Print

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

[ubertank]

Hi guys,

I'm typing this out with the minute or so of internet I will get before I get kicked out again.

For the last couple of weeks, I have been losing internet and after having an Openwork Engineer come to my premises to change all wiring, it looks to be a REIN issue.

I have had Routerstats now running for the last few days.  The broadband was getting shut down between 4 and quarter past on Wednesday, Thursday & Friday, intimating someone is coming home around that time and using an appliance of sorts which is causing this interference.

This is what my routerstats log looks like:



BT have said they have put my case forward to the REIN helpdesk, but have given no timeframe on when this is likely to be looked at, or indeed if it can be solved.

WHat I would like to know is, the way the graph is plotting the REIN bursts suggests that it is happening in pulses.  Is this the case, or just the way routerstats is plotting it?  I have tuned an AM radio to 612 and have heard the REIN through the radio exactly when the noise margin drops to 0db. 

Would the noise margin usually go down to 0 and stay there if say, a TV had been turned on that was causing the issue?  If so, what sort of appliance would sent out REIN in bursts like this?

I would really appreciate any guidance anyone can give me on this as the issue is really causing me no end of grief.  I use my line for business and it is costing me money every single day.

Thanks in advance.

[Black Sheep]

Hi

Just a quick reply as got other matters to attend to soon.

If a REIN case has been built, chances are somebody from Openreach will attend site to attempt to locate where it is being transmitted from. Things to note are ........ 1) REIN tasks will only usually be issued when workstacks are low. 2) REIN faulting is a gratis service Openreach offer, nobody pays us for our time doing this task. 3) The task issued to the REIN trained engineer will only have one hours worth of work attatched to it + Travel Time. In todays statistic-controlled environment, whereby an individual could end up on an 'Under performance plan', there's not many that will go over the allocated time. 4) If the REIN (when verified) isn't located within the hour, another task WILL be built for future attendance. 5) You will not be aware when the OR engineer is going to attend site as REIN tasks are non-appointed. 6) The REIN engineer needs to be aware of the times when REIN is service affecting, otherwise the 'Case' will be totally closed down.

It appears you have done good groundwork on this already. Further analysis you could personally do, is to ask if any neighbours are also being affected, as this could determine if the REIN is local to your premises.

To answer your Q about 'What devices could cause this REIN' ................ well, anything electrical thats going/gone faulty TBH. Central heating stats, Satellite set-top boxes, baby monitors, street lights, white goods, TV Aerial boosters ........ the list is pretty comprehensive.

If your radio is portable, leave the setting on MW 621Khz, and have a walk around outside the premises to see if it's prevalent elsewhere, if you don't mind looking like a weirdo that is.

 

[ubertank]

Thanks a lot for the reply.  Any and all information I'm given at this moment in time is appreciated.

Regarding the apparent "pulsing" of the REIN on the routerstats graph, does this seem to suggest that an item is being turned on and off very quickly and emitting this frequency, or could an appliance be left on and send out the frequency in this apparent "pulse"?

[roseway]

That graph is fairly typical of REIN. Routerstats takes a sample of the noise margin every 15 seconds (or whatever time you've configured) so it isn't a continuous curve, but just a series of snapshots. So you can't infer from the graph that something is being switched on and off rapidly, it's just a reflection of how the timing of the samples corresponds with the ups and downs of the interference.

If you walk around outside with the radio you can rotate it for maximum signal, and this will indicate the direction of the interference source - it will be on a line at right angles to the long axis of the radio. If you do this from a few different positions, you may be able to determine where the source is by working out where the lines cross.

[asbokid]

If your radio is portable, leave the setting on MW 621Khz, and have a walk around outside the premises to see if it's prevalent elsewhere, if you don't mind looking like a weirdo that is.

Great advice as ever.

Out of curiosity, how did the frequency of 621kHz come about?

cheers, a

[burakkucat]

Out of curiosity, how did the frequency of 621kHz come about?

By a typo!    B*Sheep intended to type 612 kHz.

Why 612 kHz? Because it's the second harmonic of 306 kHz and falls into the frequency band of a domestic MW radio.
So what is the relevance of 306 kHz? I forget. (Too much beer and wine has been consumed to celebrate Liverpool FC's win of the Carling Cup . . .          ) It may be explained on the REIN page here (on Kitz' site) or it may be in one of Ezzer's forum posts.

