cancel
Showing results for 
Search instead for 
Did you mean: 

RF Noise on the Line.

EnglishMohican
Aspiring Pro
Posts: 311
Thanks: 55
Fixes: 1
Registered: ‎08-04-2009

RF Noise on the Line.

I have banged on on several threads about my broadband line being affected by noise - that's RF noise - not crackles on the line.
I have been investigating and can state somethings as facts but I am now stuck where to go next. I am looking for some wise ideas.
So, the facts:-
1 Most of the time recently, my line has between 5 and 7 db of instantaneous noise margin. I guess that the exchange is aiming for 6db and what I see depends on whether the noise is a little bit higher or a little bit lower than usual.
2. Every so often the noise margin drops to nearly 0db. How close it gets seems to depend on the background level  (7 gets to 2,  5 gets to 0) with a bit of vagary  thrown in.
3  When the noise margin is 1.5db, I get about 5% packet loss on a ping test. When the noise margin is 0.5db, I get 60% packet drop and the internet is unusable.
4. On really bad days, the noise margin goes under 0db and the connection drops.
5 The source of the noise seems to start gently and then wind itself up over  maybe 3 or 4 minutes. It lasts a variable time but about 30 minutes seems common and it then turns off quite sharply. There is a slight residual noise for a minute or so. It often starts on nearly the hour or half hour.
6 There is no very firm pattern but it seems to happen maybe twice early in the morning - between 6am and 9am, then again mid afternoon and then again in the evening.
7 I can hear the noise on a transistor radio tuned to just under 600kHz. Its quite a narrow band of frequencies as I can tune out of them and back in. I know the noise is associated with the BB noise as they start and stop at the same time.
8 As I walk around the house with the radio, the volume of noise changes. Its a strange pattern but I suspect it depends on how near to the electrical wiring of the house I am. Its quite intense just near the meter box for instance and close to any sockets or plugs (or maybe it's grounded equipment - the gas boiler makes it howl and it could be the central heating piping rather than the wiring - but my bet is wiring)
9 I can find nothing in my house that works to that sort of on/off cycle - I have turned most things off at one point or another and the noise has continued.
10. I live on a housing estate and there should not be anything high powered for at least 100yds - probably much more.
11 If I lift the phone off the hook the noise margin increases (the noise reduces) at any time - 7db becomes 9db, 0.5db become 2.5db.
Questions:-
1. Could the noise be on the mains electricity supply coming into the house? Is it a high enough impedance to carry that signal around the neighborhood?.
2. If so, would running the router from a car battery (I will check the volts first) help. Would it isolate the mains borne noise from the telephone system?
3  Alternatively, if its a big enough effect, the signal may be on the telephone line as well as the mains electric. Whose problem is it to track the source down and stop it happening?
Can anyone increase my understanding of what is going on or advise a sensible way forward.
15 REPLIES 15
w23
Pro
Posts: 6,347
Thanks: 96
Fixes: 4
Registered: ‎08-01-2008

Re: RF Noise on the Line.

Sounds like REIN: http://www.kitz.co.uk/adsl/rein.htm
The trouble is this can come (via phone or electrical supply wires) from further away than you might first think.
Do you have Routerstats graphs showing the SNR margin changes?
Have you got any extension wiring?  Can you use the test socket behind the mater socket faceplate (if you have one) to isolate extension wiring (if you have any)?
Call me 'w23'
At any given moment in the universe many things happen. Coincidence is a matter of how close these events are in space, time and relationship.
Opinions expressed in forum posts are those of the poster, others may have different views.
Anotherone
Champion
Posts: 19,107
Thanks: 457
Fixes: 21
Registered: ‎31-08-2007

Re: RF Noise on the Line.

Hi EnglishMohican,
As you are discovering tracking down RFI (and curing it) is no easy task. I know from our previous exchange of posts and on this and your recent ones, that you are at the test socket with new filter and modem/router etc. From your point 6, the obvious immediate suspect would be something to do with the central heating/hot water, but as I've said previously, one of the easiest things to do (when the problem is exhibiting itself), is switch everything in the house off, switch off/disable all wireless devices (WAP's included) and any related to central heating thermostats or valves, wireless alarms etc. Although it's not likely to be a fridge or freezer still best to confirm by turning them off for a short period, but to make life initially easy everything off at the same time to start is the best approach. If the problem is still present, you know it's definitely not in the house.
More in a moment.
Edit: When I say everything off, I mean turn off the electricity supply to it, not just off on a timer or in stand-by.
x47c
Grafter
Posts: 881
Thanks: 3
Registered: ‎14-08-2009

Re: RF Noise on the Line.

