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

Login with username, password and session length
Advanced search  

News:

Pages: [1] 2 3 ... 5

Author Topic: Modem-to-wallsocket cables recap (again, groan)  (Read 5013 times)

Weaver

  • Kitizen
  • ****
  • Posts: 4004
  • Retd sw dev; A&A; 3 × 7km ADSL2; IPv6; Firebrick
Modem-to-wallsocket cables recap (again, groan)
« on: March 25, 2017, 01:54:15 AM »

I’m wondering about using RJ-11 to RJ-45 DSL cables for the modem to wallsocket link. We have talked about this before quite a few times.

I use Tandy RJ-11 to RJ-11 ones currently, but I'm wondering if RJ-45 plugs might be a better fit with the non-filtering faceplates / NTE5/A-fronts that I have. The filterless fronts came from the Andrews and Arnold Shop (https://aa.net.uk/broadband-accessories.html) It explicitly says they are RJ-45 sockets, not RJ-11, but does it even matter? Or is ‘good enough’ good enough?

The cables that I am using are Tandy (former ADSLNation) 0.5m and 1.0m ones (http://www.tandyonline.co.uk/high-speed-rj11-dsl-cable-0-5m.html) which have a real quality feel to them. They measured extremely well compared to the cables I was using before, a shocking difference, one which could be down to bad testing methodology, or shoddy construction of the other cables, or both. This has been discussed in earlier threads.

The alternative I was thinking about is https://www.run-it-direct.co.uk/rj11-to-rjJ45-patch-leads/RJ11toRJ45patchlead1M/ which is very expensive by comparison and the appearance fails to seduce: not gold-plated, unless they're keeping it a secret, twisted pair but no overall shield, no moulded plugs so less robust-looking and no metal around the plug bodies. I don't see any immediate reason why they might be an improvement, and at a far higher price
Logged

j0hn

  • Kitizen
  • ****
  • Posts: 1091
Re: Modem-to-wallsocket cables recap (again, groan)
« Reply #1 on: March 25, 2017, 02:59:22 AM »

All of the OpenReach MK(1,2,3,4) filtered faceplates also have an RJ45 socket. Most people don't even realise it's RJ45 as the ISP supplied RJ11 cable fits as expected.

I use the same unfiltered faceplate as yourself, purchased directly from AAISP. With both the OpenReach filtered and the AAISP unfiltered faceplates I prefer using a cable with an RJ45 socket. It fits much more snug, as the smaller RJ11 has a bit of "wiggle" in the RJ45 socket.

I use a RJ45 to RJ11 cable made by Mr Telephone (My mate Vince), purchased from his EBay store. I recall you saying in the past you had a bad experience with 1 of his cables. I've tried expensive Belkin cables and the Tandy cables and neither performed as well as Vinces. His cable being RJ45 at 1 end is a nice bonus.
Logged
BT FTTC 55/10 ECI Huawei Cab - Zyxel VMG1312-B10A bridge mode + Asus RT-AC68U running Asuswrt-Merlin - minted on MDWS via DslStats

ejs

  • Kitizen
  • ****
  • Posts: 1454
Re: Modem-to-wallsocket cables recap (again, groan)
« Reply #2 on: March 25, 2017, 05:33:27 AM »

Run IT Direct have changed their DSL patch cables since I bought one.

They have a CAT-5e one which is higher specification and cheaper than the one you linked to:
https://www.run-it-direct.co.uk/adsl-vdsl-patch-leads/adslvdslcat5patchlead1m/ (other lengths also available)

The one I bought used a standard 4-pair CAT-5e Ethernet cable, and only wired one pair to the centre two pins of the plugs. Now they are using 2-pair cable and wiring up both pairs to the RJ-11, not that wiring up a pair of wires which aren't going to be used for anything will make any difference.

The more expensive cable you linked to is not CAT-5, it's telephone grade CW1308.

I suppose nobody makes high specification cables with an RJ-11 plug on one end and a BT phone plug on the other, then you wouldn't need a special faceplate.
Logged

Weaver

  • Kitizen
  • ****
  • Posts: 4004
  • Retd sw dev; A&A; 3 × 7km ADSL2; IPv6; Firebrick
Re: Modem-to-wallsocket cables recap (again, groan)
« Reply #3 on: March 25, 2017, 08:18:52 AM »

Ejs wrote:
> suppose nobody makes high specification cables with an RJ-11 plug on one end and a BT phone plug on the other, then you wouldn't need a special faceplate.

Good point. I had thought about that, it would just mean a posh dial-up modem cable, and they must still be around, from someone such as Belkin. However they will all be too long by far, and you can't beat shortness.

