Print

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

[Darren]

Maybe I should of been clearer, I understood, I was just saying it can be beneficial to have one on the power cable aswell. In any case no appology necessary

[les-70]

   Currently next door has a HR fault with phones giving out odd random rings, no dial tone most of the time, sync on HH5 FTTC on and off and download down to 8mb/s c.f. attainable of 80+ when it is on. DLM is no doubt quite upset.  They have had one engineer who said an v old cable running between poles needs changing.  They are waiting for a fix. There is a distinct impact on my line.  See the two snr traces below, one with no fault next door and the other with the fault.   One day they had a full snr the trace that day was good for me.  lost their cross talk and snr remained flat.

  When next door looses sync my snr jumps up.  When the jumps up and down occur there are spikes -very large ones - in the errors.  I wondered it was the surges to and from their bell wire capacitor or maybe just the bit swapping needing to adjust the bit loading.  My capping of the line speed  seems to be keeping thing manageable on my line but without the extra 4db that I added to the snr I suspect the current crc spikes of about 500 in 30 sec would be a lot lot bigger.

[les-70]

    I spoke to an engineer at my cabinet today.  There are now 24 FTTC connections live from the 128? ECI cabinet.   After the obvious individual impacts of the first few connections the snrm seem to just be drifting slowly down now with a drop from 7.8 at the end of January to 7.3 now.  The snrm is a long way down from the 14db seen when first connected.  I also guess properties actually in my cable run may still be able to produce big drops when and if they get FTTC.   The upstream snrm is bigger at 8.5 but suffered a big drop when my upstream power was reduced a week or so ago -probably a sensible change to give others fair shares.

  Mostly I have been capping the speed to boost the snrm as the errors started rise a lot below 7.8 db snrm.  It is a pity the tolerance of the DLM to errors is not properly known.  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.

[NewtronStar]

It is a pity the tolerance of the DLM to errors is not properly known.  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.

Your looking for the Openreach DLM Holy Grail many have searched high & low for this answer many have not come back from this quest, and a few have but they no longer act like inquisitive members more like the life & soul has been taken away from them. 
but they did tell me there is a 14 day wait if things go wrong before it returns

[Bald_Eagle1]

    QLN noise as the only FTTC connection and then with just next door as well

   The impact of just one other connection was a shock to me    and dropped my attainable from 107 to 93 Mb/s  and the snrm dropped from 14 to 10.  Makes BT wholesale estimate of 80/20 for my line most unlikely as it would not take much more of this to take the snrm below 6.   I have long believed that I have a section of split pair but then again maybe this the reality of an just an average connection.  I attach the qln before and after my neighbor was connected. It is interesting to flash between them to see the exact change that a single other connection causes.  For info the HG612 latest blob 038 and the HG622 on Blob 033 both show qln and hlog for the upstream tones as well.

I realise this is a response to your original older post, but just for fun, I have animated the 2 QLN graphs.

There really is quite a significant difference between them.

[NewtronStar]

There really is quite a significant difference between them.

BE1 thats scary those peaks look like my QLN whats going on ?

[Bald_Eagle1]

This is what happened to my QLN graph when I forced a resync.

It had been up for 50 days prior to the resync, so probaly due for a refresh.

My connection's too long to use the D3 band, so QLN at those frequencies mustn't have any real effect on the useable bands.

4 days later, the gang installing new street lighting colums & some cabling cut through the underground phone cable, I can only 'suspect' they have done something else somewhere along the route, but as it's not affecting anything I have no need to complain.

The D3 band's QLN is still the strange shape since my phone/broadband services were restored & I have also reistated Asbokid's original firmware & allowed BT to remotely update it since services were restored.

In fact, my  connection resynced at almost 2 Mbps higher DS speed.

[les-70]

Quote from "So What Is In a BT Mains Conditioning Unit?"   http://forum.kitz.co.uk/index.php?topic=13570.0
 

  I have ordered one to try as an experiment.  It will be able to power the PC + screen, router and modem, but not an old laser printer (too many watts) that is connected via Ethernet to the router. I will try it powering just the modem and router first.  The printer is powered off unless in use and could be disconnected most days.

 Recalling neutral experiments reported here with batteries, I am not optimistic, but printer aside it can power all directly or indirectly connected devices.  My hope is that the odd burst errors that I get about every other day or so and come with about 10 ses may be reduced. 

My BT mains conditioner arrived and emits the low hum typical of a large transformer.  The casing seems badly designed to amplify the hum. I hate things that hum but a sound damping pad on the casing either side of the unit reduces the sound a lot and makes to seem fairly hard to notice unless your right by it.

  Before taking it apart I decided to try it out for a week. I am only powering the HG612 and the router on it at the moment. The only other connection to those items is ethernet from the PC to the router.  I am not sure what can travel down the ethernet to give noise but as noted above I can power the PC and all that is connected to it as well and may try that later on.

