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 ... 4 5 [6] 7

Author Topic: Findings with a new connection  (Read 26516 times)

burakkucat

  • Global Moderator
  • Senior Kitizen
  • *
  • Posts: 31993
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Findings with a new connection
« Reply #75 on: March 26, 2014, 10:09:54 PM »

b*cat feels somewhat uncomfortable when reading posts that make reference to "blob" . . . especially as the phrase "firmware image" should really be used. Most of the time, that inexactitude is used in the Thinkbroadband forum but now it seems to be making an appearance here.  :(

The source code of the various firmware images for the Huawei EchoLife HG612 can be obtained. There is just one component of those images for which the source code is not available -- the Broadcom driver. That driver is only made available as a binary blob.

It was with the original unlocked Beattie firmware image that Asbokid performed a binary blob transplant. And it was subsequent to that event that the incorrect usage of the word "blob" first started to appear.

"Blob", "Binary Blob" -- No.  :no:
"Firmware Image" -- Yes.  :)
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.

les-70

  • Kitizen
  • ****
  • Posts: 1254
Re: Findings with a new connection
« Reply #76 on: March 27, 2014, 10:01:49 AM »

  I can't disagree but the binary large object tends to determine the dsl performance characteristics of the firmware and with a few firmware versions it is the reference number of that object which often needs a reminder.  I am afraid I am inclined to be a bit sloppy that way but I will try to do better as I agree that in the end being careful over descriptions will help readers.
Logged

burakkucat

  • Global Moderator
  • Senior Kitizen
  • *
  • Posts: 31993
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Findings with a new connection
« Reply #77 on: March 27, 2014, 03:10:03 PM »

 ;D
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.

les-70

  • Kitizen
  • ****
  • Posts: 1254
Re: Findings with a new connection
« Reply #78 on: April 01, 2014, 04:55:46 PM »

  Ouch - another nearby connection.  Another nearby property connected today.  For the last two months I have just seen the odd drops of 0.3 mb/s from time to time, adding up to about 4mb/s over two months. However with just one extra house on the road today the attainable went down 7mb/s in one go taking me to 76mb/s attainable. If all the 12 properties on the road connect and do that I could be quite unhappy  :( .  The only other connection on the road was in the first week and reduced the attainable by 14mb/s.   I expected later connections to have a smaller impact consistent with a log snrm scale.   Today's drop seems contrary to that expectation.  Given that I had already dropped from 107 to 83 I hoped that further reductions would be a slower process. 

  edit - to be clear this last reduction takes me from 83 to about 76 mb/s attainable now.
« Last Edit: April 05, 2014, 08:01:12 PM by les-70 »
Logged

les-70

  • Kitizen
  • ****
  • Posts: 1254
Re: Findings with a new connection
« Reply #79 on: April 02, 2014, 02:13:56 PM »

   I decided I would try adding a RF3 to my connection. I don't believe that they are intended for FTTC so I was expecting an unfavorable impact.  The results on the snr are below and a drop of 17mb/s occurred in the attainable speed.  The RF3 has been removed already so I can't say what influence it might have had on errors but a fair test would have involved capping the speed to be the same with and without the filter.  Had I been on a 40/10 connection and not worried by the drop in attainable I might have left it in to see the effect on errors.
Logged

Chrysalis

  • Content Team
  • Addicted Kitizen
  • *
  • Posts: 6369
Re: Findings with a new connection
« Reply #80 on: April 02, 2014, 04:43:08 PM »

les another 9mbit drop and you have a crosstalk loss not too far of mine currently although at its worst was below 65.

my attainable is now currently 69, dont know when it increased as since I upgraded my rig (last week) I havent been monitoring my hg612 and I only checked in the gui today.
Logged
AAISP - Billion 8800NL bridge & PFSense BOX running PFSense 2.4 - ECI Cab - LINE STATISTICS CLICK HERE

les-70

  • Kitizen
  • ****
  • Posts: 1254
Re: Findings with a new connection
« Reply #81 on: April 05, 2014, 03:18:30 PM »

    Probably why were told to leave our FTTC modems on!

   The two neighbors who give me a lot of cross talk have taken to turning off their modems at night. (I do too so I can't really grumble).  See the consequences in my SNRM when they switch on - you can see from the times that they are slow starters!  This is happening every morning.  My speed capping at 64Mb/s down is maintaining a good margin but if left to itself the dlm would probably end up reacting in one way or another.   I have wondered over split pairs but I suspect it just cross talk.

Logged

les-70

  • Kitizen
  • ****
  • Posts: 1254
Re: Findings with a new connection
« Reply #82 on: May 30, 2014, 06:12:07 PM »

  After another two months not much to report.  Running with the HG612 the attainable has dropped another 1Mb/s to 75Mb/s so that from 108 to 75 so far and worryingly most properties round here are still on adsl2.  I am about to move the Zyxel VMG8324-B10A which from tests should gain at least 1Mb/s back.  I would have moved to it before but the power lead proved too short for just a simple swap.

