Print

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

[burakkucat]

Having said that, turning off the mains power at the fusebox doesn't disconnect neutral from the internal wiring so it could be conducted emissions causing the initial problem.

I thought that all main switches were double-pole? 

Why would you think that?

The way this is supposed to work in the UK is that you fuse live and RCD earth in the premises. You fuse live at the tails on the meter.

Edit - to clarify, the only current path to neutral/earth is via the fused/switched live side so why would you expect switches that isolate neutral?

I don't want to argue about this but my choice of wording ("I thought that . . .") was really just me being tactful.

To be honest, every main isolator switch that I have examined (and that is a significant number -- but not as many as oldfogy, retired electrician and Black Sheep, ex-electrician) have all been double-pole, without exception. If I could be bothered to do so, I could remove the appropriate covers, photograph both main isolator switches (here in The Cattery) and post the images. Viewers of those images will see double-pole switches.

[sheddyian]

After a good few weeks of relatively trouble-free ADSL, tonight the modem lost synch again, and re-synched around 2k lower than is typical.  Switching on the Medium Wave radio, I can hear a very loud buzzing that's been absent for some time.

The problem has occurred on a Sunday evening once more, although there is one difference - the neighbours 2 doors up, whom I previously suspected, are away on holiday.  No cars on drive, no lights in house. 

Interference once more seems loudest from immediate neighbours, even though I'd assumed I'd ruled them out when I switched off their mains at the fuse box and the interference carried on.

And then, at 23:03 the noise (which I was quietly listening to on the radio) was interrupted with a click and then it faded away over perhaps 20 seconds.

The previously noisy MW frequency is now once more silent.  And a reboot of my modem gives me normal expected synch speeds.

Baffled now.
What would YOU do next?

Ian

ps occurrence of noise gave me chance to re-test a previous experiment, namely running the modem from a 12 volt battery.  When I'd done this test before, there was no interference, and I couldn't detect any difference between battery or mains adaptor operation.

Since this noise is mains and RF borne, I wondered if disconnecting the modem from the mains would achieve anything.

The answer was around 500k gain when switching to battery during the RF interference.

Not sure if that's very significant though.

Am currently still on battery power, wondering how long it will run for

[burakkucat]

How very perplexing! 

It seems as if the only way this problem will be resolved is by you acquiring the services of a flock of Black Sheep, SFIs and PTOs from the ranks of Bettie's elite.

[Black Sheep]

Hmmm, tis indeed a tricky one ??

I'm afraid that, due to the random timings of the REIN, it would probably end up at the door of the PTO's, after an initial task has been raised through the normal fault channels with the ISP. The PTO's have far more advanced equipment than us mere mortals and are more likely to be able to attend site on 'Overtime', again, due to the lateness of the REIN emission. I know this to be the case as I've had call to bring them in on a couple of jobs on my patch, in the past.

I haven't got the time to pour over the whole thread again, but all this would only take effect if the full set of criteria has been met. IE: A PQT pass with an AC Balance of at least -50dB, preferably -60dB. RF3 and SSFP installed. Sometimes, the REIN team will request an E and D side change, but they are just desk-jockeys who are script-reading and I will politely explain that this is not required. Sometimes, I kid you not, they will have the engineer put their van radio on 612Khz and turn the volume up, then they will determine if it's REIN or not from what they hear on the other end of the phone.
Again, as a REIN trained engineer myself, such a ludicrous request is usually met with short thrift.

Obviously, this particular circuit isn't subject to prolonged REIN conditions, so the EU's claims (via their ISP's retrospective monitoring tool) will have to provide evidence of when it is actually occurring. If it's a classic BT DSL circuit, then we have the facility (WHOOSH) to view the circuit ourselves, which is a great help. Trying to get the necessary info from a Level 1 call-centre advisor is like pulling teeth, not their fault as they're probably not trained on what to look for ??.

