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:

Author Topic: Stats and a Quiet line noise question on 17a..  (Read 6950 times)

popup

  • Just arrived
  • *
  • Posts: 6
Stats and a Quiet line noise question on 17a..
« on: April 05, 2013, 08:33:15 PM »

Since getting FTTC 80/20 3 weeks ago, I have been playing around with all sorts of variables to get some sort of optimised connection.  ISP is BT, My cabinet is a Huawei and my distance is unknown as the cable could take 3 different routes but could be as short as 400-500m and my installation modem was an ECI.  I initially connected at near max with the engineer stating that it should 'stay put there' but a day later it had dropped to 65, then 58 (which was my estimate funnily enough) and I'm still pleased as it's better than my old 24meg on ADSL but it would be interesting to know if there are any line fault-type issues from my stats anyway.

  After reading a few pages on matching modems with cabs, I purchased an unlocked Huawei 612 and, although the sync was slower on the Huawei and has never connected above the ECI, the difference is only about 2-4Mbit and I enjoy having stats to harvest.  I then read about the newer BLOB Firmware (A2pv6C035m.d22g) from http://huaweihg612hacking.wordpress.com/  and after reading about it's tested sync speed increase, I installed it and gained nearly 1Mbit in both attainable and Path. When I get bored, I will swap back to the ECI though I expect.

So, lol, today I got the Tandy 0.5M high quality RJ11 lead!..... and got a noticable boost to both attainable and actual on download which quite frankly amazed me.  One odd thing here though is that the quiet noise graph immediately got a little thicker with this cable which leads me to my first question.  Why?  Could it be that the sync increase had something to do with it?  Why would everything else point to a better solution but that or am I off the mark completely?

Here are the before and after stat graphs for psychological breakdown and any observations that you kind experts may have.  any help is greatly appreciated for this forum noobie :-[

Before cable!



After cable:


Over time:-


Before and after GIF

« Last Edit: April 05, 2013, 09:02:28 PM by popup »
Logged

popup

  • Just arrived
  • *
  • Posts: 6
Re: Stats and a Quiet line noise question on 17a..
« Reply #1 on: April 05, 2013, 10:33:57 PM »

I just checked a load of telephone numbers for stats in the area and worked out the route.  It is 520M from my property to the cab.
Logged

Bald_Eagle1

  • Helpful
  • Kitizen
  • *
  • Posts: 2721
Re: Stats and a Quiet line noise question on 17a..
« Reply #2 on: April 05, 2013, 11:15:54 PM »

Since getting FTTC 80/20 3 weeks ago, I have been playing around with all sorts of variables to get some sort of optimised connection.  ISP is BT, My cabinet is a Huawei and my distance is unknown as the cable could take 3 different routes but could be as short as 400-500m and my installation modem was an ECI.  I initially connected at near max with the engineer stating that it should 'stay put there' but a day later it had dropped to 65, then 58 (which was my estimate funnily enough) and I'm still pleased as it's better than my old 24meg on ADSL but it would be interesting to know if there are any line fault-type issues from my stats anyway.

It may be interesting to see the line_stats__FULL_MONTY_P-xxxxx montage of all the ongoing graphs.
You can post it here as an attachment instead of uploading it, if you wish.

Quote
After reading a few pages on matching modems with cabs, I purchased an unlocked Huawei 612 and, although the sync was slower on the Huawei and has never connected above the ECI, the difference is only about 2-4Mbit and I enjoy having stats to harvest.  I then read about the newer BLOB Firmware (A2pv6C035m.d22g) from http://huaweihg612hacking.wordpress.com/  and after reading about it's tested sync speed increase, I installed it and gained nearly 1Mbit in both attainable and Path. When I get bored, I will swap back to the ECI though I expect.

How did you upload the new BLOB firmware.
It does seem to need to be done via the LONG reset method to ensure it is fully installed.
My own testing (on a much slower connection) achieved a slightly higher DS sync speed & the blue 'Shared/Other' tones almost disappeared at the first (lower frequency) band & completely disappeared between the next tone bands (see attached).

Quote
So, lol, today I got the Tandy 0.5M high quality RJ11 lead!..... and got a noticable boost to both attainable and actual on download which quite frankly amazed me.  One odd thing here though is that the quiet noise graph immediately got a little thicker with this cable which leads me to my first question.  Why?  Could it be that the sync increase had something to do with it?  Why would everything else point to a better solution but that or am I off the mark completely?

Maybe it was a slightly 'noisier' connection at the time it resynced?

Quote
Here are the before and after stat graphs for psychological breakdown and any observations that you kind experts may have.  any help is greatly appreciated for this forum noobie :-[

Your Hlog graph has changed slightly at the highest frequency tones (looks like slightly improved attenuation hence the speed increase?).

But, the new cable may be slightly more susceptible to 'noise' interference.


Logged

popup

  • Just arrived
  • *
  • Posts: 6
Re: Stats and a Quiet line noise question on 17a..
« Reply #3 on: April 06, 2013, 12:16:12 AM »

Thanks for the reply.  I have 3 days of stats graphed so far, so not that much but my pc is on 24/7 so will keep logging (if that is how it even works).  I uploaded the newer firmware with the long reset method as the other way did not seem to work despite acting like it did (the correct latest version is now shown in the firmware menu of the modem).

  The new cable seems almost military grade compared to the 3 meter piece of string that came with the modem and is much better shielded and despite 7 resets in the last 3 days, has been one of the only significant jumps I have seen in stats (and the others were all after actually changing something first) so I might just leave it be :)  Rather oddly, I have noticed that a couple of syncs have given me a really unresponsive experience with 140ms pings to BBC despite the same sync speed and yet a quick reboot and the problem is gone!  You have to love uncertainty.


