General Discussion
  >> Fibre Broadband


Register (or login) on our website and you will not see this ad.


Pages in this thread: 1 | 2 | 3 | (show all)   Print Thread
Standard User WilliamGrimsley
(committed) Sun 19-Mar-17 14:47:11
Print Post

Strange Downstream SNR Margin


[link to this post]
 
Hi guys,

I can only think this is caused by some interference, but does anyone who has better understanding of this than myself know what could be causing this? Should I be concerned?

https://i.gyazo.com/71f0472065c8f24c21e64b8714a7f8bb...

Cheers!

William
Standard User troublegum
(member) Sun 19-Mar-17 16:54:49
Print Post

Re: Strange Downstream SNR Margin


[re: WilliamGrimsley] [link to this post]
 
Not this again.

It's dropped from what, 4.2 to 3.8 a few times? If it didn't cause any issues at the time, then no it's nothing to be concerned about.

I think you need to switch off all your monitoring software and stop looking for faults.
Standard User WilliamGrimsley
(committed) Sun 19-Mar-17 16:56:03
Print Post

Re: Strange Downstream SNR Margin


[re: troublegum] [link to this post]
 
I've turned on the monitoring software for diagnosis. There's a fault there, but it's very low scale. I should probably investigate it but it's not affecting service yet.

Edited by WilliamGrimsley (Sun 19-Mar-17 17:00:24)


Register (or login) on our website and you will not see this ad.

Standard User RobertoS
(elder) Sun 19-Mar-17 17:26:45
Print Post

Re: Strange Downstream SNR Margin


[re: WilliamGrimsley] [link to this post]
 
Particularly on ADSLx, less so on FTTC, downstream noise margin is very likely to fall by anything up to 3dB overnight. The interference hitting the copper between you and the DSLAM/MSAN is greater at night.

A simple fact. It's to do with reflections off one of the atmospheric belts when the sun is down. Which is why it is always recommended to connect well inside daylight hours for highest speed, and at night for greatest stability.

Think about it. For decades the default sync-time noise margin was 6dB. SpeedTouch modems would lose connection if it went below 3dM. Netgear would hold it as low as 0dB or even sometimes -1dB.

Why did BT Wholesale Sky and TalkTalk all set it at 6dB? Because it was deemed to be the lowest safe level. Nowadays with improved modem technology at both ends, BT Wholesale allow DLM to set connection time SNRM at 3dB. Openreach are trialling 3-5dB as you know for VDSL2.

It's a safety margin. It is expected to fall. It is rarely a fault, and if it is there will be much more serious and noticeable symptoms that a slight overnight decrease. Remember this page?

Kindness isn't going to cure the world of all its awfulness but it's a good place to begin. Daisy Ridley.
My broadband basic info/help site - www.robertos.me.uk. Domains, site and mail hosting - Tsohost.
Connection - AAISP Home::1 80/20. Sync 65618/13914Kbps @ 600m. BQMs - IPv4 & IPv6
Standard User WilliamGrimsley
(committed) Sun 19-Mar-17 17:33:11
Print Post

Re: Strange Downstream SNR Margin


[re: RobertoS] [link to this post]
 
I understand the normal fluctuations in SNR margin, but this is different, the downstream SNR margin is regularly fluctuating in short intervals.

Edited by WilliamGrimsley (Sun 19-Mar-17 17:34:08)

Standard User lee111s
(experienced) Sun 19-Mar-17 21:17:06
Print Post

Re: Strange Downstream SNR Margin


[re: WilliamGrimsley] [link to this post]
 
Is it service affecting?
Standard User WilliamGrimsley
(committed) Sun 19-Mar-17 21:24:45
Print Post

Re: Strange Downstream SNR Margin


[re: lee111s] [link to this post]
 
Yes, in terms of throughput.

Edited by WilliamGrimsley (Sun 19-Mar-17 21:26:30)

Standard User MrDiode
(learned) Sun 19-Mar-17 21:38:35
Print Post

Re: Strange Downstream SNR Margin


[re: WilliamGrimsley] [link to this post]
 
Sudden drops of snrm always makes me think of close by equipment interfering with the modem/router when switched on
Standard User RobertoS
(elder) Mon 20-Mar-17 01:07:27
Print Post

Re: Strange Downstream SNR Margin


[re: MrDiode] [link to this post]
 
DECT phones are particularly suspect in my case. Cause exactly what we see on the graph linked in the OP.

