Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[sevenlayermuddle]

Thanks for the replies to my 612kHz question, I need time to digest.

But meanwhile... good to hear from Ezzer again 

[Black Sheep]

Its not 612kHz thats the reference point but nearer 300 khz that actually IS the "reference" point. Its just that a typical radio dosn't go that far down, you can use the next harmonic which is at 600 or 150khz.

Any major REIN at that frequency will make a marked affect on dsl service. To a lesser degree further up the frequency range will do as well. Unless its at a harmonic (by harmonic I mean nice round intergers like double or half trebble/third,quad'/quarter. The effect depreciates a degree each harmonic further away from the key frequency your concerned with.

Its the same principle with musical notes, octives, keys etc.

Havnt been on this site in ages and the 1st thing I click on is REIN, typical. If you want nasty broadband try the U.S.A. I feel as if I'm getting the service as if I'm on the end of an 11km long line I often worked on in norfolk.

Ha ha ..... yeah, right !! I bet you've been looking in on this site for months just waiting for a REIN query to drop in ??!! 

To mirror Walters sentiments, welcome back Ezzer, sir. Good to have you on board again. Walter's alluded as to my previous alias, time enough has gone by for it to be known that I was once the evil tyrant 'razpag'.

How's the great US of A treating you mate ??

[Ezzer]

Hence Black Sheep 

Well, I've got Married. We had our own little ceremony with my parents looking on via skype. Even when I'm getting wed I cant get away from the net. Although we will still have a wedding proper in August as this one was more of a registry type wedding. In fact you can just go into the car registration office, get the paper, pop out sign it, get some witness to sign, nip back in and thats it. So we wil have a ceremony under our terms up in the mountains. Parents there in real life this time though. Just hoping they dont get altitude sickness. The rest of the family here suffers from it but I seem to thrive

Enjoying lovely warm weather. Temperatures up in the 90's in old money most days. It rained a couple of nights ago so the grass has done its annual trick of going from a straw colour to green as if a switch was flicked in the night. Although we're due for more snow on the weekend. Ever watched South Park ? well I was by there 2 weeks ago but I'm by north park. And thats sort of what it's like

Well I've only just got my Authority to work. The immigration proccess is long and expensive. half a mo' I think this reply belongs in chat,,,,

[Black Sheep]

Ha ha ...... yes Ezzer, hence BS.

Will follow you over on 'Chat'.

[JoshShep]

Thanks to Bald_Eagle1 I now have some stats to look at, only thing is I don't really understand them, take a look and see what you think please?

[JoshShep]

A few more.

[asbokid]

Thanks to Bald_Eagle1 I now have some stats to look at, only thing is I don't really understand them, take a look and see what you think please?

Baldie is the man for a proper answer, but my 2p's worth.. you have a fair bit of noise between DMTs #250 - #650.

It's tempting to blame cross-talk from adjacent pairs carrying ADSL2+ in your distribution cable bundle. However, ADSL2+ crosstalk should stop at DTM#512 (2.2MHz), the highest subcarrier used by ADSL2+. Yet the noise continues to DMT #650.  That suggests the noise is crosstalk from other VDSL2 pairs, perhaps aggravating existing ADSL2+ crosstalk.

That noise is causing a notch in the SNR in that downstream (DS1) band (DMT#32-DMT#859) and this is impacting bit loading. For some reason, even when the noise subsides around DMT#650, the remaining subcarriers in the band remain unused. Presumably a consequence of Lantiq's implementation of Dynamic Line Management? Hence the 30000Kbps cap on your throughput.  The combination of that noise and the DLM means you are losing around 8Mbps potential bandwidth.

cheers, a

[JoshShep]

Crosstalk, sounds like it could be possible. Would that explain why I sometimes get upload speeds as low as 0.8Mbps? There the same results I got on ADSL2.

What would you recommend me to do?

[Bald_Eagle1]

One thing has just dawned on me.

The scripts were written specifically for the Huawei HG612, connected to a Huawei DSLAM (well before I had even heard of the ECI modem & the slightly different plans used from an ECI DSLAM).

All the relevant data for the graphs as collected from the HG612 is formatted into just 2 columns.

i.e. Tone No. in column 1 & data in column 2.