I have some REIN on occassions so can offer a few comments.
It might either be mains borne or coming in via the phone line or both.
I can certainly hear my problem on 612 KHz on the radio when its "on" - a sort of pulsing noise above the general background hiss.
It is most audible near to the electrical fuses/board etc in the house, however runnning router off the UPS and laptop off the battery and with the power plug to the UPS pulled out does not stop it.
So I'm thinking its partly mains borne (but this part is being filtered out by the router/UPS) and it is being picked up by the phone line en route.
Mine starts at variable times late morning most days incl weekends but may disappear totally for quite long periods at a time.
So its not an auto switching device.  To be honest mine is an irritation rather than a problem - though as an engineer it represents an interesting challenge for me to find.
Most of these are faulty switched mode power supplies and they affect frequencies in the broadband which are multipe harmonics of the switching frequency of the electronics inside the faulty device.
It can been seen in routerstats by looking at the noise trace.  REIN shows up by a very jerky/jumpy/jagged line in the noise margin which suddenly changes from a more flatter line normally.
If you know what you are doing and have a suitable router the latest ver of RS will allow you to monitor the SNR of only certain tones in Telnet access.  If you choose these to be one/more of the tones affected by the REIN you can see even more clearly the start/stop times of the REIN.  You can find which tones are affected by comparing the SNR/Tone charts in RS when the rein is off/on. I actually copied the SNR/Tone output from the RS terminal window to an excel spreadsheet and then plotted them all at different times so that I could more clearly see what was going on.  Faulty S.M power supplies will show up as lumpy peaks/valleys up the SNR/Tone chart at each harmonic.
Next you go round the area with an old style analog transistor radio - the one with a ferrite core aerial in them, and try and find the source (tune to 612 KHz)
you need to use it like a WWII radio direction finder in a Lancaster bomber.
Turn the radio around as you walk.  When you find it the sound , then turn the radio through 360 degress until the noise is at its LEAST.  The direction of the REIN will now be aligned with the LONG direction of the radio. (The ferrite core runs along the back length of the radio and when it is line with a source it picks up the least amount).  Obviously you have no idea which end of the radio's length is the source.  Make a note of the line/direction on a local map copy.
So you now move somewhere else and do the same, and maybe again somewhere else.  You should now be producing a triangulation of the source area with your lines converging.
I have mine down to a group of 3 houses about 0.25 to 0.5 miles away from me.
Now comes the difficult bit - what to do about it - if you know the occupants its all easy - you ask them.  This is when you need your accurate REIN diary showing the precise times it has been occuring over the last X weeks etc.
BT aren't really interested as its not a defect in their kit - and they don't have any powers to order a piece of equipment to be switched off (Cease and desist power)
Serious cases they will investigate (Whole villages BB being wiped out by an old granny's dodgy TV)  but they seem to want to go through all manner of pointless excercises before admitting that the problem is REIN.  So they change D sides, E-side,drops, lifts and shifts you name it first, which if it's REIN is all a waste of time usually.
Then they get the REIN team out - who come along with a RF444B detector - which is usually worse than using a radio as above.
They really need to turn out with a £30K+ spectrum analyser and direction aerial - fat chance.
REIN is VERY time consuming to find.
You could try ofcom - snag is they are only interested in people knowlingly pollution the RF spectrum - pirate radios etc not people doing it inadvertantely.
Yours and mine best hope is the piece of crappy defective electrical gadget whatever and wherever finally packs ups/explodes and it ceases forever.
You would be surprised at how much RF [Censored] can be pumped out by a dodgy power unit on a relatively small consumer product.
All that said, the fact that you see SNR changes when you lift the phone receiver suggests that you may have other problems additional to the REIN. This should not be happening.
What do I think mine REIn source is?
Well I think that it is a faulty router power supply.  It is on continuously and can be detected continously nearby the source by radio.  But it only affects me when something else happens to trigger its widespread RF rather than localised RF.  So I'm guessing that switching on a computer attached to this wretched router causes the RF interference to be propagated through the electrcial power system via the computer.   My house then picks it up via the incomming phone line off the power lines...................
Anotherone
Champion
Posts: 19,107
Thanks: 457
Fixes: 21
Registered: ‎31-08-2007

Re: RF Noise on the Line.