Perhaps Eric, Walter or B*Sheep could explain, please?

[asbokid]

There is a recent patent filed by Alcatel for a "method for remotely detecting electromagnetic interference on a data transmission line of a communication network, in particular for electromagnetic interference caused by a power supply unit located at a client site.. 

"The method is based on REIN: Repetitive Electrical Impulse Noise Detection"  [1]

From a brief browse of the patent, it is based on analysing the Quiet Line Noise (QLN) data obtained from the CPE.. 

The noise peaks from PSU equipment tend to be at the switching frequency of the regulator in the PSU, and at the harmonics of that frequency.

Analysing the distance between those harmonics (their periodicity) is used to discover the fundamental frequency of the switching regulator.

The discovered switching frequency is then looked up "in a predetermined information list relating to predetermined types of power supply units and their associated switching frequencies, thereby identifying said type of said power supply unit."

In the Alcatel patent, the QLN data is automatically sent for analysis to the Communication Provider, presumably via the TR-069 remote management framework that is built into the firmware of most modern modems.

Of course, the REIN detection method is limited insofar as the switched-mode PSU is only one source of REIN.

Can you obtain QLN data from your modem?

[1] http://www.wipo.int/patentscope/search/docservicepdf_pct/id00000015527702?download

[roseway]

Why 612 kHz? Because it's the second harmonic of 306 kHz and falls into the frequency band of a domestic MW radio.
So what is the relevance of 306 kHz? I forget. (Too much beer and wine has been consumed to celebrate Liverpool FC's win of the Carling Cup . . .          ) It may be explained on the REIN page here (on Kitz' site) or it may be in one of Ezzer's forum posts.

Perhaps Eric, Walter or B*Sheep could explain, please?

My memory of the details is vague, but I think that it just happens that 612 kHz was the frequency used by a particular piece of test equipment. It's more or less in the middle of the ADSL1 band, and is well clear of any radio transmitters. Because of the nature of REIN, it actually interferes over a fairly wide range of frequencies, so the frequency used by the radio / test equipment isn't critical.

[Black Sheep]

Why 612 kHz? Because it's the second harmonic of 306 kHz and falls into the frequency band of a domestic MW radio.
So what is the relevance of 306 kHz? I forget. (Too much beer and wine has been consumed to celebrate Liverpool FC's win of the Carling Cup . . .          ) It may be explained on the REIN page here (on Kitz' site) or it may be in one of Ezzer's forum posts.

Perhaps Eric, Walter or B*Sheep could explain, please?

My memory of the details is vague, but I think that it just happens that 612 kHz was the frequency used by a particular piece of test equipment. It's more or less in the middle of the ADSL1 band, and is well clear of any radio transmitters. Because of the nature of REIN, it actually interferes over a fairly wide range of frequencies, so the frequency used by the radio / test equipment isn't critical.

Eric is 100% bob-on.

I posed this particular question to one of our PTO experts, and the answer he gave was that 612Khz MW has nothing else near to it that could 'bleed' onto that frequency. He gave me another frequency for the LW range, but I'm buggered if I can remember it

[Black Sheep]

Out of curiosity, how did the frequency of 621kHz come about?

By a typo!    B*Sheep intended to type 612 kHz.

Why 612 kHz? Because it's the second harmonic of 306 kHz and falls into the frequency band of a domestic MW radio.
So what is the relevance of 306 kHz? I forget. (Too much beer and wine has been consumed to celebrate Liverpool FC's win of the Carling Cup . . .          ) It may be explained on the REIN page here (on Kitz' site) or it may be in one of Ezzer's forum posts.

Perhaps Eric, Walter or B*Sheep could explain, please?

Ha ha ha ....... well spotted Cat !! Yeah, it is indeed a typo and 612Khz is correct. PS .... I've posted a comment about 'The Reds' win in the 'Jokes' section pal. 

[jeffbb]

Hi
quote so it isn't a continuous curve, but just a series of snapshots

True but every snapshot is a plot on the graph so if there is a sequence of plots where there is no change on the graph then  this demonstrates that the SNR has not changed for those plots. A sudden drop does indicate an increase in noise so reducing SNR margin for that sample  . The important thing to remember is if the time between samples is too long then its possible to have more undiscovered changes (drops) in between the sample times.
Regards Jeff

[ubertank]

I will adjust the Routerstats time to less than 15 seconds later so I can have a look.