@j0hn - thanks for the tip about your good experience with Vince. The one that I had was simply falling apart, perhaps I would have done fine with a different individual unit. I should perhaps give Vince's ones another try. I have to say though that measuring these things is tricky because of DLM messing you about and if I were doing the comparison now I would be much more careful about it. (See http://forum.kitz.co.uk/index.php?topic=15905.0 )
Logged

Weaver

  • Kitizen
  • ****
  • Posts: 4004
  • Retd sw dev; A&A; 3 × 7km ADSL2; IPv6; Firebrick
Re: Modem-to-wallsocket cables recap (again, groan)
« Reply #4 on: March 25, 2017, 09:08:38 AM »

I have decided for the sake of science to compare a short 0.5m Vince CAT6 RJ11-to-RJ45 cable with a Tandy one. I will do so carefully this time and report back. I think that he even does 0.3m ones, if memory serves.

What is the best thing to capture? Simply go on sync rates with repeated a-b comparison? Do we try and get some information about SNRM over time? Or some kind of error rates?

One way of looking at it is to think of science only and try to decide on which measure is the most revealing in some aspect or other than you have to argue about the relevance of that measure.

The other way is just to decide what you want to see optimised and your preference is what it is. In my case I am just trying to optimise sync rates, first downstream and then upstream. And I am trusting the systems to keep to a standard level of reliability so that we are comparing like with like in the sense that we don't by some strange miracle end up with modem plus cable A going faster than B but at the expense of differences in SNR.

My modem of one of these cables is that it has a certain transfer function and lets in a certain amount of noise. Is that fair enough? Do I need to mentally include impedance mismatch/ power transfer?

Someone who has ready access to full stats from their modem might be in a better position to make the comparison. Even if I were to isolate a modem and telnet into it I can't use the good continuous stats capture tools that most of us have because of the firmware in question. Any volunteers to try the two cables back to back?

I have the advantage of a very weak DSL signal so I tell myself that these small effects are more significant in my case than for someone with a short line and hence high voltage. However, I am not testing higher frequencies so that could mean that the reverse argument could be made, that a VDSL2 user might be able to contribute something that I cannot. Perhaps we need a slow long line user and a VDSL2 user both?
Logged

ejs

  • Kitizen
  • ****
  • Posts: 1454
Re: Modem-to-wallsocket cables recap (again, groan)
« Reply #5 on: March 25, 2017, 09:30:53 AM »

All I'm going to say on the subject is that all the theoretical improvements made to my setup, getting a modern master socket, CAT-5e modem cable, even getting rid of the star wiring and numerous superfluous crimps and little bits of wire, made no significant difference to the line rate.

I wouldn't be at all surprised if any differences between the cables are so small that it's difficult to measure accurately given all the uncontrollable external factors.
Logged

tubaman

  • Reg Member
  • ***
  • Posts: 550
Re: Modem-to-wallsocket cables recap (again, groan)
« Reply #6 on: March 25, 2017, 11:36:44 AM »

I have a star-wired extension setup and when I used to have ADSL the single biggest improvement was to disconnect the ring wire going to the extensions.  This pushed my ADSL speed from 5Mbps up to about 6.5Mbps.
Fitting a filtered faceplate then gave me about another 0.5Mbps on top.
 :)
Logged
BT FTTC 80/20 Huawei Cab - Zyxel VMG8924

sevenlayermuddle

  • Helpful
  • Kitizen
  • *
  • Posts: 3341
Re: Modem-to-wallsocket cables recap (again, groan)
« Reply #7 on: March 25, 2017, 11:45:24 AM »

Trouble with any such testing is, no two syncs, even with the same cable and minutes apart, are guaranteed to be exactly the same.  It might depend upon whether the central heater stat has kicked in, or whether radio 4 is broadcasting laughter or silence, or what direction the radar station at the local airbase happens to be pointing, or dozens of other random factors. 

I would argue any tiny differences that might appear to be attributable to cables are pretty much indistinguishable from these random factors.  You can reduce the random effects by doing enough tests that the random stuff statistically cancel out, but DLM would spotted all the disconnects, and would have intervened long before you got to that point - ruining the experiment.

On the odd occasion I have tried to do any such testing, I have found no solid evidence that posh cables make the slightest difference.
Logged

ejs

  • Kitizen
  • ****
  • Posts: 1454
Re: Modem-to-wallsocket cables recap (again, groan)
« Reply #8 on: March 25, 2017, 11:57:15 AM »

I was talking about star wiring connected to the line before the master socket. If a filtered faceplate would have solved the issue, it would have been far simpler to fix.

----

Anyway, back to the subject of cable testing. Even if you swap the cables very quickly and find you get a higher speed with one cable, you might find that later, under different environmental conditions, at a different time of day or night, the other cable is better.

