gbh62 wrote:
Is this general consensus that a "reliable" 32/5 profile will enable a customer to have a 3HD/1SD service at some point? SomeJoe has posted some great stats in regards to the various profiles. Those stats seem to indicate that a minimum 40Mbs max sync on a "clean" copper pair will be all that's needed to ultimately get 3HD/1SD.
Is my interpretation of this correct?
If it ends up being a stable 32/5 profile, there should be no problem w/a 3d HD. The added 7M in the
profile download is to handle that 3d HD (and/or 24M internet). Think the 40M is an arbitrary point, a
user w/a 38M max sync and very clean line may get it and another user w/42M max sync and a noisy
line won't get it or gets 2286 or 4484 on the upload side. ATT is trying to give it to as many users as
possible, w/in the guidelines, so it will be stable and not generate tech calls. 
Chris
Please NO SD stretch-o-vision or 480 SD HD Channels
1-800-983-2811 to avoid Mr. Voice Recognition
YRMV IMHO Simply a U-verse user, nothing more
SomeJoe7777
SomeJoe7777
SomeJoe7777
It might be a good idea to make 2 graphs, one for VDSL and one for VDSL2. In my case on VDSL my
current noise margin was 26.5 and on VDSL2 it was 24.5 (yesterday & today it is 25). This could be
statistically important as I've seen other posts that their noise margins went down a few points too.
Chris
Please NO SD stretch-o-vision or 480 SD HD Channels
1-800-983-2811 to avoid Mr. Voice Recognition
YRMV IMHO Simply a U-verse user, nothing more
SomeJoe7777
mibrnsurg wrote:It might be a good idea to make 2 graphs, one for VDSL and one for VDSL2. In my case on VDSL my
current noise margin was 26.5 and on VDSL2 it was 24.5 (yesterday & today it is 25). This could be
statistically important as I've seen other posts that their noise margins went down a few points too.
I thought about that, but I've also noticed that when people move from VDSL1 to VDSL2, their noise margin goes down but the max line rate goes down also. Thus, the data point moves, but it stays on the profile line.
I think the difference there might not show up in the graph since the VDSL2 data points don't seem to form their own profile line vs. VDSL1.
SomeJoe7777
Re: Line Stats Analysis Results
OK, here is a revised Noise Margin vs. Max Line Rate graph, now that it is known that the data shows a correlation between profiles rather than wire gauge.
This graph shows a bit more information than the previous one.
For properly working service that won't have a lot of errors, you need two things:
1) Sufficient max line rate so that the line capacity doesn't exceed approximately 80% when compared to your profile rate. Line capacities exceeding approximately 80% causes the line to begin to take some errors.
Line Capacity = (Downstream Profile Rate / Max Line Rate) * 100
2) Sufficient noise margin such that the line does not take errors. From people's stats, I am estimating the minimum required noise margin to be around 12.0 dB.
These two parameters gives rise to the two red areas on the graph, where the noise margin is too low for reliable service, or the max line rate is too low even for the 19/2 profile. The minimum required max line rate for the 19/2 profile is given as 21800 kbps. This is where the 19/2 profile line (derived from customer data) hits the 12 dB noise margin, for a line capacity of 88%. This line capacity is a little high, but since the alternative is no service, it may turn out to be sufficient.
Similarly, the 25/2 profile line intersects the 12 dB noise margin limit at 30400 kbps, establishing that as the minimum required max line rate for the 25/2 profile, at a line capacity of 83%.
Finally, the 32/5 profile line intersects the 12 dB noise margin limit at 39700 kbps, establishing that as the minimum required max line rate for the 32/5 profile, at a line capacity of 81%.
What is very interesting here is how we now know that the noise margin changes significantly with the selected profile rate. We also know that the max line rate stays relatively constant when the profile rate is changed (see djrobx's data a few posts above). Because of this nearly constant max line rate and the consistency of the customer data which has given us the profile lines, we can now predict where you will end up on the graph when your profile rate is changed.
For example, let's look at the customer who is currently on the 25/2 profile line (red) at an 18 dB noise margin and 41000 kbps max rate. When this customer is upgraded to 32/5, the max line rate will stay nearly constant at 41000 kbps (it may go down slightly, no more than a few hundred kbps), but his service point will jump horizontally to the left (constant max line rate) to land on the blue 32/5 profile line at a much lower noise margin. We can predict from the graph that his noise margin will be around 13.0 dB.
Since we know what the new noise margin will be due to the profile lines, we can now also identify by noise margin alone those people who should be able to move to a higher profile. For example, people on the 19/2 profile who exceed a noise margin of 19.0 dB should be able to move to the 25/2 profile, since their new noise margin will be above 12.0 dB. Similarly, people on the 25/2 profile whose noise margin exceeds 16.5 dB should be able to move to the 32/5 profile, again because their new noise margin will be above 12.0 dB.
