Kitz Forum
Broadband Related => ADSL Issues => Topic started by: HPsauce on September 11, 2014, 10:24:35 AM
-
Intriguing one this........
I moved from Be to PlusNet some months ago. On Be I'd been stable for years on a3dB profile, but with a non-standard (low) level of interleaving that was enough to contain occasional unidentified interference that would make the non-interleaved line unstable occasionally.
At PlusNet I soon found you needed interleaving had to be off otherwise upload speeds were very limited. It didn't make sense to me, but that's the way it was and you had to have interleaving on or off the same on both Up and Down (why?).
I let DLM do its thing and ended up with a 3dB profile which worked fine though speeds were nonetheless about 10% to 15% lower than with Be, both ways. :(
On Tuesday this week I noticed lots of service interruptions during the day; I also saw some BT guys working on my local cabinet when I passed it (just saying........ ;) ) it's some distance away and not visible from my house.
Late in the evening I had an interruption that didn't recover and after a couple of router reboots could see that it wasn't getting DNS so called PN. A few minutes wait (it was 11pm or so) and got someone helpful.
At first they thought a change of DNS would sort it but then realised I was connecting, and re-connecting, to the same overloaded "rack" - their word. So after some reconfiguration at their end and a router power cycle all was well again.
The tech said (I knew) there had been loads of resyncs during the day (one of which must have triggered the DND weirdness) and that he would monitor the line and call me the next day near the start of his evening shift if there was anything significant.
Well there was, loads more resyncs yesterday (Wed) and he duly called me about 5pm. I was running DSLstats (thanks ;D ) and could see the noise margin varying through the day and resyncs when it got down to about 2. After multiple resyncs it went to 6dB but after a while (hour or so) reverted to 3dB again and the frequent resync cycle restarted.
After some discussion he said that it looked like DLM was trying to go lower than 3dB so he would make some changes to stop that, wasn't 100% clear to me what changes but I asked that interleaving stayed OFF if possible.
Another resync after his changes and I was back on 6dB but oddly at a much lower speed (13.6mbps) than during the day when on 6dB (about 15.5mbps and 17mbps on 3dB).
Then, strangely, about 2 hours later (about 8pm) the SNRM reported by the router jumped by 3dB to around 9, where it still is now over 14 hours later. :'( The connection speed is unchanged, no resyncs logged, and has been constant since 6pm yesterday.
This seems very odd and I'm not sure I believe the "DLM trying to go below 3dB" analysis at all.
Any thoughts from the gurus, other than some massive source of interference was switched off about 8pm last night and is still off?
I can also see on the SNRM graph a plateau of higher value (by maybe .4dB) from about 11:30pm to 7:15am which presumably represents some source of interference that was turned off for that time. Nothing in my house matches that for sure.
-
Haven't had time to digest your post HP, just saw lots of LOS mentions. Wondered if this may have played a part .....
http://forum.kitz.co.uk/index.php?topic=14412.0
-
You got a call back! :)
Although, this is both good and bad.
Emails full of BS can be dissected. A telephone conversation gives you nothing to refer to.
I would still commend him on the callback :D
-
I don't think the BT work was relevant, in fact I "think" I've found the cause of the disconnections.
Yes, it's a dodgy switched-mode PSU! >:D
A fairly chunky one as it powers a TV/monitor - 12V/5A. Binned and replaced, lucky I had a "spare" (multi-voltage laptop unit with exchangeable connectors).
I've also rebooted the router and it's gone back to 6dB margin and a half-decent speed - 15.2mbps at 5.9dB.
I still don't believe the "DLM trying to go below 3dB" analysis though, I think that's just increasing interference driving the SNRM down.
-
yeah the analysis was wrong, DLM doesnt have lower than 3db profiles and there is no SRA on BTW based adsl so they cant change target snrm on the fly either. You simply had some form of noise injection which thankfully you have found.
-
I still don't believe the "DLM trying to go below 3dB" analysis though
No me neither. The DLM doesnt work like that. It has a fixed set of parameters.
-
Thanks, what I thought too. 8)
With my replacement PSU I can still see a smallish (0.3dB or so) change in SNRM when I turn the TV/monitor (it does both jobs - 19" monitor with Freeview tuner built in) on/off, but it's stable around that, unlike the previous one.