Rather than compare two high-specification cables, perhaps also compare the ordinary lead that came with the modem.
Logged

j0hn

  • Kitizen
  • ****
  • Posts: 1091
Re: Modem-to-wallsocket cables recap (again, groan)
« Reply #9 on: March 25, 2017, 12:29:41 PM »

The cable I bought from Mr Telephone increased my sync by 2-4mb (taking account of usual variations). My line attenuation also dropped by 0.1
VDSL2 certainly appears more sensitive and is more likely to give gains with a better modem cable.

After ADSL running along nearly 8km of copper, the last 0.5m is never going to make a huge difference. If I were you though Weaver I couldn't help but try anyway  ;D
Logged
BT FTTC 55/10 ECI Huawei Cab - Zyxel VMG1312-B10A bridge mode + Asus RT-AC68U running Asuswrt-Merlin - minted on MDWS via DslStats

Weaver

  • Kitizen
  • ****
  • Posts: 4004
  • Retd sw dev; A&A; 3 × 7km ADSL2; IPv6; Firebrick
Re: Modem-to-wallsocket cables recap (again, groan)
« Reply #10 on: March 25, 2017, 02:23:00 PM »

I beg to differ about the last small length. This is my thinking: The last x km are clean electrically and then suddenly we enter a house full of electronic noise. The other thing is that attenuation means that we are dealing with very small signal voltages so (noise / signal ) ratio is most significant at my end of the downstream. I may of course be well wrong as I have no idea about the magnitude of these numbers even if the logic is sound.

The other thing is that like you, I got a really substantial difference by switching cables before something roughly >20%. But possible bad testing methodology back then is something that bugs me.

Logged

Weaver

  • Kitizen
  • ****
  • Posts: 4004
  • Retd sw dev; A&A; 3 × 7km ADSL2; IPv6; Firebrick
Re: Modem-to-wallsocket cables recap (again, groan)
« Reply #11 on: March 25, 2017, 02:32:32 PM »

I completely agree with the earlier posters about the noise in the test results anyway, and indeed the effects might be too small (by a huge long way) for the resolution of the procedures we have at our disposal.

But my thinking is that if anyone has a chance of seeing such effects (and perhaps no one has a chance) then it's me, because of my low signal levels and uncannily clean house-external noise environment (apart from things like radio stations), because most people have human civilisation all around their conductors’ run outside their premises. Having said that, it is perhaps an unhelpful thing that I am not testing umpteen higher frequencies at all when compared with normal people.
Logged

burakkucat

  • Global Moderator
  • Senior Kitizen
  • *
  • Posts: 20325
  • Over the Rainbow
    • The ELRepo Project
Re: Modem-to-wallsocket cables recap (again, groan)
« Reply #12 on: March 25, 2017, 09:45:21 PM »

Just a few comments --
  • I fail to understand how a screened cable with metallised plugs on its ends can be superior to other cables if the screening is not connected to a good signal earth.
  • As much as I like Vince (Mr Telephone Supplies), I disagree with his policy of attempting to make patch cables out of the solid core cable used for structured wiring. Structured wiring -- solid core. Patch cables -- stranded core. As other have discovered, the plugs crimped onto the solid core wires of Vince's patch cables will come off, sooner or later.
  • As ejs has mentioned, Run IT Direct now have a different set of patch cables available. (8P4C modular plug and 6P4C modular plug, respectively, at opposite ends.) I would be tempted to use one of those cables.
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.

sevenlayermuddle

  • Helpful
  • Kitizen
  • *
  • Posts: 3341
Re: Modem-to-wallsocket cables recap (again, groan)
« Reply #13 on: March 25, 2017, 10:33:01 PM »

  • I fail to understand how a screened cable with metallised plugs on its ends can be superior to other cables if the screening is not connected to a good signal earth.

I know opinions differ within the forum, but I strongly agree with  Burakkucat on this aspect.

In fact, I would argue such a cable ought be inferior to a normal unscreened cable, for the reason that any increase in conductor surface area will lead to an increase in currents induced from interference.  If the energy produced by these currents cannot be transferred to a good earth conductor, then that energy will have to go somewhere else.   In the absence of an earth, the energy will transfer to the internal conductors, adversely affecting SNR, and making matters worse.

But I did say, opinions differ.  I would not be surprised if other opinions were to come along soon. :)
« Last Edit: March 25, 2017, 10:39:32 PM by sevenlayermuddle »
Logged

roseway

  • Administrator
  • Senior Kitizen
  • *
  • Posts: 38807
  • Penguins CAN fly
    • DSLstats
Re: Modem-to-wallsocket cables recap (again, groan)
« Reply #14 on: March 25, 2017, 10:40:51 PM »

I won't be contrary, I agree with the two of you. :)
Logged
  Eric
Pages: [1] 2 3 ... 5
 

anything