I am the son of and IT staff for an Xfinity customer.
Starting a couple of weeks ago, my mother experienced occasional service dropouts. Her computer could access the router and cable modem interfaces, which showed nothing unusual; the only issue was that the cable modem couldn't access the wider internet. The problem would come and go, which of course made it more difficult to diagnose.
Trying an alternative modem, patch cable, and router didn't fix the problem. Changing the modem did get me the "Xinfinity wants you to verify that you're really you" page, which I went through.
Then the problem became seemingly permanent. When rebooted, the modem would now get to "upstream ranging" and remain "in progress".
The next day, it recovered, and we again got the "equipment change" verification page. However, we got the same page the next day. And again this morning. This despite the fact that we hadn't swapped in any new gear during that time.
After verifying, we appear to be okay for the rest of the day. But is there a way to reassure Xfinity that it need not repeat this verification step every day?
Solved! Go to Solution.
What do the modem's signal stats look like ? Try getting them here http://192.168.100.1 or here http://10.0.0.1
Please post the *Downstream Power Level*, the *Upstream Power Level*, and the *SNR* (Signal to Noise Ratio) numbers.
Downstream power level = -1.8 dBmV, signal/noise = 37.7 dB, at 573 MHz
Upstream power level = 50.7 dBmV, at 14.8 MHz
Refreshing gets changes of a few tenths of a dBmV.
The upstream power is on the high side and it may be intermittently fluctuating even higher to out of spec levels. That can cause random disconnects, spontaneous re-booting of the modem, speed and latency problems, loss of provisioning.
In a self troubleshooting effort to try to obtain better connectivity / more wiggle room, check to see if there are there any excess/unneeded coax cable splitters in the line leading to the modem that can be eliminated/re-configured. Any splitters that remain should be high quality and cable rated for 5-1002 MHz, bi-directional, and no gold colored garbage types like GE, RadioShack, RCA, Philips, Leviton, Magnavox, and Rocketfish from big box stores like Home Depot, Lowes, Target, Wal-Mart etc. Splitters should be swapped with known to be good / new ones to test
If there aren't any unneeded splitters that can be eliminated and if your coax wiring setup can't be reconfigured so that there is a single two way splitter connected directly off of the drop from the street/pole with one port feeding the modem and the other port feeding the rest of the house/equipment with additional splits as needed, and you've checked all the wiring and fittings for integrity and tightness and refresh them by taking them apart then check for and clean off any corrosion / oxidation on the center wire and put them back together again, then perhaps it's best to book a tech visit to investigate and correct.
Thank you. I'm not currently at my mother's, but am reasonably certain the gear is of decent quality and there aren't undue splitters. I'll give it a check. Taking connectors apart, cleaning, and reassembling shouldn't pose too much difficulty. Between the modem coax connector to the wall and the outside world, there are connectors that haven't been touched in years; oxidation is indeed possible... will post again after trying all that.
O/k if that doesn't help the constant reactivation problem you can also try doing this;
Call 1-800-comcast and ask a rep to completely remove the modem from your account and to "re-provision" it from scratch. Also ask them to confirm that the provisioning and account data in their database is correct and complete and that the correct configuration / bootfile for your subscribed to speed tier is being pushed to the modem. Good luck with it !!
OK, got back to my mother's place, disassembled some rather clean-looking connectors, cleaned them anyway, and reassembled. It then occurred to me that I'd swapped cable modems as part of this, and had therefore disconnected coax from the old one and connected it to the new one... finger-tight. Not a good idea, of course, and could easily explain poor signal. I used a wrench this time. Upstream power level is 47.2 dBmV and holding very steady (no changes when I refresh). S/N is 37.8 dB, downstream power level 1.0 dBmV.
A bit of searching suggests that 47.2 ought to be about right, and 50.7 was, in fact, marginal (seems to be much disagreement on what a "good" level is). Fingers crossed that we may have success here... won't really know unless we go a bit without resets.