|
|
Another question about signal/noise margin!
My router (ST585 v6) syncs happily for days on end to maintain an IP Profile
of 5.5 Mbps. The diurnal variation of the noise margin is ~ 5.5 to 13
dB. My router is located at and directly connected to the master socket
filter faceplate but occasionally there is a brief noise spike that according to
RouterStats plunges negative off the scale. This spike is quite random and
occurs at no particular time of day or night and upon daytime resync the margin is ~12dB.
Is there any way of suppressing this brief spike to smooth out the noise margin and maintain sync, in a similar way to a power surge limiter?
Line Stats for reference:
Bandwidth (Up/Down) [kbps/kbps]:448 / 6,432
Output Power (Up/Down) [dBm]:11.5 / 19.5
Line Attenuation (Up/Down) [dB]:25.0 / 41.0
SN Margin (Up/Down) [dB]:22.0 / 11.5
Andrew
|
|
|
Yes, slow the broaband down to the point where you have enough spare margin that it does not disconnect.
A spike once or twice a day is one of those you have to live with issues on ADSL.
|
|
The author of the above post is a thinkbroadband staff member. It may not constitute an official statement on behalf of thinkbroadband.
|
|
|
Thanks Andrew, not the answer I was hoping for
A spike a day keeps the margin at bay
Andrew
|
|
Register (or login) on our website and you will not see this ad.
|
|
|
This is what I'm up against!
Now, if someone could just invent something.........
Andrew
|
|
|
Do these "spikes" occur during daylight hours, or just between dusk & dawn?
An all too familiar picture and one I used to regularly see until I switched from a BT wholesale ADSL1 (ADSLmax) product to BE's ADSL2+ LLU product.
The culprit may be traceable (if you have a suitable router).
The 2-wire 2700HG or 2700HGV (often branded as a BT business hub) has an "impulse noise" figure on its web interface.
Normally this figure sits at zero, but occasionally starts to rise and suddenly goes through the roof (at which point a resync occurs).
Using a 2-wire router, I traced one of my noise sources to a cheap ebuyer DVD player, which would give off massive amounts of noise every time it accessed a disc.
In answer to your question "can it be filtered".....
No, unfortunately not.
If the noise is at ADSL frequencies of anything up to 1.1MHz (or 1100kHz in AM radio speak) for ADSL1 or 2.2MHz for ADSL2+; if you filter out the noise, you filter out the ADSL signal too.
Unfortunatly ADSL works up into the same frequencies as AM (MW) radio, so anything which makes a "pop" on AM radio, can affect ADSL broadband (lightning strikes, tube lighting, TVs, microwaves, street lights, and several thousand other things - sometimes it appears just looking at an ADSL router will force it to drop sync).
Ade
ADSL2+ with BE
DL Sync around 4.8Mbps
UL Sync 1088kbps
DG834GT with DGTeam firmware
|
|
|
That is not a spike! I would call that a "burst" and for RSL to pick it up it must have a duration of several seconds. You need to look close to home and see if there is anything causing it. Possibly a freezer which may only come on twice a day
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
M H C
taurus excreta cerebrum vincit
|
|
|
That is not a spike! I would call that a "burst" and for RSL to pick it up it must have a duration of several seconds. You need to look close to home and see if there is anything causing it. Possibly a freezer which may only come on twice a day
I had similar in the summer, and my ISP called out BTOR who found three dodgy connections *outside* my flat, but *inside* the block of flats, not touched since built in 1971....
--
James - be* pro, on THFB exchange with a Draytek 2820Vn and a BeBox (585v7) BQM
|
|
|
That is not a spike! I would call that a "burst" and for RSL to pick it up it must have a duration of several seconds.
Something arching then? It can happen at any time of the day so perhaps so as you say, maybe my freezer thermostat.
My phone line also passes under but at right angles to a 11kV power line but it's only me on the end of the power line!
Andrew
|
|
|
In answer to your question "can it be filtered".....
No, unfortunately not.
OK, thanks
Andrew
|
|
|
Where's the transformer?
My broadband basic info/help site - www.robertos.me.uk
My domains,website and mail hosting - Tsohost. Internet connection - O2 Standard.
|
|
|
Possibly or a motor having trouble starting.
I suggest freezer as a fridge would be more frequent and if you have a dual compressor (one each) fridge freezer it is possible. Plenty of others including a neighbour on a wall backing on.
At 90 degrees no issue with normal noise but a burst could cause it. Is the transformer a tap switching version? So as the incoming voltage drops/rises it will step up/down?
See if you can get a few more occurrences and measure the mains before and after.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
M H C
taurus excreta cerebrum vincit
|
|
|
Is the transformer a tap switching version? 
I don't know if you are following on from my question about the transformer, but I was referring to whatever is providing the 230V supply from the 11kV feed. I doubt if the OP knows anything about it.
Goodness knows what could be going on in the cabinet holding that.
My broadband basic info/help site - www.robertos.me.uk
My domains,website and mail hosting - Tsohost. Internet connection - O2 Standard.
|
|
|
Where's the transformer?
About 75 feet from the phone line, at the top of a pole.
Andrew
Edited by Andrew_W (Sat 06-Nov-10 14:19:13)
|
|
|
Possibly or a motor having trouble starting.
I suggest freezer as a fridge would be more frequent and if you have a dual compressor (one each) fridge freezer it is possible. Plenty of others including a neighbour on a wall backing on.
At 90 degrees no issue with normal noise but a burst could cause it. Is the transformer a tap switching version? So as the incoming voltage drops/rises it will step up/down?
