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: Slow upstream - why  (Read 3613 times)

Weaver

  • Senior Kitizen
  • ******
  • Posts: 11459
  • Retd s/w dev; A&A; 4x7km ADSL2 lines; Firebrick
Slow upstream - why
« on: August 02, 2022, 04:21:56 PM »

Why do you think my line 4 upstream is so slow? It doesn’t change over time, by the way. See the comparison between line 4 and 3 or 1.

Line 1: Firebrick upstream PPP rate: 542.640 kbps  |  Link sync rate upstream: 646 kbps

Line 2: Firebrick upstream PPP rate: 464.520 kbps  |  Link sync rate upstream: 553 kbps

Line 3: Firebrick upstream PPP rate: 597.240 kbps  |  Link sync rate upstream: 711 kbps

Line 4: Firebrick upstream PPP rate: 295.680 kbps  |  Link sync rate upstream: 352 kbps
Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Slow upstream - why
« Reply #1 on: August 02, 2022, 04:52:43 PM »

My guess:

All four of your lines do not have the occasional application of around 95V AC, the telephony ringing voltage, applied to them as they are DSL only circuits. However there is a constant 50V DC on the pairs. All the joints slowly become conditioned with that steady DC voltage and, in the presence of moisture, start to show semi-conductive and/or high resistance tendencies. A blast of AC would probably clear the problem. (And deter the haggis from nesting in the joint closures.)
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.

Weaver

  • Senior Kitizen
  • ******
  • Posts: 11459
  • Retd s/w dev; A&A; 4x7km ADSL2 lines; Firebrick
Re: Slow upstream - why
« Reply #2 on: August 02, 2022, 04:58:36 PM »

A couple of questions to be considered:

1. Why would that line, line 4 have been ‘born bad’ ? Was upstream-slow when I first got it.
2. Why the ×2 difference between line 3 and line 4 ?

Do these pose problems for your theory?
Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Slow upstream - why
« Reply #3 on: August 02, 2022, 05:05:12 PM »

"Non-linear response at radio frequencies" would be my utterance. That can be the source of many a DSL circuit problem.
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.

Weaver

  • Senior Kitizen
  • ******
  • Posts: 11459
  • Retd s/w dev; A&A; 4x7km ADSL2 lines; Firebrick
Re: Slow upstream - why
« Reply #4 on: August 02, 2022, 05:35:10 PM »

Non-linearity indeed, agreed. Why would it only affect the lower frequencies and so upstream? Upstream slowness seems to be completely independent from downstream slowness - there seems to be no correlation.
Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Slow upstream - why
« Reply #5 on: August 02, 2022, 05:47:18 PM »

If we consider the non-linearity as also applicable to the frequency domain then there can be significant attenuation at the lower (radio) frequencies whilst the higher (radio) frequencies are less affected.
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.

Weaver

  • Senior Kitizen
  • ******
  • Posts: 11459
  • Retd s/w dev; A&A; 4x7km ADSL2 lines; Firebrick
Re: Slow upstream - why
« Reply #6 on: August 02, 2022, 08:28:44 PM »

Would that be because a higher frequency means "more zero-crossings per unit time" ? And being near a zero-crossing point (ie instantaneous voltage value V is very low) is a bad thing, so the more times you are well away from the region V=0 the better. I’m not at all sure about the mathematics of that though.
Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Slow upstream - why
« Reply #7 on: August 02, 2022, 10:00:24 PM »

Would that be because a higher frequency means "more zero-crossings per unit time" ?

I really cannot say . . . as I have not studied the topic. However higher frequencies ==> greater number of zero-crossings (per unit time). So you might have intuitively grasped at a concept that is applicable.  :)
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.

Weaver

  • Senior Kitizen
  • ******
  • Posts: 11459
  • Retd s/w dev; A&A; 4x7km ADSL2 lines; Firebrick
Re: Slow upstream - why
« Reply #8 on: August 02, 2022, 10:05:25 PM »

The wetting current thing is non-linear around zero that’s why DC bias works, as it gets us away from V=0, and AC bias gets us away from zero for a decent fraction of the time. Agreed?
Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Slow upstream - why
« Reply #9 on: August 02, 2022, 10:22:55 PM »

The DC bias is essentially "charging up" small electrochemical cells at each iffy joint in the metallic pathway. Eventually those series connected cells perturb the metallic pathway sufficiently that the wanted RF energy is highly degraded. Possibly in a frequency dependent fashion.

The AC ringing voltage will "reset" each of those small electrochemical cells (i.e. pair joints) back to a linear state . . . only for the DC bias to start the unwanted effect once again, as soon as the AC voltage is removed.

In terms of AC power transmitted through the pair, that of the ringing voltage is far greater than that of the xDSL transceivers at either ends of the link.
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.

Weaver

  • Senior Kitizen
  • ******
  • Posts: 11459
  • Retd s/w dev; A&A; 4x7km ADSL2 lines; Firebrick
Re: Slow upstream - why
« Reply #10 on: August 02, 2022, 10:25:28 PM »

Ah! Light-bulb appears!  :)
Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Slow upstream - why
« Reply #11 on: August 02, 2022, 10:36:50 PM »

Ah! Light-bulb appears!  :)

  :thumbs: 

Hence why I suggest that the application of the ringing voltage, a "blast from the Megger" or a vigorous cranking of a telephone magneto may clear the fault.
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.

kitz

  • Administrator
  • Senior Kitizen
  • *
  • Posts: 33881
  • Trinity: Most guys do.
    • http://www.kitz.co.uk
Re: Slow upstream - why
« Reply #12 on: August 03, 2022, 12:17:12 AM »

It seems to be one of things for which there is no definite answer and we've exhausted the most obvious reasons. Sometimes there is no answer, it can be applied to just about anything.   Why do some lightbulbs last longer than others, why do some electrical appliances fail, when an identical model can last for years. 

Whilst you may think all lines are identical, there's the possibility that it could be on a different MSAN, or different line card.  Sometimes a certain line card may not perform quite as efficiently as it could, but it is still within acceptable parameters.  Line 4 could also be on a port which is affected by a crosstalker that doesn't come into contact with the other lines.. 

Crosstalk can rob adsl2 lines too.  We dont really talk about it so much on ADSL when compared to VDSL because the speed losses are smaller (usually kbps rather than Mbps) and lack of decent monitoring.... but its still there.  When BE* came to this exchange, no one else had really heard of them..  I was one of their first connections and had practically zero crosstalk for quite a while .  When I visited  my exchange the MSAN only had about 2 dozen lines. That changed when O2 bought them out... and even more changes when Sky bought them out.  What previously was an easy 24Mbps Annex_M dropped to more like 16Mbps by the time I moved away.  A single disturber could literally just wipe 1Mbps off my sync speed at the flick of a switch.   You can't always spot crosstalk on adsl graphs are there are fewer tones, but sometimes you could see the shallow bowl like dips on the bit load graphs.  250kbps FeXT is quite possible on ADSL2.
« Last Edit: August 03, 2022, 12:30:25 AM by kitz »
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

Weaver

  • Senior Kitizen
  • ******
  • Posts: 11459
  • Retd s/w dev; A&A; 4x7km ADSL2 lines; Firebrick
Re: Slow upstream - why
« Reply #13 on: August 03, 2022, 01:33:32 AM »

@Kitz - weird. Thanks for that.
Logged
 

anything