I have to be honest, the OP is gonna have trouble getting this resolved if it is REIN. The intermittency of it makes it damned near impossible to predict when to attend site. Bear in mind nobody is paying 'us' (and at OT rates) to find the REIN, it really would be hard to justify dipping into the coffers, with a good chance of not getting a result.

Sorry to be all negative, or realistic as I'd call it  , but I think it's a case of sheddyian collating as much evidence as he can per-chance he does pursue it via his ISP.

[sheddyian]

Hmmm, tis indeed a tricky one ??

Sorry to be all negative, or realistic as I'd call it  , but I think it's a case of sheddyian collating as much evidence as he can per-chance he does pursue it via his ISP.

I quite accept that it's going to be tricky to resolve, and your comments and insight are helpful, not negative

As you say, realistic.

I always hate intermittent problems, especially where they seem to defy logic (as in this case).

Since Sunday, the connection has been fine.

Wednesday has also been a popular day for interference, so well see what happens tomorrow.

Curiously, Wednesday is also a day that the immediate neighbours do church-related activity.

Wednesday and Sunday.  The days I'm most likely to get the REIN problem.

Wonder if they're praying too hard and it's interfering with my router and MW radio? 

Ian

[Black Sheep]

They do say that God works......................... nah. too corny by far. 

[burakkucat]

Perhaps, when time permits, TobyPTO could offer some suggestions as to the techniques which can be used when attempting to trace such an intermittent source of interference? 

[sheddyian]

It ocurred to me earlier that when I switched off neighbours mains electricity at the fuse box, I adh the MW radio in their housem and it continued to buzz away loudly.

But what if that's irrelevant?

In MY house, noise around 600 - 700Mhz on my MW radio will co-incide with loss of synch, lower speed resynch and high CRC errors.

Maybe the noise is ALWAYS present in their house, for whatever reason?

Yet their activity seems to influence the interference.

So, when convenient, I need to switch off their mains electric then listen to interference in MY house. Did it have any effect?

Thankfully, they are very amenable to all this, and I've explained that I don't blame them personally. At worse it's some equipment of theirs causing this, but they know nothing of it. Or that it's utterly nothing to do with them.  I've stressed that it's a puzzle I'd like to solve

Ian

[sheddyian]

Update :

After a trouble free week (including Wednesday evening, which has been a popular time for problems in the past), on Sunday evening around 7pm I lost synch and reconnected around 2K lower than usual.  Switching on the MW radio showed the loud hum to be back.

Later that evening (around 11:30pm) I switched the radio on again and the hum was gone, so I rebooted the router and got my lost 2K back.  I noticed that in the intervening 3 hours I'd clocked up 6,000 error seconds!  (10 - 15 an hour is more usual for me).

When the modem lost synch the immediate neighbours were out, the the ones further up the road were in - though they were on holiday last week when this last happened 

Why Sunday evenings and occasionally Wednesday evenings? 

Trying to approach this logically : What electrical equipment might someone have on a timer or somesuch, that only runs once a week, on Sunday evening, for 3 hours or so?

Ian

[JGO]

"Trying to approach this logically : What electrical equipment might someone have on a timer or somesuch, that only runs once a week, on Sunday evening, for 3 hours or so?"

Watering for greenhouse plants  ?

Just a memory jog,  most portable radios have a ferrite rod aerial which is directional. It isn't as good as a proper DF frame aerial and off hand I don't know  the angle of the nulls in relation to the aerial ( always use nulls, they are sharper, maxima are flat)  but afer a dummy run on the local radio station, so you know the location,  two or three checks should may localise the RFI source. 

[sheddyian]

Just a memory jog,  most portable radios have a ferrite rod aerial which is directional. It isn't as good as a proper DF frame aerial and off hand I don't know  the angle of the nulls in relation to the aerial ( always use nulls, they are sharper, maxima are flat)  but afer a dummy run on the local radio station, so you know the location,  two or three checks should may localise the RFI source.

I was looking this up the other day, to check which way the aerial is most receptive 

.. which is basically along the length of the ferrite coil I think, with the nulls at each end.

           
             max
null ===ferrite=== null
             max

(I found a site with a better picture than that, can't find it again now...)