  I am with TalkTalk so the usual BT line info facilities are not available to me.  I have however been tracking my next door neighbor on the BT wholesale speed estimator.  The estimator seems to be trying to demonstrate its incompetence,  earlier figures which ranged from 80-80 on on non impacted, to 80 on impacted high and 67 on impacted low has moved in two separate jumps to now be 69-53, and  60-34 respectively.  That is a crazy jump of a factor of two showing that one estimate or the other was/is quite misleading.
Logged

les-70

  • Kitizen
  • ****
  • Posts: 1254
Re: Findings with a new connection
« Reply #83 on: July 08, 2014, 01:56:10 PM »

  A small update.  Today started with most of my crosstalk missing and an attainable of 93Mb/s, by 8.00am it had dropped to the usual of about 76/77 Mb/s,  but Oh dear a man was seen at the cab   :( and shortly after I am down to 73Mb/s attainable.

 The Zyxel VMG8324-B10A was sent back due to very high errors and strong suspicions of a fault.  I now have a Billion 8800NL (which I am pleased with) and I have been trying to compare error rates of the Billion  8800NL  with those of a HG612.  I discarded all periods with thunderstorms and  after a few week and lots of swaps have concluded that the with exactly the same capped speed the error rate of the 8800NL is 20-30% greater than the HG612.  With capping that can be offset with a small extra speed cap on the Billion.

 I also have a STC Saudi Arabian HG658B which does not support capping. (cost US$8 and $15 postage -- they come up from time to time on Taobao) It has the odd behavior of showing a higher attainable but giving a lower sync than the HG612. In an error comparison with the Hg612 which lasted two weeks and swapped them every day the HG658B has an error rate about 30% less than the HG612.  That was in spite of the HG612 being capped at 72 and the Hg658B running at about 75  The HG658B is unlocked and works with Dslstats (but it took months to find the uername and password for telnet) so I have been considering using it but my large attainable swings each day make me worried about possible resyncs.

« Last Edit: July 08, 2014, 02:58:31 PM by les-70 »
Logged

les-70

  • Kitizen
  • ****
  • Posts: 1254
Re: Findings with a new connection
« Reply #84 on: January 13, 2015, 05:53:33 PM »

Almost exactly one year from connection I thought I would give an update.  The attainable has dropped from 107Mb/s to 70 mb/s (currently higher as someone seems to have gone away) and is still dropping.  The bad news is that most neighbors don't have FTTC so there must be plenty of scope for more drops. The only good news is that the cabinet is 75% full so only 25% to go.

  I have retested my various efforts at optimization of the line.  Although the two go together my main aim has been to keep errors low rather than get a better speed. The speed is way more than enough for my use and the only thing that has annoyed me is two occasions when interleaving was applied and I actually did notice an impact.  I am sure that what is best to reduce errors depends a lot on noise that previals on the line.  I have a general tick over rate of erors coupled with what seems to be SHINE events that occur about 3-6 times a day with potentially many CRC and one to several SES each time.  As noted below the SHINE and some of the general error count seem be mains borne. 

  My biggest benefit comes from running the modem and router, mainly a HG612 and Dlink Dir-655, on a BT mains conditioner. The Main conditioner seems best, if contrary to recommendations,it is unearthed!

 Another a big benefit is also gained if there is NO extra wired ethernet connection to the modem/router but with all access wireless.   I guess this set up is close to running on batteries.  This has not proved a disadvantage for me but as a matter of principle I do prefer a wired connection.

 A smaller but noticeable benefit comes from using a home made common filter using toroidal ferrites.  The does not reduce the background error rate but does help reduce the impact of the SHINE like events.

  I was previously earthing the unused drop wires but with the new set up that hardly makes a difference and those earth connection been disconnected.

   When compared to a straight mains connection - and using just the mains conditioner, the SES rate, a good indicator of SHINE, drops from ~6-10/day to ~1-2/day. With both the conditioner and wireless access only, the SES rate drops to ~0.5/day.  With both improvements in place the overall ES/day is also reduced but by only about 30-50%.   
« Last Edit: January 20, 2015, 09:26:23 AM by les-70 »
Logged

NewtronStar

  • Kitizen
  • ****
  • Posts: 4844
Re: Findings with a new connection
« Reply #85 on: January 13, 2015, 10:22:58 PM »

I have to say being on a long Distribution-side 997 meters with interleaved applied and a SSFP MK3 the SES errors look fanstantic on a daily basis 0 Down and 0 Up.

with the standard MK1 would have seen 1-2 SES per day, so the MK3 for this line is doing a better job than the MK1 and MK2  ;D
Logged

les-70

  • Kitizen
  • ****
  • Posts: 1254
Re: Findings with a new connection
« Reply #86 on: January 14, 2015, 07:28:01 AM »

  When interleaved with my current set up (the cause of the interleave is suspected to be work at cab) the ES rate for 10 days was 0.1ES/day and 0.0SES.

