 m, then rapidly decrease thereon after as the Line Attenuation increases due to the general resistance within the wires.
At a distance from the cabinet of say 750 m, assuming the wires are in good condition, a reasonable line speed of 35 Mb can be expected. Allowing for general overheads, an achievable & sustainable download speed of 33 Mb can be expected, suject to contention.
Signal to Noise Ratio also affects line speed. A high SNR = good, a low SNR = bad.
Electrical noise can be introduced into the line from various sources e.g. induction from street lights, nearby power cables, radio frequency transmissions, fridge motors etc.
SNR is usually at its best during the daytime. In the evenings, SNR can lower substantially
Line speeds & therefore upload/download speeds are continually & automatically adjusted by Dynamic Line Management (DLM) to suit prevailing line conditions.
Following initial installation, the DLM will allow the highest speed possible to the master socket, constantly checking for things such an increase in attenuation, a lowering of SNR (the more interference, the lower the SNR), regular disconnections for whatever cause.
To keep the line working, the DLM will lower the line speed to restrict errors & maintain stability.
Over the first 10 days or so, the line speed usually trains & stabilises at a sustainable value, although a succession of incidents such as repeated disconnections can restart the line training period at any given time.
Line speeds are lowered quite quickly by the DLM, & via FTTC can take around 10 days or so to recover to the previous sustainable speeds.

Now, in my own circumstances, for the first couple of weeks, I solidly achieved download speeds of 32-33 Mb, against the installing engineer's confirmation of a line speed (Maximum Rate) of 35(ish) Mb.
Following a number of disconnections & a complete loss of phone & broadband service for 4 days, the line speeds eventually reduced to around 8.5 Mb.
Every now & then the line would pick up speed, but would be reduced a short time later by another succession of disconnections.

Having sustained download speeds of around 32-33 Mb for a couple of weeks, my line can be considered to have completed its training period, & given another suitable training period with very few disconnections, high resistance, other interference etc. may be expected to gradually regain those sort of speeds.
this may take another 10 days or so.

It is my understanding that DLM adjusts line speeds "on the fly" & does not in itself cause disconnections.
I assume therefore that the disconnections have been intermittently caused by other unknown events.
There does not appear to be a pattern to these events.
The weather can be warm & dry, or wet, cold & windy.
It could be during the daytime or at evenings, or even the early hours of the morning.
A number of days can pass with no disconnections, sometimes spanning whole weekends.

My line currently does appear to have quite a high attenuation in comparison to its likely overall physical length. This suggests high resistance, but of an intermittent nature, otherwise, surely I could never have achieved the initial high download speeds.

It is however impossible to monitor line/connection statistics via the BT modem.
The cynics amongst us may well believe this was an intentional move by BT, purely with the aim of concealing the facts that FTTC isn't quite ready for the market as advertised & is only really fit for purpose for anyone living within 100m or so of the FTTC cabinets.

Maybe the new & improved version 2B modem will be nore resilient to these intermittent events.
As the "special" engineer didn't test the quality of the line in any way, the question regarding general line quality cannot be answered at this stage.

In your opinion, is this matter really worth pursuing? I won't give up if there is a reasonable chance of a permanent reolution, but I don't wish to keep flogging a dead horse.

In many respects I should be happy with my current download speeds of around 22 Mb (it is after all a lot more than many users can achieve), but in reality it is only just over half the speed that FTTC promises.

Sorry for the length & rambling nature of this post. Do you have any experienced comments or advice for me?


Paul

Edit: Have BT or Plusnet manually introduced a Max Rate cap onto my line, that could/should now be removed to at least test the line at full speed again?
The engineer said he didn't test the speed as it leaves the DLM at the FTTC cabinet as that wasn't his job.
Would this be a basic test that "special" engineers should do as part of an investigation into poor speeds at the end users master socket? I know it would be if I was investigating for example poor water pressure at a tap. I would first check that the pressure at the source (mains) was O.K., especially if it was a known fact that it had been good pressure at the tap at one time.
« Last Edit: August 10, 2011, 09:34:36 PM by Bald_Eagle1 »
Logged
Pages: 1 ... 3 4 [5] 6 7 ... 44
 

anything