Kitz Forum
Broadband Related => FTTC and FTTP Issues => Topic started by: jaydub on January 18, 2018, 12:27:22 AM
-
It has been noted on here before that I have quite an elevated noise floor, particularly on my downstream (see attached Quiet Line test results)
I'm only a recently paid up member of the MDSW community, so have less than a months history to look at. In that time my attainable rates have not changed very much, but I've managed to dig out my attainable figures from November 21016 when I first had a FTTC connection and my attainable rates were originally US 25.828 Mbps / DS 105.494 Mbps. These have now dropped to US 24.327 Mbps / DS 91.622 Mbps.
Is the probable cause of this change crosstalk at the cabinet?
I still sync at 80/20 so life could be much worse.
However, I would like to understand the likely impact of crosstalk on my connection beyond the impact of sync speed.
I have a history of somewhat inconsistent single thread speeds on the TBB tester. If I do three consecutive tests on TBB, I always tend to get at least one result that is noticeably slower than the fastest.
Could this be down to usage on the line that is providing the cross talk interference? Or is it just down to the vagaries of my connection?
Any thoughts appreciated.
-
. . . (see attached Quiet Line test results) . . .
b*cat thinks he hasn't gone blind, so perhaps you forgot to attach an attachment or provide a link to the image . . .
-
I’ve never heard or experienced crosstalk effecting speed tests, all crosstalk should do is eat away at the SNR.
A drop in attainable would correspond to crosstalk, looking at the graphs you have a significant level of crosstalk however with how close you are to the cabinet I can’t see you dropping below 80.
I’ve attached a side by side of your QLN and mine to highlight how it could be worse, mine is kind of the worse case scenario. With how much crosstalk we both have the chances of more crosstalk are reduced as at this stage every line that effects yours with crosstalk is probably active.
(https://s13.postimg.org/nkntocpnb/AA5_CCD71-2_D60-433_E-_A836-_B7364758776_A.jpg)
-
b*cat thinks he hasn't gone blind, so perhaps you forgot to attach an attachment or provide a link to the image . . .
The joys of posting at a time of night when you should be in bed.
I woke up this morning and thought did you attach the screenshot? Many thanks to S.Stephenson for doing the biz on my behalf.
-
I’ve never heard or experienced crosstalk effecting speed tests, all crosstalk should do is eat away at the SNR.
A drop in attainable would correspond to crosstalk, looking at the graphs you have a significant level of crosstalk however with how close you are to the cabinet I can’t see you dropping below 80.
I’ve attached a side by side of your QLN and mine to highlight how it could be worse, mine is kind of the worse case scenario. With how much crosstalk we both have the chances of more crosstalk are reduced as at this stage every line that effects yours with crosstalk is probably active.
Many thanks for posting my QL stats image.
I am fortunate that the cabinet is the other side of our front garden wall and the telegraph pole fibre and voice cabinets are all pretty close to each other. I'd be amazed if there is more than 30m of copper between us and the the fibre cabinet. Too bad that it's an ECI cab, but we both have that cross to bear.
What is interesting is that my downstream seems to be much more affected by noise than my upstream, whereas it isn't so clear cut for you as your U2 is carrying more noise than mine.
Not sure what this all means in reality, but it's interesting to do the comparisons.
-
Crosstalk definitely can't cause single thread speed issues.
Crosstalk can only really effect throughput if it creates so much noise that it causes the line to have huge numbers of ES. That's more likely to give you packet loss.
As mentioned above crosstalk can have an impact on SNR/QLN/Sync, things on the xDSL layer.
-
Crosstalk definitely can't cause single thread speed issues.
Crosstalk can only really effect throughput if it creates so much noise that it causes the line to have huge numbers of ES. That's more likely to give you packet loss.
As mentioned above crosstalk can have an impact on SNR/QLN/Sync, things on the xDSL layer.
OK. Thanks for the clarification.
As I have spare attainable, my SNR is happily sitting at 9dB both up and down. I also have very few errored seconds, so in my case there doesn't look to be much to worry about.