Dodgy microwave seals perhaps. Who knows?

Kindness isn't going to cure the world of all its awfulness but it's a good place to begin. Daisy Ridley.
My broadband basic info/help site - www.robertos.me.uk. Domains, site and mail hosting - Tsohost.
Connection - AAISP Home::1 80/20. Sync 65618/13914Kbps @ 600m. BQMs - IPv4 & IPv6
Standard User RobertoS
(elder) Mon 20-Mar-17 01:08:55
Print Post

Re: Strange Downstream SNR Margin


[re: WilliamGrimsley] [link to this post]
 
In reply to a post by WilliamGrimsley:
Yes, in terms of throughput.
Have you got evidence if that? At the specific times?

Kindness isn't going to cure the world of all its awfulness but it's a good place to begin. Daisy Ridley.
My broadband basic info/help site - www.robertos.me.uk. Domains, site and mail hosting - Tsohost.
Connection - AAISP Home::1 80/20. Sync 65618/13914Kbps @ 600m. BQMs - IPv4 & IPv6
Standard User WilliamGrimsley
(committed) Mon 20-Mar-17 08:23:15
Print Post

Re: Strange Downstream SNR Margin


[re: RobertoS] [link to this post]
 
I can confirm it's nothing in the house and the phone was not used during these blips. Anyway, it's been dropped to 4 dB now so obviously it's not causing a problem.
Standard User TheEulerID
(experienced) Mon 20-Mar-17 08:43:43
Print Post

Re: Strange Downstream SNR Margin


[re: WilliamGrimsley] [link to this post]
 
A reduction in the SNR margin will only cause throughput issues if there is an increase in uncorrectable errors requiring a retransmit. It would be best to monitor error rates to see if there is a correlation.

As for what's causing fluctuations, both long and short term, then the former is normally down to things like atmospherics whilst short term fluctuation cans be caused by any number of sources of EMF.
Standard User WilliamGrimsley
(committed) Mon 20-Mar-17 08:47:26
Print Post

Re: Strange Downstream SNR Margin


[re: TheEulerID] [link to this post]
 
I understand that, there were no errors but the amount G.INP retransmits going on would suggest the reduction in throughput.
Standard User Icaras
(experienced) Mon 20-Mar-17 18:20:21
Print Post

Re: Strange Downstream SNR Margin


[re: WilliamGrimsley] [link to this post]
 
In reply to a post by WilliamGrimsley:
I understand that, there were no errors but the amount G.INP retransmits going on would suggest the reduction in throughput.


You really need to stop obsessing about your connection.
Standard User WilliamGrimsley
(committed) Mon 20-Mar-17 18:41:29
Print Post

Re: Strange Downstream SNR Margin


[re: Icaras] [link to this post]
 
It may come over that way, but I'm not.
Standard User TheEulerID
(experienced) Mon 20-Mar-17 18:56:14
Print Post

Re: Strange Downstream SNR Margin


[re: WilliamGrimsley] [link to this post]
 
g.inp retransmissions are errors, just corrected ones, so if that is getting very high it could affect throughput to some extent, but high is the operative word. It would be transparent to the IP layer, so would not cause the sort of problems you see with packet loss, so it would never be that marked.
Standard User WilliamGrimsley
(committed) Mon 20-Mar-17 21:01:35
Print Post

Re: Strange Downstream SNR Margin


[re: TheEulerID] [link to this post]
 
I already know that (that's meant in a positive not targeting way). smile
Standard User MrDiode
(learned) Mon 20-Mar-17 23:55:40
Print Post

Re: Strange Downstream SNR Margin


[re: WilliamGrimsley] [link to this post]
 
Just looking into your history you mentioned that you use ethernet power line adapters they are not the cleanest way to connect to a modem it has drawbacks increased FEC errors which is extra noise and can only guess that is what is causing your upstream to bounce from g.inp to fastpath and downstrean to sometimes enter the interleaving profile.