As a precaution I've ordered a brand new replacement anyway. The current PSU was for other purposes, powering laptops.
Meanwhile I didn't get a 2nd callback (that was promised!) from PN :'( so I've just updated the ticket requesting a reset back to a normal profile so that DLM can do its thing again, hopefully getting me back to 3dB.
-
At least you found out what it was.
When you say slight .3dB when you turn on/on, is it just a quick spike, or more permanent. Sorry I couldnt figure out which you meant.
-
Sorry, permanent. That's the problem, I know what I mean, but the words are open to interpretation. :-[
e.g. 5.9dB just now (unplugged), 5.8/5.7dB (on standby) and 5.6dB in use.
All nicely shown and graphed by DSLstats.
-
Damn, I was hoping for your sake that it would be a quick spike. If its a decline then thats not good, because if it happens each time, its further going to eat away your SNRm
Hopefully things will improve with a new replacement. :fingers:
-
Not worry I have plenty of LCD and LED TV's around the house when switching on each TV one by one I see the SNRM drop 0.3dB and the next TV 0.6dB and then the next one the SNRM has gone down by 1.0 dB.
Have posted about this issue many months ago with LCD/LED TV's effecting SNRM loss on the Modem, but have not had a satisfactory explanation why these TV's are interfering with the modem, some of thoses TV's are many meters away from the modem.
-
Interesting; I wonder if it's radiated interference or injected into the mains?
If I can be bothered I might try my unit on a 12V battery and see whether it is the screen or PSU that is worse.
PSU (the newer one) on its own (not connected to TV/monitor) seems to have no effect, but maybe without any load the guts do very little.
-
Interesting; I wonder if it's radiated interference or injected into the mains?
If I can be bothered I might try my unit on a 12V battery and see whether it is the screen or PSU that is worse.
PSU (the newer one) on its own (not connected to TV/monitor) seems to have no effect, but maybe without any load the guts do very little.
You will find it's only when the screen is on it effects the SNRM that would say the PSU is not the cause but the workings inside the screen when in the on position that is causing interference.
-
Well, my problem earlier in the week was definitely the PSU, but that was probably masking smaller effects from the screen.
Now I see the largest step is from no power to standby, then a smaller one to fully on.
When I get the proper replacement PSU I'll be able to do some comparison tests; all very interesting nonetheless. :graduate:
-
Well, my problem earlier in the week was definitely the PSU, but that was probably masking smaller effects from the screen.
Now I see the largest step is from no power to standby, then a smaller one to fully on.
When I get the proper replacement PSU I'll be able to do some comparison tests; all very interesting nonetheless. :graduate:
You have to remember even on standby a TV and Moniters well receive a supply of voltage to the internal circuitry it's like PC sleep mode, my samsung 40 inch LED TV sends out a 7Mhz signal from the board when in standby mode, My Old 32 inch cathode ray TV was a lot quieter :o
Yes HPsauce it would be excellent if you could find and cure the relationship between Moniter and the SNRM loss to modem as it's been annoying me for well over 3 months, haved tried the RF3 and it made things worse (more SNRM loss).
-
A Faraday Cage will feature in my experiments; I just need to decide what to make it from.........
I also happen to have 2 separate downstairs ring mains in my house, so I'll play with that aspect, though the TV is currently (ha ha) on a different one to the router.
-
DECT, either of you?
-
Yes I have two DECT phones, will this cause the SNRM to drop when the LED/LCD TV in on ?
-
I know off topic but newtron my friend, where is your graph on mydslwebstats :)
-
posted about this issue many months ago with LCD/LED TV's effecting SNRM
My first experiment was to move the offending unit to another room, further away from the router (and phone line) but on the same ring main.
Distance is maybe double, but there's a fair bit of additional structure in the way now.
Interference is still there, following the same pattern but at around half the level. Difficult to be precise giving the low values and rounding.
So my first hesitant conclusion is that it's radiated interference, probably not carried over the mains.
-
And for my second test I used the same PSU to drive a spare router which would be "live" but lower power. Nothing.
So I moved it to be right next to my router and phone socket, on the same power outlet. Still nothing.
