Print

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

[Brian Coat]

Hi all,

I am getting SNR downspikes (from 6 dB down to zero then straight up again i.e. a big noise), most often at 4pm ... but not always at 4pm and the 4pm one does not happen every day.

It does not hang around long enough to affect ES/hr, FEC or CRC rates but the DLM does not like it and sometimes dibs my speed down. See attached capture as an example.

My IP profile seems to be 'sticky' it  is very reluctant to 'learn' back up again even after prolonged low noise periods - so the IP profile ratchets down over time, possibly due to this noise effect? About once every 6 months I have to call the ISP and get the IP profile reset to a sensible level.

Anyone who knows what a pain in the botty it is to get BTW to reset an IP profile will understand that I have a vested interest on tracking down this otherwise minor noise issue, just in case it is the cause of my IP profile ratcheting down.

I have checked that there are no timed events like heating coming on at 4pm. No swarms of unsuppressed mopeds riding by at 4pm etc.

No neighbours really, so its probably me.

What are the other usual suspects to check??

Any suggestions gratefuly received.

[kitz]

Attaching QLN, Hlog and bit loading may show if the noise is at a particular frequency.

Problem being is that QLN & Hlog are moment in time graphs taken during the sync up phase which isn't much help if the noise is REIN or SHINE like as yours appear to be

[skyeci]

Sorry to say but I had 3 -4 months of random snr spikes both on ds and us. At one point I had a minus figure for snr.. both would drop to zero, sometimes forcing a resync. Even though openreach investigated it lead to nothing. At first it was implied no fault by isp. I supplied 3 months of charts to isp and openreach which clearly showed the random issues. Worst drop out was 2 hours plus with no service, which acknowledged eventually somewhat reluctantly the final outcome was decided as non fixable. Thankfully it's cleared off at the moment so I hope it does not come back anytime soon.....

[Brian Coat]

Thank you, both.

Kitz: There is nothing in the frequency domain to suggest a problem frequency, neither QLN or Hlog. If it is a very short total drop-out this won't necessarily show up in the frequency domain, I guess.

Skyeci's testimony suggests Openreach technicians will not be able to diagnose it. I agree this is likely because it is intermittent.

My next thoughts are to log my home's mains current behaviour and RF noise to see if there any anomalies coinciding with the dslstats-logged events.

This is a Raspberry Pi project in itself, so it will have to wait until work quietens down a bit.

[kitz]

Kitz: There is nothing in the frequency domain to suggest a problem frequency, neither QLN or Hlog. If it is a very short total drop-out this won't necessarily show up in the frequency domain, I guess.

Yup.  As above both of those graphs are only a snapshot in time and the odds of them catching bursts of SHINE are practically zilch.  You may have more chance of catching REIN, but that would depend upon the type of REIN duration and being able to perform a resync during that time.