VDSL Performance & Tuning

Forums › Snap › VDSL Performance & Tuning

1 | ... | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | 50 | 51 | 52 | 53 | 54 | 55 | 56 | 57 | 58 | 59 | ... | 87

845 posts

Ultimate Geek

Trusted

Reply # 736187 21-Dec-2012 23:36
Well this is odd! I've got my SNRM set to 13, but I'm getting up to 15 on the line!

Dairusire

39 posts

Geek

Reply # 736190 22-Dec-2012 00:08
Well, I honestly do think I will need to do a new Line drop from the demarc and also get a Master filter. Considering what you fellas have all said seems probably the best thing to do, with apparent degradation of speed etc because of jacks in the walls etc. And maybe the fact my house is 50 years old lol.

Reido

26 posts

Geek

Reply # 736609 23-Dec-2012 17:14
Hi again, just wondering how to tell what DLM you are on?

sidefx

1072 posts

Uber Geek

Reply # 736612 23-Dec-2012 17:19
Reido: Hi again, just wondering how to tell what DLM you are on? Look at your downstream\receive latency and INP(impulse noise protection) and match it to the table in the first post: DLM-1 1ms 0INP DLM-2 8ms 0INP DLM-4 8ms 1INP DLM-6 8ms 2INP DLM-3 16ms 0INP DLM-5 16ms 1INP DLM-7 16ms 2INP DLM-8 16ms 4INP

Reido

26 posts

Geek

Reply # 736615 23-Dec-2012 17:33
So DLM1 now? Whats the main difference of DLM? latency / stability trade offs?

stevehodge

159 posts

Master Geek

Reply # 736618 23-Dec-2012 17:47
Reido: So DLM1 now? Whats the main difference of DLM? latency / stability trade offs? Yes, that's DLM-1. I don't think anyone has experienced instability due to the DLM level. The settings Chorus are using seem to be quite conservative. The lower latency and INP levels will give you faster sync speed and lower latency but higher error rate. If your error rate is too high then you'll drop down to a less aggressive DLM level. This really only matters if latency is important to you, i.e. if you do a lot of online gaming. We don't have a good idea of what error level is too high to stay on DLM-1 - previously the cutoff seemed to be about 0.5 CRC error per minute but lately people seem to be staying on DLM-1 with much higher error rates.

SamF

845 posts

Ultimate Geek

Trusted

Reply # 736652 23-Dec-2012 19:21
sidefx, did you mean: "The higher latency and INP levels will give you faster sync speed and lower latency will give you higher error rate" ?

stevehodge

159 posts

Master Geek

Reply # 736656 23-Dec-2012 19:35
SamF: sidefx, did you mean: "The higher latency and INP levels will give you faster sync speed and lower latency will give you higher error rate" ? That was me, not sidefx, and no, that's not what I meant. Lower latency and INP (i.e. "better" DLM profile) will give you three things: higher sync rate, lower latency (obviously), and higher error rate.

SamF

845 posts

Ultimate Geek

Trusted

Reply # 736681 23-Dec-2012 20:39
Lol, sorry man, too much eggnog already! :D Right, gotcha, right you are.

Jekkyl

41 posts

Geek

Reply # 737654 27-Dec-2012 11:58
Chorus reckon a faulty modem because they can see the errors coming from my end(New Solid Cat5E cable has been run from ETP to Jack. Going to try a straight RJ11 on the end of the CAT to 100% rule out internal wiring.) SNAP have no test modems to send out currently...so not looking good to getting this resolved anytime soon! I really didn't want to start playing with the stability settings for what seems like a fault as it drastically reduces the Speed, I lost 1/4 of the speed just dropping it down to balanced and still had quite a lot of errors. Anyone have any other thoughts?

quakeguy

79 posts

Master Geek

Trusted

Reply # 737698 27-Dec-2012 14:13
Jekkyl: Chorus reckon a faulty modem because they can see the errors coming from my end(New Solid Cat5E cable has been run from ETP to Jack. Going to try a straight RJ11 on the end of the CAT to 100% rule out internal wiring.) SNAP have no test modems to send out currently...so not looking good to getting this resolved anytime soon! I really didn't want to start playing with the stability settings for what seems like a fault as it drastically reduces the Speed, I lost 1/4 of the speed just dropping it down to balanced and still had quite a lot of errors. Anyone have any other thoughts? Interesting. Is your modem (or line) near something that might generate high frequency noise, like a CRT, television, ...? Tim.

Jekkyl

41 posts

Geek

Reply # 737720 27-Dec-2012 15:22
Na nothing like that in the room. Ran the cable through the roof away from any other cables as best as I could

fellaintga

172 posts

Master Geek

Reply # 737762 27-Dec-2012 17:48
Jekkyl: Chorus reckon a faulty modem because they can see the errors coming from my end(New Solid Cat5E cable has been run from ETP to Jack. Going to try a straight RJ11 on the end of the CAT to 100% rule out internal wiring.) SNAP have no test modems to send out currently...so not looking good to getting this resolved anytime soon! I really didn't want to start playing with the stability settings for what seems like a fault as it drastically reduces the Speed, I lost 1/4 of the speed just dropping it down to balanced and still had quite a lot of errors. Anyone have any other thoughts? I wouldn't worry too much about the errors I just checked a business where they have the same attenuation and with similar errors and are on DLM1.

stevehodge

159 posts

Master Geek

Reply # 737773 27-Dec-2012 18:15
fellaintga: I wouldn't worry too much about the errors I just checked a business where they have the same attenuation and with similar errors and are on DLM1. You can't compare error rates between DLM levels - errors increase significantly on the low latency DLM levels. A CRC error rate of 0.66 on DLM-1 seems to be ok. A CRC error rate of 2 on DLM-3 is not good at all.

Jekkyl

41 posts

Geek

Reply # 737989 28-Dec-2012 11:09
So line test is back and it's looking pretty bizarre, so SNAP are going to try and get a test Fritz out to me asap to rule that out before getting chorus involved, as I can't really afford the callout if its a faulty modem and not the line!

1 | ... | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | 50 | 51 | 52 | 53 | 54 | 55 | 56 | 57 | 58 | 59 | ... | 87