And just for fun, we can also estimate that those people on the 32/5 profile with a noise margin exceeding some value would theoretically be able to move to an even higher profile. Assuming the noise margin drops 6 dB for an increase in the profile rate of 8-10 Mbps, we could theoretically predict the existence of a new 40/5 profile that would be eligible for rollout to people on 32/5 who exceed 18 dB of noise margin and 50000 kbps of max line rate.
Now, since noise margin and max line rate are locked together on these profile lines, predicting whether a person has a noisy line or a bridge tap becomes more problematic. The Dohrenburg DMT plot can be instrumental here. Removing a bridge tap or removing a noise source can raise the max line rate (and noise margin with it), possible allowing someone to upgrade their profile. However, this has to be done on a case-by-case basis, as there is no guarantee how much additional max line rate/noise margin you can achieve.
SomeJoe7777
SomeJoe7777
Diamond_Jim wrote:
SomeJoe7777,
I have ANOTHER question for you amigo...
Why do you have Uverse?
I mean, with as much time as you put in to helping all of us on this forum do you even have time to watch a TV?
Found 3800HG HTML version: 2
GOOD
Your gateway is using profile 25208
The reported distance from the VRAD is: unknown distance.
Probably pair-bonding.
Based on your logs your last retrain SNR average was: 13.5 db
Pair bonding profile. Most likely your line will not qualify for the 25 or 32 Mbps profiles. Based on your gateway's log history the average maximum attainable (estimated) sync rate is: 32683 Kbps. Remember that this number is an estimate based on number of downstream bits, tone SNR (signal noise ratio) and attenuation.
Analyzing bitload distribution:
Tone [0-31] Total bits in extended upstream band = 0 ............. 0.00% of the total upstream bits
Tone [32-255] Total bits in extended downstream band = 2002 ......... 27.50% of the total downstream bits
Tone [256-869] Total bits in VDSL downstream band 1 = 4747 ......... 65.20% of the total downstream bits
Tone [870-1205] Total bits in VDSL upstream band 1 = 747 ......... 100.00% of the total upstream bits
Tone [1206-1972] Total bits in VDSL downstream band 2 = 532 ........... 7.31% of the total downstream bits
Total bits in all upstream bands = 0 + 747 = 747
Total bits in all downstream bands = 2002 + 4747 + 532 = 7281
Last tone with bits assigned: 1443 (frequency 6.222938 MHz)
BAD. Your gateway have no bits allocated above 7.8 MHz
Analyzing each VDSL band:
Tones with assigned bits in extended upstream band start with tone: none and end with tone: none Searching for problems ...
No problems found.
Tones with assigned bits in extended downstream band start with tone: 32 and end with tone: 255 Searching for problems ...
Found a potential source of interference! zero bit load found in tone 64 (frequency 0.276000) [ ADSL pilot tone ]
Tones with assigned bits in VDSL downstream band 1 start with tone: 256 and end with tone: 851 Searching for problems ...
Found a potential source of interference! zero bit load found in tone 322 (frequency 1.388625)
Found a potential source of interference! zero bit load found in tone 323 (frequency 1.392938)
Found a potential source of interference! low bit load found in tone 336 (frequency 1.449000 MHz)
Found a potential source of interference! zero bit load found in tone 392 (frequency 1.690500)
Found a potential source of interference! zero bit load found in tone 576 (frequency 2.484000) [ VDSL2 normal value ]
Hey Joe. I know I'm late but what do you think here.
Current Rate: 25208 kbs 2040 kbs Max Rate: 47815 kbs Not Available Current Connection: Current Noise Margin: 20.5 dB Not Available Current Attenuation: 15.0 dB Not Available Current Output Power: 14.2 dBm -11.8 dBmCurrent Rate: 25208 kbs 2040 kbs Max Rate: 47815 kbs Not Available Current Connection: Current Noise Margin: 20.5 dB Not Available Current Attenuation: 15.0 dB Not Available Current Output Power: 14.2 dBm -11.8 dBm
SomeJoe7777
ChrisPC,
Your new DMT plot looks better. There are more high frequencies present now. If you got this result by plugging the RG into a different jack in your house, that indicates what you need is a home run of Cat5e from the NID to your RG, and bypass your internal house wiring. Depending on the phone wiring you have now, you could pick up additional line rate and/or noise margin.
dapperdan,
Looks like a bridge tap to me:
The line is really noisy and taking errors as well. You need to have tech support roll a truck to come take a look and see if they can get I&R to come out and condition the line.
jgnall,
No problem with your line, and yes, you're eligible for 32/5. Your may sync rate is over 53 Mbps.
tthomaso,
Yes, you're eligible for 32/5.
jjs2132,
You have the 25 Mbps profile, but your max sync rate and noise margin is too low for the 32/5. I don't see a bridge tap, but your line is very noisy, and that is probably reducing your max sync rate. You may want to get a home run installed from the NID to the RG if you don't have one already.
camp80,
Perfectly working line.
