Print

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

[splbound]

Hi all. After trawling through this very informative forum and learning as much as I can I'm stuck.
I'm getting this consistent interference on my line. Not sure if its REIN interference or just Crosstalk.

Master socket was fitted pretty recently and the connections look good.
I have no extensions connected to it. I still have to check the telephone line run from where it enters through the roof to the Master Socket.

Interference is lower in the tone bands, higher tones look ok.

The QLN graph peaks look harmonic and occur every 100Khz across the affected range.

Any ideas of possible causes will be welcome.

I'm going to double check turning off each electrical appliance again just to make sure I didn't miss anything.

[burakkucat]

Welcome to the Kitz forum.

Looking at you QLN graph, we see that your average noise floor is somewhat higher that the typical -140 dBm/Hz and, overall, it is a somewhat "messy" plot.

I assume you have taken into account your geographical location? For example, are you near to a radar installation or a broadcast transmitting complex?

[JGO]

As it seems the interference covers (some) of the Medium Wave band, suggest a hunt with a transistor radio (NOT at 612 kHz !)  Some sort of switched PSU seems more probable in a domestic environment, although agree with burakkucat it could be a radar with a 10 u sec pulse width. If you can DF it with a radio that would help !   

[splbound]

Thanks for the help. I'm located in Beckenham BR3 so Crystal Palace is relatively close to me.
There is MW transmission on 558, 720 and 1035Khz so suspect the dips in the respective areas of the graphs caused by them.

The DP is a pole not far from the property so I'm looking at it thinking that's a great way to pick up interference.
Not only that but the pole serves all the properties in my street (short cul de sac) of around 35 properties. Not ideal for crosstalk or induced noise to the lines from any of the other properties. The rest of the run to the CAB is underground which runs by a hospital and a petrol station with the cab being about 500M away.

I'll buy a radio, tune in and hopefully it will lead me somewhere.

[JGO]

Aha yes overhead line -  a good way to pick up both crosstalk and interference !


I take it you have a Mk2 filtered faceplate ? so shouldn't need to add an external RF3 as already included .
If so the path could be via power wiring,  so probe that with the radio.   I wonder - if both  modem and router are fed from a 50 Hz splitter box, can it's input be filtered with one device ( Maplin ?).


You say "consistent interference";  is this 24 hrs/day ? 

[JGO]

Further thoughts.

I think you are far enough from Crystal Palace not to get interference from it at MW frequencies; if not a lot of other people would be worse off !   I'm inclined to suspect something domestic, switching PSU, street lights or Plasma TV - the time of day when it happens may give you a clue.

Regarding a portable radio, for DF there is are sharp nulls along the line of the ferrite rod aerial, usually across the width of the case. The radio's Automatic Volume control (AVC) will try and smooth it out so you will find the null is quite sharp - practice on a BC station.
Also getting too close looses accuracy since an aerial say 10m long covers a angle of about 6 degrees at 100m but only 0.6 at 1000m - more use in rural areas.

A radio can be used to probe for interference on a wire by just holding it near the suspect, judging the strength by how far away to loose it.

Don't forget you can't eliminate interference except at source; I'm about as far as you from Droitwitch and there is a slight dip at 198 kHz. Getting rid of it would cause a worse overall loss. 
 

[splbound]

Thanks JGO. Yes, the faceplate is already MK2 so no RF3 needed.

The interference is consistent 24-7 so it pretty much rules out any electrical device I have that doesn't run all the time.

I run a server,router, switch and modem from a UPS (APC smart-UPS 1500VA) so should I be right to think that the power to them is being appropriately filtered.
There may be a the possibility that the UPS itself is inducing noise into the power lines or phone lines somehow. I'll try eliminating this as a possible source tonight.

The server, switch or computers near the modem are definitely not the cause, totally powering them down and pulling the plugs on them don't do a thing to change the graphs.

Thanks for the radio tips. I'll practice first.

Failing this I can probably just live with it, wait for FTTP or move to an FTTP area (My 40ish Mbps speeds aren't as bad as some of the other not so lucky in this forum). My other choice is getting my ISP to send Openreach out and footing the bill if nothing is found.

[JGO]

I run a server,router, switch and modem from a UPS (APC smart-UPS 1500VA) so should I be right to think that the power to them is being appropriately filtered.
There may be a the possibility that the UPS itself is inducing noise into the power lines or phone lines somehow. I'll try eliminating this as a possible source tonight.

Should is the operative word; recently someone reported a 50Hz splitter board which was arcing. When fixed his noise level dropped 3 dB !
Suppression components can go faulty and it isn't unknown for electricians to just remove the dud component, no replacement.
Take nothing on trust !  Good Luck.   

[NewtronStar]

Thanks JGO. Yes, the faceplate is already MK2 so no RF3 needed.

I was under the illusion when the BT80 RF3 is installed along with the MK2 you are supposed to get less RFI than with just the MK2 FP, or has this all gone out the window, I do feel there is a lot of inconsistency when it comes to RF3 and MK2 on this forum 

[burakkucat]

My understanding of the situation is that the Mk 2 SSFP was designed to incorporate common-mode rejection circuitry which should have made a BT80-RF3 unnecessary but the final product, manufactured, distributed and deployed in the field was found to be lacking -- not performing to its design specification. Accordingly an Openreach directive was issued from "on high" to all technicians to fit a BT80-RF3 to any circuit which was troubled with RFI. The Grimbledon Down¹ wizards were tasked to "look again" and, thus, I expect that a Mk 3 SSFP will eventually be seen in the wild.

