Basic Internet Connectivity And Modem Help

[dyndragon]

Hey all,

In Schaumburg, IL, losing connection to the internet every 30 seconds to 3 minutes. Cablemodem lights may go out, or sometimes not, and sometimes recovers within a few seconds, or a few minutes. Anyone else experiencing this? All this snow might have something to do with it, maybe.

--Hans 

[EG]

---

dyndragon wrote:

Hey all,

In Schaumburg, IL, losing connection to the internet every 30 seconds to 3 minutes. Cablemodem lights may go out, or sometimes not, and sometimes recovers within a few seconds, or a few minutes. Anyone else experiencing this? All this snow might have something to do with it, maybe.

--Hans 

---

Want advice here ? Then you need to do the self troubleshooting stuff in the links to the sticky topics below, then post the requested stuff:

http://forums.comcast.net/comcastsupport/board/message?board.id=5&message.id=69354&jump=true

http://forums.comcast.net/comcastsupport/board/message?board.id=5&message.id=69355&jump=true

[dyndragon]

Thanks for the link. I actually work as an IT professional for a living, so I've actually gone beyond many of the points on those links already. Router's not the problem, swapped cable modems, made sure of my own wiring, no firewalls interfering, etc etc...

I do note that my upstream power level is at 55dBmV, sometimes up to 65dBmV, so that may be the cause of this. But, it has been like this for as long as I can remember, and I've never had problems before. Nevertheless, I'm having a tech come out wednesday to take a look. All the other reported levels check out.

The only other questionable thing is the wiring outside. We live in a condo complex, and we have the little green "pods" that are in our back lawns that our coax wires go to. They never buried the wires. In fact, I'm pretty sure that my wire, as well as some of our neighbors' wires are just laying in a puddle of water/ice under the snow. They might even be split off of each other, but who knows...

Anyway, was just gathering some intelligence on whether anybody else in my area was having issues, or if it was something to do strictly with my neighborhood and housing units here.

--Hans 

[dyndragon]

Sorry, I forgot to add--

This just started tonight, around 6pm CST (coinsidently, right around the time the snow started to come down hard). Earlier I had no problems.

When i run a continous ping, I just sporadically get time outs to the remote server (i.e. google.com)

Traceroutes don't come up with anything, as the only hop that responds after a network interruption is my own router. 

[EG]

---

dyndragon wrote:

Thanks for the link. I actually work as an IT professional for a living, so I've actually gone beyond many of the points on those links already. Router's not the problem, swapped cable modems, made sure of my own wiring, no firewalls interfering, etc etc...

Being an IT pro doesn't make one an HFC based network pro.

I do note that my upstream power level is at 55dBmV, sometimes up to 65dBmV, so that may be the cause of this.

I'd have to say BINGO here..

55 dB is already too high, and intermittent fluctuations to 65 dB are absurd !

This will definitely cause the modem to re-boot spontaneously which you already stated will cause the modem's indicator lights to behave the way that they are..

Post the contents of the modem's error logs so that we *may* see evidence of a loss of block synch.

The only other questionable thing is the wiring outside. We live in a condo complex, and we have the little green "pods" that are in our back lawns that our coax wires go to. They never buried the wires. In fact, I'm pretty sure that my wire, as well as some of our neighbors' wires are just laying in a puddle of water/ice under the snow. They might even be split off of each other, but who knows...

Guy. This sounds like quite a local issue to me...

---

Message Edited by EG on 01-12-2009 11:33 PM

[dyndragon]

Oh, I don't mean to say I know HFC that well. I am saying though that I know my own network and my own gear doesn't have any issues, that's all!

We'll see what the tech says after wednesday.

I get a lot of these errors:

2009-01-12 21:19:15 3-Critical R005.0 Started Unicast Maintenance Ranging - No Response received - T3 time-out

2009-01-12 20:30:35 3-Critical R004.0 Received Response to Broadcast Maintenance Request, But no Unicast Maintenance opportunities received - T4 timeout

They seem to coinside with the lights on the cablemodem going out. But, they don't coinside with the cablemodem lights NOT going out, but me losing internet connectivity for 5-10 seconds.

