Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Grapefruit]

Just another strange question from me....

What could cause the dropped pings, but yet my connection does not drop and speed/SNRM don't change significantly either?

[Grapefruit]

Pings seem to be getting silly this morning. Averages are climbing fast due to current ping average of 4-500.

[Grapefruit]

Latest update from PN:

[Grapefruit]

Strangely also at the time PN phoned to advise of their update, the SNRM chose to do an up-spike rather than a dip. The ADSL connection also reset after nearly 3 constant days connected. It seems puzzling that it only resets when PN phone!

Looking at their previous updates, the HR fault due to phone calls seems obvious even at the beginning as the "mean time between errors upstream" has remained in the hundreds.

[Grapefruit]

Going back to electrical interference you mentioned which can affect SNRM....could the answerphone kicking in have an affect upon it? Only reason I ask is I've just received a call, checked the SNRM and it didn't spike or dip. Though thinking back i'm not sure if the answerphone kicked in for PN earlier, they certainly didn't leave a message.

[kitz]

So its been more stable for the past few days?  or youve just had less phone calls?

>>> Going back to electrical interference you mentioned which can affect SNRM

Sorry in a head-spin atm trying to catchup on a several things..  do you mean the 1dB electrical interference between mid-day - mid-night or the large spikes?
If you mean the 1dB..  then no.   If its the spikes.. not sure 

Increased latency can be a result of errors on the line.   If youre getting lots of CRC's then things do start to slow down as data has to be retransmitted.  I dont know if this is the reason for your latency issues as you'd need to look at your linestats when its happening to see if you can see the errors racking up in real time.

The upspike is really weird.   I believe there are times when the ring can temporary clear some HR faults.   For example when I had a HR fault myself.. the SNRm would go on a slow decline.  The phone would ring and I'd see that massive spike...  but afterwards my SNRm would be better.  Its something to do with the current from opening the line is just sufficient to temporary bridge damage from an oxidised joint.  [Someone else may give the proper electrical reason]...  but that spike looks nothing like that.   

[roseway]

I think the upspike may be a graphing error in DSLstats. It was probably a deep but brief downspike. If you (Grapefruit) still have access to that part of the graph you might try changing the clipping or switching off the autoscale option and using the manual scaling controls instead - very probably, some scaling combinations will show the spike going down.

[kitz]

Ah yes!   Thank you eric

I think you are correct.
I hadnt noticed that the spike ended at 20dB which is the clipping setting.  I have seen it do that once before..  but it had slipped my mind.

[Grapefruit]

So its been more stable for the past few days?  or youve just had less phone calls?

>>> Going back to electrical interference you mentioned which can affect SNRM

Sorry in a head-spin atm trying to catchup on a several things..  do you mean the 1dB electrical interference between mid-day - mid-night or the large spikes?
If you mean the 1dB..  then no.   If its the spikes.. not sure 

Thank you once again for your reply Kitz, you don't know how much it is appreciated.

Averaging about 5 calls a day it would seem, sometimes more. Sometimes less. It doesn't seem to reset the connection all the time, but almost always causes a spike.

I was wondering whether it was the answerphone kicking in (rather than a call being answered) that could cause the spike, though am unsure of this. Is it worth trying the line without the phone plugged in at all to see whether it still spikes, or is this pointless?

I've tried changing the clip level up to 100db but it still spikes all the way to the top of the scale. Roseway may well be right about the scale causing it to appear this way. Unfortunately I can't go further than 100db but it's safe to assume this is the case. I also tried to scale negatively but it would only let me go to -2.

I've got an engineer booked for Thursday morning so I shall be sure to demonstrate calling the phone when the JDSU is plugged in. I can also show him the SNRM graphs i've been logging. Is there anything else that might help him identify the fault itself, and where it may lie? or that I should illustrate to him?

Thank you all once again, your help is immensely appreciated.

[kitz]

It doesn't seem to reset the connection all the time, but almost always causes a spike.

Mine was the same, it didnt always knock the dsl out, and the sharpness of the spike could vary.

Is it worth trying the line without the phone plugged in at all to see whether it still spikes, or is this pointless?

You could try it just for the sake of elimination, Im honestly not sure whether it will make any difference.

I've got an engineer booked for Thursday morning so I shall be sure to demonstrate calling the phone when the JDSU is plugged in. I can also show him the SNRM graphs i've been logging. Is there anything else that might help him identify the fault itself, and where it may lie? or that I should illustrate to him?

Show him the images where your attenuation has increased. 
I also showed the engineer my dslstats graphs live so he could see the spikes when the phone rang.   He was able to see the errors rack up when the JDSU was plugged in..  but dont forget to make sure he rings your line whilst doing the test.

As to tracking it down and which joint is perhaps corroded... thats his job and not something we as plebs have the tools to diagnose.  Even with the JDSU and other tools at their disposal its not always easy for them to work out which bit of cable or joint is causing the problem. 

Good luck for tomorrow 


----------------------------

FYI I attach a couple of my graphs when I had a HR fault. 
The first graph is when the fault first appeared - it was the day after I ordered fttc!!   But I attach it because you can see I also got the upwards spikes clipped at 20dB.   The SNRm had actually gone into negative figures, which the router doesnt expect and it was recording it as some weird amount.

The other 2 are just random graphs I picked from no particular day but when on vdsl, other than to show just how unpredictable the spikes could be other than when the phone rang.

I had what the engineer called a 'double whammy'.. one of them was easy to find..  the other wasnt.
This was the easy to find one, where the bottom white wire disintegrated when he poked it with a screwdriver.



 The 2nd one took several visits and when it was finally fixed it took the poor guy the best part of the whole day to sort it.   BlackSheep saw a copy of my notes and said that the guy must of worked his socks of to fix it that day as he had a heck of a lot of re-crimping to do.  By this time though my fault had got so bad that what started off as just broadband eventually affected voice too.