   I really did not expect any impact but there is clear reproducible impact tried 3 times and always the same. (with fingers crossed re DLM)  With my normal plain mains I powered off and then on again via the conditioner, then back to normal and repeated this 3 times with a self speed cap to exactly the same sync of 60Mbs down.  The snrm is normally very stable in the day and the test consistently gave +0.2 or +0.3 db with the conditioner c.f. without it, matched with about 0.9 Mhz attainable increase with it.  Below I post the snr with and without the conditioner, the visible difference was consistent each time.  The difference is however not entirely positive and the extra noise at some frequencies puzzles me.

   The initial boost of ~0.2 db is tiny but I am encouraged to now leave things be and see how the error stats over a few days behave.   It is nice to see some effect 

[les-70]

 An update. 

  1. Running everything off the mains conditioner was worse, it removed about half the benefit.  I assume noise from the PC etc was added back in.

  2. Turning off the router gained a further 0.1-0.2 db but is clearly not a viable option as I need the wireless and more LAN ports.

  3. MUCH MORE PUZZLING -  - I have found moving ethernet, power supply and dsl cables about causes changes comparable with the impacts of the conditioner.  What ever is changed the conditioner still helps.  I found this when I decided to tiidy up with some 1m ethernet cables rather than 2m ones.  Using the 1m cable caused a drop bigger snrm drop than the conditioner gain!! It was linked with much more noise in the highest frequency band. I am quite puzzled by this as I thought ethernet cables were alll much the same unless run over long lengths. Maybe all this a symptom of something noisy aorund the pc area that I have not tracked down.

[waltergmw]

@ Les-70,

If you think you have ethernet cable problems, you could make up your own leads from FTP cable but only earthing the drain wire at one end.

Kind regards,
Walter

[les-70]

            Ethernet puzzle.

  As noted above I was puzzled by sync changes when changing the ethernet cables used to connect the Hg612.  I ended up with a laptop on its batteries and the HG612 in sync but not connected to anything other than the lan2 going via the ethernet cable to the laptop.  With nothing else at all was power up, this looked as clean a test as I could conduct.  The connection is vdsl with an attainable of 83mb/s and during the test no resyncs were made.

 Just changing the Ethernet cable caused the snrm to consistently change, usually difference were between 0.1 and 0.3 db but a couple of cables in a fair sized collection gave changes of 1-2 db.  The worst were cables from  BT home hub 3 pack.  The BT wan link cable with a red connector was the very worst at about drop of 2.5 db snrm. I don't know if that is a non standard cable not for a lan, or just a bad cable.

 Stats from the modem suggest the "bad" cables give noise mainly in the highest frequency vdsl band.  I therefore guess that there may be no effect on an adsl connection.  I also note that lan2 is not expected to be in use, but I doubt that is relevant. 

[NewtronStar]

 Just changing the Ethernet cable caused the snrm to consistently change, usually difference were between 0.1 and 0.3 db but a couple of cables in a fair sized collection gave changes of 1-2 db.  The worst were cables from  BT home hub 3 pack.  The BT wan link cable with a red connector was the very worst at about drop of 2.5 db snrm. I don't know if that is a non standard cable not for a lan, or just a bad cable.
 

I don't understand how changing the ethernet cable from BT Router to HG612 could effect the SNRM as the HG612 modem gets its information from the DSL cable from the master socket and not from the Router 

[burakkucat]

Just changing the Ethernet cable caused the snrm to consistently change, usually difference were between 0.1 and 0.3 db but a couple of cables in a fair sized collection gave changes of 1-2 db.  The worst were cables from  BT home hub 3 pack.  The BT wan link cable with a red connector was the very worst at about drop of 2.5 db snrm. I don't know if that is a non standard cable not for a lan, or just a bad cable.

It is not immediately obvious where the Ethernet cable, to which you refer, is connected. Is it the cable that you are using to connect the LAN2 port of the HG612 to the laptop computer?

Like N*Star, I can't see how an Ethernet cable can be having any effect on the SNRM of the CO - CPE segment of the circuit. 

[les-70]

   In the test  the Hg612 has 3 connections.  1 - power, 2 -vdsl, and 3 pc on batteries via Ethernet only to lan2.   lan1 is not connected in the test so the modem is in sync but there is no active ppoe session on lan1.  Nothing else in the room has power. 
 
    I found the same effect in complete "system" of modem/router/pc but have tried to isolate things to a test.  I need to try the same test with a router wan connected  to lan1 and changing that cable but I have been trying one thing at a time in an effort to track down the puzzling impacts. I was puzzled    to begin with but having narrowed down to this simple configuration am perhaps more puzzled    .  Perhaps the appearance of extra noise in the highest frequency vdsl downstream band may suggest noise is getting into the router due to or through the Ethernet cable.  Perhaps the variability with the cable suggests that the "bad" cables may lack "pair balance" or something else must be odd with offending cables.

[burakkucat]

Ah, I see. 

<snip>
but there is no active ppoe session on lan2.
<snip>

Reading the above "lan2" as "LAN1". 

Perhaps you could set up a test environment in the sheddyian style? Get out the two or three old car batteries that you have stashed away (for a rainy day), connect them to a pair of power bus-bars and then run the HG612 and the computer from the common battery supply. The mains supply can then be isolated / completely off throughout the house.