Thanks for the help!

--Hans 

[EG]

I wish you the best for a speedy resolution to this issue !

Please post back on your progress !

[andyross]

I'm also in Schaumburg (near Schaumburg/Plum Grove roads) and have had no problems since Dec 26. Mid Oct until then was a nightmare due to the D3 and bandwidth/usage monitoring upgrades.

The upstream issue you mentioned is most likely the primary issue. I've had issues in the past where my upstream was inconsistent and would occasionally max out the modem. Anything over 50 should be considered a problem. Add in the bitter cold and snow recently, and some connections or amps may be flaking out.

Message Edited by andyross on 01-13-2009 04:19 PM

[Armandez]

I got this same exact issue. Tech came out said my Up was all wrong. Apparently I was splitting my cable line too much. Said I should go from a 3way to a 2way. For the **bleep** of it that is what I did and I didn't get anywhere. I live out in Elkhart, Indiana. All that WONDERFUL joyful fluffy white stuff your getting, im getting, but my problem started Sunday. Getting a little annoyed though.

[dyndragon]

So my connection came back, and upstream power is near 51-52dBmV.

Just so I'm understanding this correctly, so I can help the poor, poor tech that has to come out and work in this blizzard, upstream power is the autosensed power necessary for the cablemodem to actually maintain a connection with the Comcast head end.

My house is not a big one, and as far as I know, there are only 3 active cable jacks in my house. I believe at the point of entry, the cable from the outside is put through a 3 way splitter to get to those 3 jacks in my house.

Jack 1 (living room) has a 1-to-2 splitter for my dual tuner card on my computer.

Jack 2 (loft) has no splitter and is directly connected to the cablemodem.

Jack 3 (master bedroom) is unused.

Since I only use 2 of the jacks, will I improve the upstream power used if I change the 3-way splitter at the point of entry to a 2 way splitter?

Are there any other ways to improve upstream power on my end?

How can it be improved on the outside, between my house and the head end?

Thanks!

--Hans 

[andyross]

Many 3-way splitters have unequal losses. One will be -3.5dB, the other two will be -7.0dB. If it's not marked, the -3.5dB is the usually at the end, closest to the end where the input is. Try putting your cable modem feed on that line.

  |

+-+---+

|        |

|      +++

|      |  |

1      2  3

[dyndragon · Accepted solution]

Found it.

There was a hidden splitter in a crawlspace that was wrecking my connection in my loft. Turns out that the 3 way splitter at the service point wasn't even necessary because some moron decided it would be a fine idea to split the signal between the living room TV and the loft cablemodem connection. So the other 2 splits on the 3 way splitter weren't even used.

Removed the 3 way splitter, installed a tap type splitter on the split between living room and loft to give the cablemodem in the loft a stronger signal, and now I'm down to 42dBmV upstream power at my cablemodem.

Just a note, the Motorola Surfboard SB5100 modem report upstream power 2dBmV higher than it really is, so keep that in mind for anyone who's reading that. This was verified with a tech's testing equipment, and also from previous testing of other similar Surfboard modems (update to cover the bases as per EGs post below).

Solved!

--Hans 

Message Edited by dyndragon on 01-14-2009 07:38 PM

[EG]

---

dyndragon wrote:

Found it.

Good to hear !! Happy surfing !!

Just a note, the Motorola Surfboard modems report upstream power 2dBmV higher than it really is, so keep that in mind for anyone who's reading that. This was verified with a tech's testing equipment

I'd like to also add "just a note" to that comment if I may.

That is a bit of a broad statement as there have been, and are many different models of the Moto Surfboards. There are also other variables involved including the accuracy and consistency of his tool (though these tools are much more accurate than any cable modem's measuring mechanism is), this particular testing environment, and the fact that this is only one single sample here.

---