I have spoken to a couple of neighbours who don't see to be having the same problem, so the REIN seems to be just affecting my line.  I have literally only spoke to those either side of me, so this doesn't really give a lot of information - I will enquire further afield later.

If it's just my lines that are being affected by the REIN, does this not mean there is an issue that can be solved by changing the lines?

There is no cable in my area for me to change to a Virgin service, not is there good 3G coverage for a dongle (which are pretty bad anyway), so if my adsl doesn't get fixed it literally means having to move house or change business. 

[Black Sheep]

I will adjust the Routerstats time to less than 15 seconds later so I can have a look.

I have spoken to a couple of neighbours who don't see to be having the same problem, so the REIN seems to be just affecting my line.  I have literally only spoke to those either side of me, so this doesn't really give a lot of information - I will enquire further afield later.

If it's just my lines that are being affected by the REIN, does this not mean there is an issue that can be solved by changing the lines?

There is no cable in my area for me to change to a Virgin service, not is there good 3G coverage for a dongle (which are pretty bad anyway), so if my adsl doesn't get fixed it literally means having to move house or change business. 

If REIN is only affecting your line, there are 2 possible reasons why.

1) The REIN source is local to your premises, and the chances are you will be charged by Openreach for their time spent localising this.

2) Your pair of wires from the Exchange to your premises (MPF), aren't of a high-enough quality to support higher frequency carriers. This usually presents itself to us in the way of a really poor AC Balance. There are different views on what is best practice. Our REIN Helpdesk (and SIN 349) insist it should be -50dB or lower. Dont forget, we are in 'minus' territory here, so 'lower' would mean a higher number. I have documentation stating -65 is ideal, and other stuff saying -70 is ideal. My experience tells me that -50 is probably on the boundaries for receiving fault-free DSL.

If your line has a higher reading (EG- -33dB), you will be susceptible to all kinds of noise levels, thus de-stabilising your line.

<Disclaimer> With you commenting that it happens only at certain times, my guess would be that it is still REIN, rather than a poor quality MPF. Remote faulting of a circuit is pretty hard and can only be done so on the information given.     

[waltergmw]

@ UberTank,

IF all your efforts come to nought you might like to investigate WiBE devices which can perform significantly better than a mobile etc.

http://www.deltenna.com/products/wibe.aspx

These are some test results I did over the hill closeby.

http://www.scribd.com/doc/76790433/All-Albury-WiBE-Items

I believe they are about to launch a HSPA+ device possibly capable of 25 Mbps download and quite fast upload too.

The units are designed to sit on a windowsill but they are also experimenting with an external unit perhaps for chimney mounting.

Kind regards,
Walter

[ubertank]

THanks a lot for your replies.  Some really helpful information.

I walked up and down about 3 or 4 houses either side last night.  It appears nobody else is having the issue.  I would be discouraged and think the source of REIN could be anywhere along the line between me and the exchange, however because I can clearly hear it at 612kh I'm more encouraged it's something pretty local to me.

I spoke to my elderly neighbour yesterday who the timing of the REIN seems to coincide with relatively well and she put on her Alba TV around the time I had REIN.  SHe turned it off for me so I could check routerstats and it was clear for about 30 seconds, then I got hit with a blast.  I'm not sure if REIN may continue for a short while while the appliance is on standby, if is does, it could have been the TV. However, last night I had bad REIN until  1AM, them it was totally clear (she may have gone to bed), but then I had s single spike on routerstats at 5.34, another single spike at 6.14, and another spike at 6.45.  I ten had continuous RAIN from 7.09 to 7.38.

The pattern is strange and I can't imagine what sort of appliance would emit signals like this.  I did take a look at her boiler which is admittedly pretty old looking (I didn't want to turn it off in case I couldn't get it back on) and on a timer, but these REIN times don't make much sense if it's a thermostat clicking on and off.  It wouldn't go on literallly once or twice throughout the night, then pretty much stay on between 7.09 to 7.38..at least I can't imagine it would.

Anyway, I have a REIN engineer coming out today between 1 and 5.  I have stated the REIN is off until 4 normally but they said he'd be out between this time and that would be that.

Seemed pretty pointless to me.  Luckily the lady next door is in by 12 today so if it is her, hopefully I'll be getting something.

Fingers crossed.

Thanks again guys.  I'll let you know how it goes.