jstearns68
Premium
join:2009-04-18
Columbia, IL
|  DSL connection issues - SNR Fluctuations
Is it odd to see my downstream SNR stats jumping around like crazy?? At best it's sitting at 8.5 (which I gather from posts on this forum isn't great to begin with) but when I'm looking at my modem stats I see it drop as low as 0.5, -1 and anywhere in between. The only time I've seen it stabilize is when connecting it unfiltered directly to the NID. I've been instructed to do this by tech support. I feel like the this lack of stability could be causing my intermittent disconnects.
Is it best practice to use a DSL splitter close to the NID? I've installed mine approx 50' from the NID. So I'm running Cat5 from the NID across my garage, through my basement to my distro panel for my phones and all my AV stuff.
The modem is only 3' away from the splitter and I have a router nearby to distibute the Inet connection throughout my house. |
|
tschmidt
Premium,MVM
join:2000-11-12
Milford, NH
 ·Hollis Hosting
 ·Verizon Online DSL
·Fairpoint Communic..
| Whole house POTS/Splitter splitter does a better jobs isolating Voice from DSL. A splitter will not solve all problems. For example several years ago lost DSL completely, phone service was fine. We use a splitter. Turned out several phone jacks were corroded and bridged line 1 to line 2. Was not severe enough to affect phone but completely took out DSL.
Ideally margin should be above 10 dB, between 10-6 dB line will be susceptible to disturbances. At 6 or under modem will not be able to maintain sync.
Doesn't really matter where you locate splitter but most convenient location is typically near the NID to make rerouting existing phone wiring easy. Are you absolutely sure everything except DSL modem is behind splitter?
Using a splitter there should be no significant difference between connecting modem directly to NID and data terminals on the splitter. As a test record download/upload stats connected directly to the NID, then normal DSL connection, then disconnect inside wiring from "phone" terminals on the Splitter. Reboot modem for each test. That should flag where the problem lies.
/tom
|
|
jstearns68
Premium
join:2009-04-18
Columbia, IL
| reply to jstearns68
well I had the phone company come out and install the whole house POTs splitter. The tech tested for a ton of different things. The stress tested my lines to the CO, looked at the noise stats at four different points all of which were consistant at 6.5 - 7.0 snr downstream. This lead me to believe that my cat5 runs are good and I'm not adding noise between the NID and my modem.
According to the tech I'm roughtly 13k + from the CO. He admitted the 6 SNR was low.
So the tech leaves and I setup my modem to watch the SNR stats more and things look stable at 6 - 6.5 until about the two hour mark and I start seeing big drops again.
I'm thinking my last step here is to drop the speed of my service. Currently I'm at 5Mbps down / 765 Kbps up. The next lowest package is 3 down / 512 up.
Is this significant enough of a change to improve the SNR? I have a feeling it's going to be a 'try it and see' situation. |
|
tschmidt
Premium,MVM
join:2000-11-12
Milford, NH
·Hollis Hosting
·Verizon Online DSL
·Fairpoint Communic..
| 6 dB noise margin is pretty low. Reducing speed will improve margin.
Here in Verizon land at 13,000 feet you would not even qualify for 3000/768, cutoff is 12,000.
Who is your ISP? Can you post modem stats?
It wouldn't hurt to temporally connect your modem directly to the Phone Company NID test jack. That disconnects inside wiring from phone line. While a whole house POTS/DSL splitter does a much better job isolating DSL and tech tested your residence nice to make sure for yourself. If stats improve something within your residence is degrading performance either wiring or equipment.
My guess is you will not see any improvement connected directly to NID. Reducing speed is the only other way to improve stability.
/tom
|
|