In direct answer to your questions -
1) I think walker23 has covered that
2) That would not be advised, but in any event would only be considered as an absolute last resort if you couldn't maintain a stable connection which you can, albeit under-performing. It would also depend on the modem/router as to whether a 12V DC. supply was suitable.
3) If it's coming up the mains supply (ignore the phone line for the moment) and affecting nearby neighbours in some way, then talking to the Electricity Supply company (not who you pay bills to) - the Distribution company that you would speak to about power cuts/faults eg. Central Networks, Western Power Distribution etc. - whoever is applicable to your area. But anyway that's a few jumps ahead from now.
From your list of points, 7.) is the best tracking method, but I'd also be using Tools such as DMT or RouterStats to monitor the BB noise so you can see the immediate effects. I wouldn't waste time at looking at individual tones at this stage, a more general approach initially is best.
8.)  As far as pipework goes that may depend on how good your earthing arrangements are, and again I'd be suspicious of the boiler and/or any controls especially if any are wireless and want to eliminate it as quickly as possible.
11.) It's suggesting that the RF field may be quite local as well as getting into the phone wiring, but then you said it's a DECT, so turn it all off handsets, base unit, the lot, and use a corded phone to see if the effects are similar.
Oh, and make sure any mobiles are turned off as well to be certain they aren't having an effect.
Edit: x47c has given an excellent overview of most of the aspects of REIN and the problems of tracking it, so I've avoided repeating the whole picture.
EnglishMohican
Aspiring Pro
Posts: 311
Thanks: 55
Fixes: 1
Registered: ‎08-04-2009

Re: RF Noise on the Line.

Thanks all for the replies. It all adds to my ammunition!
I attach a routerstats plot from this morning. My current router cannot display the tones - or nowhere I can find at least (DGN1000)
As stated already, the normal level is about 6db and you can see the effect of the interference where the plot drops to nearly 0db at roughly 7:30. This time it lasts a bit longer than usual but the duration is variable in any case. The little blip up towards the end of the plot is where I answered the phone so this plot illustrates most of the problem.
I do think that the 6db base level is also affected by noise, small spikes occurring almost all the time - though quiet often very, very early in the morning. Those spikes also often show a pattern of increase and decrease that could indicate it might be the same source.
When it is in the bad condition, I get huge packet loss on broadband so it destroys my internet connection for the time it is interfering. Just hope they do not turn it on all the time!
I will do some more checking around the house but I am pretty convinced it is coming from outside. Nothing in the house has timers on it that match this sequence of on's and offs and I doubt the freezer thermostat or similar would keep this regular a schedule - besides, I would hear it. running. Easy enough to switch it off during a bad period however so I will try that. I think the central heating effect is something to do with either the earthing  - the casing will be  a big ground plane or possibly just the amount of wiring around it.
So it looks like I will be out with my little transistor radio noise hunting.
Any further thoughts gladly received.
jelv
Seasoned Hero
Posts: 26,785
Thanks: 971
Fixes: 10
Registered: ‎10-04-2007

Re: RF Noise on the Line.

If your router and PC are running on a UPS surely a quick test to see if it from inside your house is possible when the noise is occurring - turn of the mains at the distribution board and see if the noise margin jumps up - you'd only need the mains off for a couple of minutes to prove the point. Will the UPS keep everything up for that long without initiating a shutdown?
jelv (a.k.a Spoon Whittler)
   Why I have left Plusnet (warning: long post!)   
Broadband: Andrews & Arnold Home::1 (FTTC 80/20)
Line rental: Pulse 8 Home Line Rental (£14.40/month)
Mobile: iD mobile (£4/month)
x47c
Grafter
Posts: 881
Thanks: 3
Registered: ‎14-08-2009

Re: RF Noise on the Line.

Yup, I've certainly done that one:
Radio next to electric distribution board tuned on 612KHz, computer running routerstats, REIN on.
Radio records pulsating noise, RS records reduced jagged SNR trace.
Go out to garage and cut power to house totally, computer/router running on UPS, radio obviously battery powered.
The Result:
Radio pulsating stops as the radio is picking the RF up from the electric boards in that room - so no electric power now and no noise.
Take radio out to main switch/incomming feeder in garage - pulsating noise on radio returns there.
The computer/router continues as before on the UPS and records no change at all to the ADSL signal - still the jagged edge reduced SNR.  
So the telephone line is picking up the RFI from outside en-route to the house.
Switch everying back on - no change again to ADSL signal on routerstats.
Final test: walk down road 200yds to where main power line feeding local area comes out ground and yes, stand next to pole and I can hear the pulsating noise - though there is obviously a huge amount of electrical hiss/buzz in the background.
As I said I only loose 0.6 SNRmargin plus gain a load of CRC's usually so its niggling (and interesting) issue rather than being a 'problem' for me.
Perhaps I get the courage to politetly "enquire" from the suspects sometime - but it would need very careful tact  Stand next to the BT pole feeding these 3 houses (which indicentally has no electric stuff nearby) and the pulsating on the radio is more like a hammering.
Anyway back to the OP:
Their problem is a big issue - from the magnitude of the SNR drop
The quality of their RS trace during/after the interference seems similar and the gradual detioration over 7 minutes of the SNR magnitude is unusual for a sudden switch on of a single source of interference.  It could though represent a gradual startup of multiple sources
More than that I've not got many more ideas to add.
jelv
Seasoned Hero
Posts: 26,785
Thanks: 971
Fixes: 10
Registered: ‎10-04-2007

Re: RF Noise on the Line.