Now, with the radio aligned along the max of the ferrite, the interference is strong enough to cause the "tuned" led to light.

And certainly walking along the street I can determine where the interference fades out, and get a rough idea of direction (the two neighbours houses, but I can't decide which)

But you say to use the nulls to locate the signal?  How would I go about doing that? 

Thanks

Ian

[JGO]

"I was looking this up the other day, to check which way the aerial is most receptive 

.. which is basically along the length of the ferrite coil I think, with the nulls at each end."

That is what I thought but it is read across from a loop, so wasn't sure.

"But you say to use the nulls to locate the signal?  How would I go about doing that?"

The idea is that the rate of change of signal with angle is greatest at a null and zero at a peak so the nulls define a direction better. The technique is to find the directions where the signal drops below receiver noise and returns the other side and average them to approximate the true null.  Repeat  from two or three locations to sort out the 180 degree ambiguity of a single bearing  and then move in that direction to narrow the zero signal angular range.
 If you get too close the radiator isn't a point source so you can't get a clean bearing, but that puts you quite close. ( I seem to remeber reading that the French Resistance used to use a telephone wire miles long as an aerial to fool the Gestapo, and this could be replicated with overhead power lines and a powerful interference source - but that isn't a faulty appliance it is a*!&? spark transmitter, banned since 1938 ! ) 

P.S.  perhaps I should add thet the receiver AGC tries to flatten out the peaks so nulls are a better bet for this reason too. 

[shirts]

I am so pleased to find out about the problems you are having. No, it is not because i'm a nasty piece of work, it's just that I have been having EXACTLY the same problem and I am running out of hair to pull out !
The sounds of the buzzing that you posted were spot on. EXACTLY the same, although i'm getting it at different times. In the colder weather, I had the interference between 1800hrs and 2100hrs, during which time my errors rocketed from nil to multi-millions. Just when I think I have found the culprit and get a few days of nil errors, it then comes back when least expected with a vengeance. It now seems to occur Saturday and Sunday evenings.
Can anybody recommend a mains filter which would give me a source of clean power to the hub just in case it is the mains which is the source of this interference ? I have already tried cat6 cables and am running out of options.........  

[sheddyian]

Thanks for suggestions re aerials and locating the source of the interference.  Wednesday evening went without incident, and everything has been fine since last Sunday - so I await this Sunday with interest, especially as the suspect-neighbours are away again for a bank holiday caravan break

I did try a wander up and down the street the other evening when I was NOT having interference issues, and the radio was largely silent - though there was a slight buzz around the immediate neighbours house, which was interesting.

Ian

[sheddyian]

The sounds of the buzzing that you posted were spot on. EXACTLY the same, although i'm getting it at different times. In the colder weather, I had the interference between 1800hrs and 2100hrs, during which time my errors rocketed from nil to multi-millions. Just when I think I have found the culprit and get a few days of nil errors, it then comes back when least expected with a vengeance. It now seems to occur Saturday and Sunday evenings.
Can anybody recommend a mains filter which would give me a source of clean power to the hub just in case it is the mains which is the source of this interference ? I have already tried cat6 cables and am running out of options.........

That does sound quite similar to my problem - though Saturdays is usually trouble-free for me 

My feeling on this is that a mains filter won't help.

If you can hear the noise on an AM radio, it's being radiated out from the source(s) and so your router/phone cabling/lan cabling/computer cabling etc etc will all be susceptible to picking it up and it getting into the router.

Even if you were able to put all these things in a Faraday cage (to eliminate radio frequency interference) and run from a 100% clean power feed, chances are the phone line under the road or on the poles will be picking it up.

(My own experiment with disconnecting my router/modem from the mains and running it from a 12 volt battery proved this, I still suffered a loss in speed until the interference ceased)

It's a case of tracking the noise down and removing it at source.  Which I initially found easy (I discovered a faulty computer PSU making a dreadful noise on AM) but then later discovered there was another source outside of my house that I've yet to pin down.

Ian