how-to block ads
|
AlEinstein
join:2004-10-08
Hughson, CA
| reply to justbits
Re: Any 5100b technician page DMT/SNR data experts
Bits,
If your graphs are correct, you should not have the "holes" in the graph. The left "hump"(horizontally short)is your upstream and the right "hump" (horizontally long)is your downstream. Each graph you posted, except the first, report "holes" or slices in the downstream portion. On the graph they are the white "slices" where the red is missing so to speak. The overall shape looks decent. But the holes show trouble. 1 skinny one is sometimes OK, but you show more, and one of the graphs is just full of spikes. Spikes are ugly.
Are you CO or RT fed? Just curious. You may have said already.
According to your post, its good that you went right to the nid, homerun with the cat5 etc.
Any reply from posting to Direct? Probably have to wait till Monday morning I guess. Definitely will get some good analysis from those guys on your modem log and graphs.
The bottom line is, assuming outside trouble, (which I think is safe to do in this case) is to have a trouble ticket opened so a tech can dispatch and confirm bits graph holes and monitor the line with his/her test set etc and justify opening a trouble ticket to the line techs to find a pair thats not reflecting these holes and any other issues the tech may find. It may require a vendor meet too, just heads up.
Last thought, the 2Wire modems are great for analysis and troubleshooting. They give max attainable unlike the B modems. Hopefully Efficient will provide a firmware upgrade or the like to add that feature to an otherwise great modem in the B. | |
justbits
More fiber than ATT can handle
Premium
join:2003-01-08
Chicago, IL
 ·AT&T Midwest
| AlEinstein,
I am CO fed. No RT.
After having 3 horrible experiences with having to go through the same "script" with tier 1 tech support, I started reporting all information on the problem to a thread in SBC Direct. It's 5 pages long and detailed.
»[TROUBLE] Speed drop from 2400/443 to 1500/443
There have been 7 ASI trouble tickets so far. A bunch of them were closed with a simple resolution unrelated to the problem.
First visit by ASI: found two bridge taps.
Second visit by Ameritech: couldn't remove bridge tap due to satellite dish on telephone pole. See:
»OT: Satellite dish attached to utility pole?
And:
»Satellite dish on public utility pole? (part 2)
Another visit by Ameritech: Supposedly they came and removed the bridge tap(s).
Third visit by ASI: Saw a really noisy line. Still saw bridge taps. Said that DSLAM was estimating the line at 14,000 feet when in March it was estimated at 9,900 feet. The weather & temperature and humidity were pretty much the same from March until the middle of May. It's not a weather issue.
Fourth visit by Ameritech: Blamed inside wiring, even with a home run installed!
Fifth visit by ASI and Ameritech (last Monday): went through wire bundle and found a replacement pair of wire. Had to go through a bunch of pairs before finding a pair of wires that looked like it could support Pro package. Same problem occurs. This line fails even worse ... when the problem occurred on Tuesday, it couldn't even hold 784/384 and had to be forced to 384/384 to keep sync. With the line forced to 784/384, Friday and Saturday and Sunday no loss of sync occurred, but the intermittent high levels of noise (jagged/spikey graphs) still exist at approximately the same times that they have occurred in the past. We'll see if it doesn't lose sync today or tomorrow.
I think I missed a visit in there... oh wait! They did.  Last Friday, ASI was supposed to show up on-site, but missed the appointment because it was scheduled for 2008, not 2005.
I've seen people quote the Maximum attainable rate by looking at the technician data. There's gotta be a way to calculate/estimate it.
I've been trying to work with the 5100b's hidden XML interface to see if it isn't disabled in the XML interface, but am having no luck. If anyone knows anything about the 5100b XML interface, send me an IM.
--justbits | |
|