To separate DS & US data, the scripts use the tone band plans as discovered from a Huawei DSLAM:-

Discovery Phase (Initial) Band Plan
US: (0,95) (868,1207) (1972,2783)
DS: (32,859) (1216,1963) (2792,3939)


However, an ECI DSLAM uses, & the HG612 discovers these tone band plans:-

Discovery Phase (Initial) Band Plan
US: (0,95) (880,1195) (1984,2771)
DS: (32,859) (1216,1959) (2792,4083)

The differences between the ECI & Huawei band plans may not be huge, but they MAY be causing some slight graphing discrepanicies/inaccuracies when a HG612 is connected to an ECI DLSAM.

I don't really have time to adjust the scripts at the moment, but I am aware that BlackEagle is working with them to possibly create .exe rather than .bat files.

FYI, BlackEagle was also originally provided with an ECI modem for his connection, so may well have already spotted & be dealing with the slightly differing band plans.

[asbokid]

Crosstalk, sounds like it could be possible. Would that explain why I sometimes get upload speeds as low as 0.8Mbps? They're the same results I got on ADSL2.

It would make sense that noise from crosstalk is also affecting the upstream (US) bands, especially the higher frequency US1 band. Unfortunately, we cannot gauge the extent of that noise in the US1 band.  The modem cannot calculate noise levels in upstream bands - they have to be measured at the DSLAM and we don't have access to that data.

What would you recommend me to do?

Not a lot you can do. Crosstalk is perhaps the greatest limitation of xDSL.

In terms of VDSL2 it's relatively early days.

Openreach/ECI/Lantiq will be remotely monitoring statistics at the VTU-R. Hence the usefulness of TR-069 remote management. Crosstalk measurements taken in the field will be collated and overlaid onto theoretical models. These will be used in the refinement of the DLM algorithms. And as the bigger picture unfolds, those updated algorithms will be distributed as remote firmware upgrades to the DSLAMs.

If that were my line, I would be wondering why a sizeable chunk of utilisable tones in DS1 band are not being used (DMT#650 to DMT#859)..

It may be that Openreach/ECI/Lantiq are trying their best to juggle finite resources  - attempting to fairly distribute the aggregate 'Shannon Capacity' of all the pairs in a bundle, while mitigating inter-pair crosstalk.   To this end, Openreach/ECI/Lantiq will be able to see an overview of the line data for all xDSL pairs served by a DSLAM, all the xDSL pairs in a distribution bundle, etc..

No answers really.  Although in the great scheme of things, 30,000kbps is still pretty useable.

cheers, a

[Bald_Eagle1]

The scripts were written specifically for the Huawei HG612, connected to a Huawei DSLAM (well before I had even heard of the ECI modem & the slightly different plans used from an ECI DSLAM).

All the relevant data for the graphs as collected from the HG612 is formatted into just 2 columns.

i.e. Tone No. in column 1 & data in column 2.

To separate DS & US data, the scripts use the tone band plans as discovered from a Huawei DSLAM:-

Discovery Phase (Initial) Band Plan
US: (0,95) (868,1207) (1972,2783)
DS: (32,859) (1216,1963) (2792,3939)


However, an ECI DSLAM uses, & the HG612 discovers these tone band plans:-

Discovery Phase (Initial) Band Plan
US: (0,95) (880,1195) (1984,2771)
DS: (32,859) (1216,1959) (2792,4083)

The differences between the ECI & Huawei band plans may not be huge, but they MAY be causing some slight graphing discrepanicies/inaccuracies when a HG612 is connected to an ECI DLSAM.

I don't really have time to adjust the scripts at the moment, but I am aware that BlackEagle is working with them to possibly create .exe rather than .bat files.

FYI, BlackEagle was also originally provided with an ECI modem for his connection, so may well have already spotted & be dealing with the slightly differing band plans.

CORRECTION:-

When I first wrote the scripts, I did indeed use specific Huawei band plan tone numbers.

I must have realised that these would be incorrect when connected to an ECI cabinet around the time we were first made aware of the ECI modem/cabinet combination.

It may even have been JoshShep that made us aware of the ECI combination via the TBB forum?

The scripts actually do now use the band plans as discovered at Discovery Phase, so the data collected via a HG612 modem, even when connected to an ECI cabinet should split up & plot the data correctly after all.

Aplologies for the previous misinformation folks.
I'm not getting any younger & my memory must be playing tricks on me 

[burakkucat]

Aplologies for the previous misinformation folks.
I'm not getting any younger & my memory must be playing tricks on me 

   D'oh!

[JoshShep]

These errors anything to worry about? I noticed allot of them clock up sometime during the night or early morning.

[JoshShep]

.

[JoshShep]

The errors are just clocking up.