My view is that if the insertion loss can be tolerated, then fit a BT80-RF3. (The belt and braces technique.)

[1] Martlesham Heath / Adastral Park

[NewtronStar]

There is a Downside to the RF3 if you have a 40" Led TV near the BT80 for example a meter away the coils in the unit seem to pickup the noise from the TV and the effect is a loss of 0.5 dB SNRM.

Once I turn off 40" LED TV the SNRM will rise, and also the weird part of this my PC LED Moniter also the effect is a loss of 0.2dB SNRM and the other TV which is LCD also effects SNRM with loss of 0.3dB yet they are 12 meters away from BT80 if they are all on at once i'll lose 1dB SNRM which is bad as my SNRM sits at 6.5 during the afternoon with all screens off.

I am Very confused why this is happening but once the radio scanner goes near that 40" TV the interference is very loud 

[splbound]

Update. I have tested with the UPS off and just the modem connected to the router to my PC. Still the same craptastic connection so I don't think its that. Haven't had the chance in acquiring a radio so have not had the chance to test for REIN. I also put the supplied ECI modem back online so they cannot blame the unlocked HG612 for the issues.

After a few weeks of more suffering from sync loss, the DLM smacked my sync down to 30Mbps and pings up to 40ms. This was it for me so I finally called up Zen to get the engineer to come down.

Yesterday I had a very helpful OR engineer come in to look at my problems.

He connected up a JDSU from the master socket and immediately saw CRC errors coming once it had synced.

He tested again with the JDSU but this time from the junction box there the line from the pole hits my the wall of premises.
The junction box connects the cable from the pole to my existing internal drop cable. He noted that the existing drop cable wiring looked very old. I have previously tried tracing this drop cable run from the loft myself but it the area was inaccessible to me.

The test was done by connecting a long twisted pair cable from the junction box, which he trailed into the premises via the front door to near my master socket where he connected the JDSU. This time there was an increase in sync from the 30Mbps to 56Mbps with no CRC errors in the 10-15 minutes tested.

With this info, a new drop line was installed and hooked up to the master socket. The new drop cable ran from the junction box, along the wall outside of my premises. Then through a hole made next to my front door and along the floor skirting. Finally punched down to the master socket. Very nice and clean internal and external routing by the OR engineer so thumbs up to him!

The setup was tested with sync speeds around 56Mbps (Much better than my original 42Mbps and closer to the BT wholesale FTTC estimate). No CRC errors occurred in the time we tested which was about 10-15 minutes. The OR engineer and I agreed that this may have solved the issue so DLM reset and he was off to his next job.

Finally fixed!? .. or so I thought. Everything was fine until about 4 hours later. Sync loss again. Was fine again for a bit until between 8pm - 2am where there was more sync loss.

Zen this morning (excellent customer service, I was going to call them but they followed up on my situation) called me to follow up on whether the issue was resolved. They noted more disconnections since the new drop line was installed which correlated to my findings. We agreed to further monitor the situation over the weekend and if it does not settle down another engineer visit will be booked. Most likely I will have to get another engineer in.

I guess the next places to check will be between the street cabinet and the junction box before my drop cable. Hopefully it can be traced as the run is about 400m via a pole and then underground.

I also wonder if the actual master socket is faulty as this is the only thing that has not been changed on my side. I guess if it was the master socket the JDSU tests would have picked it up?

Only other thing would be REIN interference somewhere out there. I think I can pretty much rule out anything coming from my premises. I have been switching appliances on and off like a madman trying to correlate things without success.

For now, connection is a little faster but I know DLM will kick it back due to the sync losses. I can stand a little bit of slow speed but very frequent random disconnections are not acceptable since I remote a lot to work servers.

I'll keep any new findings posted.

[splbound]

Had another engineer come in last friday. I was hoping it would be the same one from last time but nope.

Tested again with a JDSU from the master socket. No fault found there so off he went to the PCP to check it from that end and didn't return.
I did saw a re-sync a few minutes later after he left which must have been a DLM reset as my speeds went back up.

Disconnections have actually decreased down to around 3-4 per day. Whatever was done at the PCP helped, but what was actually done there was not reported by the engineer to my ISP or me.

I was hoping the engineer would check the connections at the pole. Looking up there it seems like it hasn't been touched for years and all the lines on the street connect to it.

Anyway I am monitoring it for longer to see it if it all stabilises.

[NewtronStar]

I was hoping the engineer would check the connections at the pole. Looking up there it seems like it hasn't been touched for years and all the lines on the street connect to it.

Anyway I am monitoring it for longer to see it if it all stabilises.

It used to a regular sight in the 80's to see an engineer climbing up the distribution pole, these day you would be lucky to see a cherry picker near the pole unless there had been gales which caused the drop wire to come away from the pole.

I am guessing the diagnostic equipment could show a possible fault at the distribution pole if there was one 

[perry081064]

on my install day a year and a half back , i had a cherrypicker because a new line was needed to the pole and the pole was listed on their books as not to be climbed ( obviously it was too old ).

openreach had to ring to see if one could come over , otherwise the job would need to be rebooked.
luckily he rang straight away, and was at my house within 15 mins.