 I did not get any significant difference is stats between Mark 1 2 or 3 SSFP.  My home made common filter takes about 1 Mb/s of the attainable, that could be compared with an RF3 which would currently give a drop of about 12Mb/s.  However the home made one is bifilar and should be at least as effective as an RF3 but but with the big speed impact.

 I assume I get mains borne spikes and the BT mains conditioner coupled with wireless connection to the devices on the conditioner is eliminating most of these spikes.  As I noted what works best will depend on what the error source is.
Logged

kitz

  • Administrator
  • Senior Kitizen
  • *
  • Posts: 32622
  • Trinity: Most guys do.
    • http://www.kitz.co.uk
Re: Findings with a new connection
« Reply #87 on: January 17, 2015, 10:23:15 AM »

Does any one have a experience of what crc and ses  rates the DLM seems to tolerate and preserve fast path?  If the DLM works sensibly and the noise is not unusual this might be a fairly fixed value on lines where the DLM has capped speed but maintained fast path.

I know that is an older post, but since this thread has come up to the top again, I think its worth re-visiting

Documents Ive seen would seem to imply that the last known FTTC parameters are

Speed MTBE 30/300
Stable MTBE 60/600

This equates to 2880 ILQ red & 288 ILQ green


I wrote to someone the back end of last year to ask and they promised they'd get back to me if it was any different, but Ive not heard anything and I dont want to pester because I know BT are far more sensitive about DLM information since the ASSIA court case.

I know theres been discussions about a possible 1440, but iirc that was the original (not latest) figure.. and since ronski and a couple of others on MDWS seem to go over this without DLM intervention, then Im inclined to think that unless the ILQ figures were changed since the ASSIA court case, then they are 30/300
Logged
Please do not PM me with queries for broadband help as I may not be able to respond.
-----
How to get your router line stats :: ADSL Exchange Checker

kitz

  • Administrator
  • Senior Kitizen
  • *
  • Posts: 32622
  • Trinity: Most guys do.
    • http://www.kitz.co.uk
Re: Findings with a new connection
« Reply #88 on: January 17, 2015, 10:47:32 AM »

Almost exactly one year from connection I thought I would give an update.  The attainable has dropped from 107Mb/s to 70 mb/s (currently higher as someone seems to have gone away) and is still dropping.  The bad news is that most neighbors don't have FTTC so there must be plenty of scope for more drops. The only good news is that the cabinet is 75% full so only 25% to go.
 

I dont know if anyone else has noticed this or if it is just a co-incidence, but since BTw started messing with MSE bRAS, Ive been noticing my line is far less stable on the upstream and also that I'm noticing the effects of crosstalk more.

For those that dont know MSE bRAS means that more of the small to medium exchanges are now a satellite from a larger neighbouring exchange which has a MSE bRAS installed.   The effect in my case is that rather than being attached to an OLT in my local exchange, the fibre from the mini DSLAM goes direct to an OLT in the next town, meaning a much longer run of fibre before it joins up to an OLT.

This method has its benefits in that it should mean a shorter backhaul before going on to the core network and slight decreases in latency (which Ive seen), but Im not sure whether its been beneficial for crosstalk or if Ive just been unlucky with picking up some bad disturbers all at once and also seeing a much more unstable upstream since the night I was re-routed.
Logged
Please do not PM me with queries for broadband help as I may not be able to respond.
-----
How to get your router line stats :: ADSL Exchange Checker

les-70

  • Kitizen
  • ****
  • Posts: 1254
Re: Findings with a new connection
« Reply #89 on: January 17, 2015, 11:55:21 AM »

I know theres been discussions about a possible 1440, but iirc that was the original (not latest) figure.. and since ronski and a couple of others on MDWS seem to go over this without DLM intervention, then Im inclined to think that unless the ILQ figures were changed since the ASSIA court case, then they are 30/300

   I agree that on MDWS Plusnet has a few users consistently over 1440 now.  There were none back in November but that may be a coincidence.   For TalkTalk I think the jury out re 1440 or 2880.  TT state that all are on "standard" but their nomenclature is not clear and I suspect the 1440 may apply.   I recall you thinking Plusnet was speed.  I am on TTB and even if I look at a home service Plusnet works out quite a lot dearer when I take phone usage into account.  Otherwise their 2880 ES/day might capture me.
Logged
Pages: 1 ... 4 5 [6] 7