Beginning to look like it's probably the main TV/monitor that is the cause of my "residual" 0.3dB loss...........
-
Interesting, but somewhat perplexing........................
The primary "control" over interference is whether or not the TV aerial is plugged in! Even in standby mode or doing another function (VGA monitor, playing DVD's etc.). Location seems to be a minor factor.
So it's something to do with the DVB-T tuner and operates significantly even when just in standby mode, but ONLY when there's an aerial connection supplying a signal.
My aerial comes via a distribution amplifier (in the loft) and also has a masthead amplifier on the aerial, which may or may not be relevant.
Something for others with similar problems to test - pulling out an aerial lead is easy enough. ;)
-
Beginning to look like it's probably the main TV/monitor that is the cause of my "residual" 0.3dB loss...........
Oh dear, not as cheap as a cable to replace
...
and just as I was about to hit send... I got the new reply warning, which totally invalidates the above.
Not sure what to say though now about the tuner... or if its actually something being introduced via the distribution amp
-
It's going to be my problem to work around and fix (if I can) or live with.
Just really posting the info here in case it helps anyone else with similar problems.
What the actual precise cause is, and whether/how it can be fixed, I may never know.
Though I do have a little separate 12vDC DVB-T tuner box (intended for my car) to play with, so I'll try that as an input to this monitor, using the same aerial and see what happens!
-
I know off topic but newtron my friend, where is your graph on mydslwebstats :)
have not made an account on mydslwebstats reason being i'll see my stats when the PC is on and that's only 5 or 6 hours a day maybe 12 hours if I'm lucky at the weekend and only have 40GB monthly usage allowance ;)
-
A Faraday Cage will feature in my experiments; I just need to decide what to make it from.........
I also happen to have 2 separate downstairs ring mains in my house, so I'll play with that aspect, though the TV is currently (ha ha) on a different one to the router.
Regarding the F-cage, might I recommend, a cardboard box (large enough to house the device in question), aluminum foil and tape? Simply tape a layer of foil on the inside of the box, and one on the outside and make sure to use a decent overlap if the foil is not "wide" enough to cover the box on one "lane". Bonus points for not aligning the overlaps on in- and outside ;)
Best Regards
M.
-
Yes I have two DECT phones, will this cause the SNRM to drop when the LED/LCD TV in on ?
From http://forum.kitz.co.uk/index.php?topic=14336.0
4. Removed DECT phone from line
-It was permanently robbing 0.3dB SNRM. If the TV was turned on, it became even greedier and took 1.3dB! It was in close proximity to the TV and the master socket, though.
-
Beginning to look like it's probably the main TV/monitor that is the cause of my "residual" 0.3dB loss...........
Or maybe not............
My replacement PSU arrived today and I plugged it in about half an hour ago.
Initial readings suggest that it is "clean" and what's more the SNRM varies not a jot in all the configurations I've tested.
Exactly the same with everything unplugged as when fully operational.
So I'm back to thinking that whatever I'm changing, it varies the load on the PSU and the temporary one I used, whilst nothing like as bad as the one I binned, was actually the source of interference.
Obviously if your TV has the PSU built-in you can't test like I have.
Onwards and upwards! :graduate:
On a negative note, the PN rep I'm communicating with has gone on holiday. I had asked them to put me back on normal DLM but as I'm still on a 6dB profile after several days of total stability I'm doubting that. :-X
-
Initial readings suggest that it is "clean" and what's more the SNRM varies not a jot in all the configurations I've tested.
Excellent news.. lets hope it stays that way. :)
-
Thanks, it's certainly looking good several hours later. 8)
My downstream SNRM is hardly wavering at all, only up and down by 0.1dB from 6.0dB since lunchtime which is amazing.
Oddly the upstream is varying quite a bit between 6.0 and 6.6dB, often in just a few minutes.
-
Just to close this down, I did have to rather beat the PN reps into submission to get my line properly reset, but they did it in the end. It then moved to 3dB pretty quickly (DLM seems to work these days :o ) and stayed that way for some considerable time, many days, without a resync.
I then got hit by the big PN outage last weekend, but it's been totally stable at 3dB since then - almost 3 days now.
Problem solved I think. :graduate:
-
:)