It's not a diagnostic method that had occurred to me before, but I think you've demonstrated it's a good test for internally generated noise.
@EnglishMohican I suppose there's no chance of you getting a loan of a UPS?
jelv (a.k.a Spoon Whittler)
   Why I have left Plusnet (warning: long post!)   
Broadband: Andrews & Arnold Home::1 (FTTC 80/20)
Line rental: Pulse 8 Home Line Rental (£14.40/month)
Mobile: iD mobile (£4/month)
w23
Pro
Posts: 6,347
Thanks: 96
Fixes: 4
Registered: ‎08-01-2008

Re: RF Noise on the Line.

The upward blip when using the phone suggests to me something slightly corroded joint somewhere (current can flow more readily when the more 'robust' telephone signal is operational), if I'm not mistaken a 'fault' of this nature can make the line all the more susceptible to picking up interference?
Call me 'w23'
At any given moment in the universe many things happen. Coincidence is a matter of how close these events are in space, time and relationship.
Opinions expressed in forum posts are those of the poster, others may have different views.
jelv
Seasoned Hero
Posts: 26,785
Thanks: 971
Fixes: 10
Registered: ‎10-04-2007

Re: RF Noise on the Line.

Isn't that usually the extra voltage of the ringing of an incoming call? I'd have thought that would have left the SNR higher for a period after the call ended having 'burnt' a better path across the bad joint.
jelv (a.k.a Spoon Whittler)
   Why I have left Plusnet (warning: long post!)   
Broadband: Andrews & Arnold Home::1 (FTTC 80/20)
Line rental: Pulse 8 Home Line Rental (£14.40/month)
Mobile: iD mobile (£4/month)
Anotherone
Champion
Posts: 19,107
Thanks: 457
Fixes: 21
Registered: ‎31-08-2007

Re: RF Noise on the Line.

Point 11 in the OP suggests that for what ever reason, something is extremely sensitive to RF, not necessarily a corroded joint.
w23
Pro
Posts: 6,347
Thanks: 96
Fixes: 4
Registered: ‎08-01-2008

Re: RF Noise on the Line.

Depending on the nature of a bad joint, some corrosion might pass a signal better when there's a higher current flowing, some, as correctly said improves after a ringing voltage.
Definitely not 'normal' for the SNR margin to improve during a telephone call.
Call me 'w23'
At any given moment in the universe many things happen. Coincidence is a matter of how close these events are in space, time and relationship.
Opinions expressed in forum posts are those of the poster, others may have different views.
Anotherone
Champion
Posts: 19,107
Thanks: 457
Fixes: 21
Registered: ‎31-08-2007

Re: RF Noise on the Line.

Nothing in the way his SNRM is varying could be regarded as "normal". If there was a bad joint causing all this I'd expect to be seeing issues when the phone was called, not just a blip for the duration of the call or when just when picking it up.
EnglishMohican
Aspiring Pro
Posts: 311
Thanks: 55
Fixes: 1
Registered: ‎08-04-2009

Re: RF Noise on the Line.

Just an update with a bit more information - not that it makes things clearer really.
The problem co0ntinues to happen - no surprise there.
My telephone input wire, master socket and router are just beside the main electrical cable and the meter and fuse box for the house - So, I can move the router further away than it is - at least temporarily and will try doing that soon. Might reduce any airborne pickup from mains to router but I am not hugely hopeful.
If I take the radio just outside my house, the noise disappears- its also quiet as I walk past next door, starts to increase at the third house and is quite loud by the time I get to the fourth house. Then it seems to diminish again. I live in a U shaped close with my house at the bottom left of the U and the above applies as I walk up my arm of the U. If I go up the other arm, there is silence all the way.  Obviously, the street has two sides to it and the fourth house up has a partner across the street and I am unable to be sure which has the loudest noise. I only vaguely know either of them so am looking to be a bit more convinced before I go knocking on doors. Also, it almost seems as if there could be standing waves with dead spots and peaks that just add to the confusion. Incidentally, I reckon that the peak signal is at 602khz. Does that tell me anything. What is the significance of 600kHz to adsl.
I agree with the comment above that the slow start up is a bit strange.  In fact on some occasions you can see signs that it is starting up maybe half an hour before the sharp increase in noise, it builds slowly for a while and then the noise increases sharply as in the plot above. It  often then continues to increase slowly until it finally drops sharply. Even that sharp drop is not simple because it drops mostly away, then thinks about it for a minute or so and then shuts off completely. It strikes me as being much more an industrial size effect - a big motor being slowly run up or some such - but they certainly have nothing like that 3 or 4 doors away. The nearest would be a big grain drier half a mile away and its a bit early in the year for that.
Finally, Plusnet have offered to get an RF3 filter fitted for me (well maybe). I guess that simply puts a low pass filter in circuit with a cut off above the top ADSL frequency. Any views on whether that would help?