  It would be interesting to know what aspects of my stats are responsible for not allowing a higher attainable sync speed.  Mind you, the house behind mine is 30M from the Cabinet and he is fairly old so maybe there is a better tweak available after all :)
« Last Edit: April 06, 2013, 12:20:41 AM by popup »
Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Stats and a Quiet line noise question on 17a..
« Reply #4 on: April 06, 2013, 04:53:17 AM »

I uploaded the newer firmware with the long reset method as the other way did not seem to work despite acting like it did (the correct latest version is now shown in the firmware menu of the modem).

Just a quick note to mention that there are two different routes available to upgrade the firmware on a HG612 device. Each method requires the firmware to be in the correct format for the upgrade route chosen.

All versions of the currently available firmware are only valid for the upgrade process being initiated via the boot-loader. Any attempt to perform the upgrade from the GUI menu will silently fail.
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.

popup

  • Just arrived
  • *
  • Posts: 6
Re: Stats and a Quiet line noise question on 17a..
« Reply #5 on: April 06, 2013, 01:48:13 PM »

Here is my total 'Full Monty' as requested.

Logged

Bald_Eagle1

  • Helpful
  • Kitizen
  • *
  • Posts: 2721
Re: Stats and a Quiet line noise question on 17a..
« Reply #6 on: April 06, 2013, 02:56:35 PM »

Well, Your connection is still on fastpath, so quite a few error seconds & CRC errors are to be expected.

Other than the peaks around 11:00 on the 5th, it all looks pretty steady (error-wise).
Can you think of anything that may have caused those increased error rates?

Were the reasonably high number of resyncs in the graphed period all initiated by you, or are some of them DLM initiated?
There's a slight danger that DLM may start to see your connection as unstable & thus apply Interleaving etc. which will cause lower sync speeds as a 'corrective' measure to improve stability.

If Interleaving does get applied, error seconds will reduce & the other fairly low number of errors you are seeing will be replaced to a large extent by RSCorr/FEC type errors.

SNRM & Power levels don't really fluctuate, even when looking at each tone band separately & Line/Signal attenuation has remained steady.

All in all, it looks like you are achieving around the maximum possible for your connection's physical condition & hopefully it will remain that way for you.

Logged

popup

  • Just arrived
  • *
  • Posts: 6
Re: Stats and a Quiet line noise question on 17a..
« Reply #7 on: April 06, 2013, 03:10:20 PM »

Thanks very much for your input and taking the time to have a look.  All of the re-syncs were initiated by myself but I will not re-sync unless there is reason for it like a jump in attainable.  I have no idea about that chunky error spike.  Would heavy usage affect the graphs data? Also, can I ask that when you stated that my connection is still on Fastpath, do you mean it will not remain on this setting?

Many thanks again.
Logged

Bald_Eagle1

  • Helpful
  • Kitizen
  • *
  • Posts: 2721
Re: Stats and a Quiet line noise question on 17a..
« Reply #8 on: April 06, 2013, 03:48:59 PM »

Thanks very much for your input and taking the time to have a look.  All of the re-syncs were initiated by myself but I will not re-sync unless there is reason for it like a jump in attainable. 

If you do need (for whatever reason) to re-sync regularly, the advice is to leave it disconnected from the phone line for 20 minutes or so.
DLM should then see the disconnection as intentional rather than as a 'fault'

Quote
I have no idea about that chunky error spike.  Would heavy usage affect the graphs data?

I suppose that the more data being pumped through the connection, the higher the chance of some errors, but I'm not 100% sure.

Quote
Also, can I ask that when you stated that my connection is still on Fastpath, do you mean it will not remain on this setting?

I have seen some low speed & high speed connections remain on fastpath for months at a time, some that have Interleaving applied & then following a period of low errors removed again & some that have it applied within the first couple of days after installation.

Nobody seems to know the exact algorithm that makes DLM decide to apply Interleaving, Impulse Noise Protection & delay at various depths, but it will be some cumulative combination of the number of rapid disconnections/reconnections, various types of errors & MTBE (Mean Time Between Errors).

That could be a massive amount of errors/disconnections in a short period of time on an otherwise fairly error-free connection or regular but 'high' error counts over a longer period.

TBH, I have seen connections with far fewer errors than yours is exhibiting, so Interleaving MIGHT end up being applied, but perhaps at a low depth that won't have a massive negative impact on speeds.
Logged

popup

  • Just arrived
  • *
  • Posts: 6
Re: Stats and a Quiet line noise question on 17a..
« Reply #9 on: April 07, 2013, 05:31:38 PM »

Most interesting.  Thanks!

One more thing that has been bothering me is that I left the unlocked modem I purchased from Ebay as I found it and it just connected but when I look under basic settings, there are 2 connections as shown in the attached picture.  Can anyone tell if both are needed and what the TR069 connection is doing?  When I switch it off, nothing seems to change...

Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Stats and a Quiet line noise question on 17a..
« Reply #10 on: April 07, 2013, 06:31:58 PM »

With ptm1.101 disabled (i.e. the bridge between the xDSL port and the LAN1 port), you would have no Internet access. So that's a good thing to keep enabled.  :)

As to ptm1.301, that is used by Beattie's 'busy body' (the btagent) to call back to the Evil Empire's mother-ship and tell tales . . . It is also through this channel that any firmware upgrade may be pushed to the modem (without you knowing anything about it). Not necessarily a good thing . . .  :-X

Clearly we do not encourage users to disable the latter ptm but the choice is yours to take.  ::)
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.