04-26-2013 09:58:58 AM
Last Saturday, my Uverse service went down. I called support and they scheduled a tech for later that day. While I was waiting, I noticed a Uverse truck on my street. I spoke to that tech and he said he'd been working on a house near mine and he might have caused my service to go out. (He actually seemed pretty inexperienced.) I told him I had called tech support, so he didn't offer to investigate.
When the tech support guy arrived, he checked the VRAD and found the first guy had disconnected my service from the port. He reconnected it. He then came into my house and saw I had an older 2WIRE 3800, so he gave me a new 3801. While he was there, he checked my line quality and said it looked good. No sign of a bridge tap.
I fired up Uverse Realtime and showed him the program. He'd never heard of it, and thought it was very cool. The first thing I noticed was that the bitloading graph for the 3801 RG was much better than the old RG. So by the time he left, I was happily back in business.
However, since then, I've had lots of communication errors in a short period of time. The latest was this morning, when I turned on the TV and there was no signal for a few seconds. In the three years I've had Uverse, I've had rare link retrains and other errors, but never this much in such a short period of time.
Here are my UVRealtime screenshots. The bitloading graph shows areas of RF interference, but they've been there for the three years I've had Uverse and I've never been able to track down the source. (The prior 3800 RG showed them as well.)
I thought that the newer 3801 RG was supposed to be more reliable in maintaining a connection, so I'm puzzled by this. Any comments or suggestions would be much appreciated.
04-26-2013 10:13:07 AM - edited 04-26-2013 10:13:52 AM
Ponzi - I can not explain it, but I had a similar experience exactly a year ago. Replaced 3800 with 3801 - At first, it was worse than the 3800. In the morning it was fine & has been ever since. Hope yours calms down. If not maybe it is a refurb & needs to be replaced. Maybe make sure the coax connect is tight.
04-26-2013 05:56:45 PM
The 3801 may have been overzealous in trying to load up some frequencies and is having to back down on them. Hopefully it's learning its lesson and will leave them alone in the future. It's also possible that you got changed from one port (or even card) to another in the VRAD by virtue of the disconnect and the reconnect. The new one may not be as good as the old one.
04-26-2013 08:55:48 PM
aviewer: Thanks for the input. I've checked the connections and all are tight. I don't think the RG was a refurb.
JefferMC: It's possible that I was connected to another port or card at the VRAD. I have a different outside IP address, and I don't know if this means a different port or card was used. I didn't know that the RG will remember the best frequencies. I assumed it wiped the slate clean each time it rebooted. I do notice that the power output will change each time the RG reboots and the bitloading graph looks a little different. Not necesssarily fewer frequencies used, just a slight reallocation here and there.
One comment that the first tech (the one who disconnected me in the first place) said was a little disconcerting. He said that the VRAD was a mess. I'm not exactly sure what he meant by that, but it didn't sound good.
04-27-2013 05:17:19 AM
I didn't know that the RG will remember the best frequencies. I assumed it wiped the slate clean each time it rebooted. I do notice that the power output will change each time the RG reboots and the bitloading graph looks a little different. Not necesssarily fewer frequencies used, just a slight reallocation here and there.
I meant once it retrain occurs, it should start blocking frequencies that it later has issues with up until the next retrain/reboot. I can see some advantages and disadvantages to having it remember problem bands longer than that, e.g. between factory resets, but I do not think it does that.