Would get rid of any ethernet powerline adapters and see how your line performs over 2 weeks.
Standard User WilliamGrimsley
(committed) Tue 21-Mar-17 09:04:17
Print Post

Re: Strange Downstream SNR Margin


[re: MrDiode] [link to this post]
 
I wish I didn't have to use them, but unfortunately we have to for our BT TV service.
Standard User lee111s
(experienced) Tue 21-Mar-17 09:12:29
Print Post

Re: Strange Downstream SNR Margin


[re: WilliamGrimsley] [link to this post]
 
NTE shift?
Standard User WilliamGrimsley
(committed) Tue 21-Mar-17 09:16:15
Print Post

Re: Strange Downstream SNR Margin


[re: lee111s] [link to this post]
 
Too costly and I need the router next to the PC anyway. I have disconnected the powerline adapters several times and it makes no difference to line performance.

Edited by WilliamGrimsley (Tue 21-Mar-17 09:16:53)

Standard User MHC
(sensei) Tue 21-Mar-17 09:37:01
Print Post

Re: Strange Downstream SNR Margin


[re: MrDiode] [link to this post]
 
Care to explain how use of something on the Ethernet side of a modem will generate increased FEC errors on the VDSL side of the modem? A small potential for radiated and conducted RFI but that is all.


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

M H C


taurus excreta cerebrum vincit
Standard User ggremlin
(experienced) Fri 24-Mar-17 00:02:33
Print Post

Re: Strange Downstream SNR Margin


[re: MHC] [link to this post]
 
powerline adapters/ networking use similar frequencies to vdsl, so could cause problems
Standard User ukhardy07
(knowledge is power) Fri 24-Mar-17 00:10:08
Print Post

Re: Strange Downstream SNR Margin


[re: WilliamGrimsley] [link to this post]
 
In reply to a post by WilliamGrimsley:
Too costly and I need the router next to the PC anyway. I have disconnected the powerline adapters several times and it makes no difference to line performance.
Was a fault developing on your line back in Feb - no.

Do you have a fault now - no.

Are you obsessing over this - yes

Disable the stats it aint good for you, set an admin password you'll forget, grab a beer, forget about this nonsense.
Standard User Zarjaz
(eat-sleep-adslguide) Fri 24-Mar-17 06:39:08
Print Post

Re: Strange Downstream SNR Margin


[re: ukhardy07] [link to this post]
 
+1

Standard User WilliamGrimsley
(committed) Fri 24-Mar-17 08:07:23
Print Post

Re: Strange Downstream SNR Margin


[re: ukhardy07] [link to this post]
 
I'm not monitoring my line stats anymore, I just occasionally view them through the web GUI which I think is fine. If you have a problem with this then tough, because it's my decision and right to ask for help. I am only showing concern and interest.

Edited by WilliamGrimsley (Fri 24-Mar-17 08:11:15)

Standard User ukhardy07
(knowledge is power) Fri 24-Mar-17 09:41:28
Print Post

Re: Strange Downstream SNR Margin


[re: WilliamGrimsley] [link to this post]
 
In reply to a post by WilliamGrimsley:
I'm not monitoring my line stats anymore, I just occasionally view them through the web GUI which I think is fine. If you have a problem with this then tough, because it's my decision and right to ask for help. I am only showing concern and interest.
Happy to help wherever I can, as are the other guys here. I noted in a previous thread you struggled to accept nothing was wrong... Saw this going down the same path.
Standard User WilliamGrimsley
(committed) Fri 24-Mar-17 10:00:52
Print Post

Re: Strange Downstream SNR Margin


[re: ukhardy07] [link to this post]
 
I didn't struggle to accept, something was wrong and there are occasional spikes of severe interference on my line so something isn't right somewhere.
Standard User WilliamGrimsley
(committed) Sat 25-Mar-17 10:16:29
Print Post

Re: Strange Downstream SNR Margin


[re: WilliamGrimsley] [link to this post]
 
Now down to 3 dB, pretty happy we're back where we were. laugh
Pages in this thread: 1 | 2 | 3 | (show all)   Print Thread

Jump to