See if you can get a few more occurrences and measure the mains before and after.
The transformer is so old I think it was one of Faraday's first attempts  One of those oil filled ones covered in rust. The supply company said they would replace it within 2 years, 10 years ago!
Also, I live in the middle of nowhere so no neighbours.
Andrew
|
|
|
I doubt if the OP knows anything about it.
Are you suggesting I don't know what I'm talking about?
Andrew
|
|
|

'Twas posted before you said anything about it. In case MHC was meaning the power supply to the router.
My broadband basic info/help site - www.robertos.me.uk
My domains,website and mail hosting - Tsohost. Internet connection - O2 Standard.
|
|
|
Fair comment!
Andrew
|
|
|
Resurrecting this thread, I have established that my problem is my own central heating boiler!
The electrical installation is only 6 months old, the boiler is oil fired so my spike is coming from either the boiler thermostat, the fan motor starting in the oil burner or the piezoelectric spark that ignites the oil.
I have checked the boiler earthing, fitted a mains filter to the router power supply, my incoming phone line and router are on opposite sides of the house to the boiler. This has tripped my router three times in the last two hours so is obviously getting worse. It doesn't always happen when the boiler fires up, presumably dependant upon when it happens in the mains cycle.
Any bright suggestions please as to the most likely cause. Suppressing the piezoelectric, not a good idea with the forthcoming cold weather
Andrew
|
|
|
Have you filtered the boiler supply? It may have to be a ferrite on the cable.
However, it may not be transmitted through the cable but radiated and picked up by the router or last length of cable.
The boiler will ignite in a certain sequence: thermostat, water pump, oil pump, extractor fan, ignition ... can you monitor it very closely to see at what stage it happens?
You mention Piezo - is it? It could actually be a very high voltage spark and the circuit to generate that will have some nasty radiation characteristics.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
M H C
taurus excreta cerebrum vincit
|
|
|
Have you filtered the boiler supply? It may have to be a ferrite on the cable.
However, it may not be transmitted through the cable but radiated and picked up by the router or last length of cable.
The boiler will ignite in a certain sequence: thermostat, water pump, oil pump, extractor fan, ignition ... can you monitor it very closely to see at what stage it happens?
You mention Piezo - is it? It could actually be a very high voltage spark and the circuit to generate that will have some nasty radiation characteristics.
Thanks for your learned response
Last things first, looking in the book of words it mentions an ignition transformer with an output of 8kV, not piezo as I thought. Suppressing that would be difficult although it is in a steel box and of course the whole boiler is in a steel case.
The water pump is upstairs and there is no extractor fan (balanced flue) and I have checked, my router re-syncs the moment the burner ignites.
As I said, this has only recently started to be a problem. The boiler is 8 years old and serviced annually but was moved about 6 months ago to a new position in the house and rewired to a new electrical circuit. My router now shares that new circuit so I will try connecting it to the old house wiring to see if that offers any form of isolation.
When you refer to a ferrite on the boiler cable, would a clip on type be adequate like these?
Andrew
|
|
|
Generally speaking your boiler should have a dedicated circuit (I really refer to urban boilers on 15A/16A circuits).
I can't speak for oil-fired boilers though.
my router re-syncs the moment the burner ignites.
Something I've never read before.
|
|
|
On the cores, I think it is experiment - my rf knowledge in that area got pushed out of my brain some years ago
|
|
The author of the above post is a thinkbroadband staff member. It may not constitute an official statement on behalf of thinkbroadband.
|
|
|
my router re-syncs the moment the burner ignites.
Something I've never read before. 
Oh, I try to be original
Not every time though otherwise I'd be back to dial-up speeds!!
Andrew
|
|
|
On the cores, I think it is experiment - my rf knowledge in that area got pushed out of my brain some years ago
Marconi invented a way of sending "wireless" messages using the same principle. Remember that?
Andrew
|
|
|
Sounds as though you have found the actual problem.
The transformer may be in a box but there will be a piece of wire from that to the point of ignition - a great antenna! The transformer may also be operating at a lot higher than 50Hz - incoming mains is rectified and stored at 250 - 300V DC then switched with a free running oscillator which will create a whole series of frequencies in a "comb", that passes through the transformer to provide the 8kV
The ferrites are worth a try - it is a bit of a black art. You can pass the wire through or through, round the outside and through again. If you can get two of different sizes - they will have different characteristics. Fit them close to the boiler. It might also be worth fitting one one the control/thermostat cable.
Even with a balanced flue there could be an internal fan - my boiler has one.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
M H C
taurus excreta cerebrum vincit
|
|
|
Again, thanks for the reply.
Sounds a bit like the line output transformer principal in a proper TV
I will try the ferrites as you suggest!
Andrew
|
|
|
Andrew said that it doesn't happen every time, so I'd guess that it's not the ignition per se, but whether it's the transformer or a fan there's an inductive load suddenly being switched in- do that at the wrong moment in the mains cycle and there'll be quite a spike.
A filter in the supply to the router might do it... I run mine from the UPS, but that might be an expensive cure if there isn't one already there!
|
|
The author of the above post is a thinkbroadband moderator but it does not constitute an official statement on behalf of thinkbroadband.
|
|
|
|
as I understand it, in a conventional oil boiler, the oil is ignited by a spark, this spark continues until heat is detected.
it may be that over the years the spark gap has been eroded, and requires adjustment. or the heat detection need cleaning.
this may not stop the noise all together, but would make it shorter so the router may not resync.
there is also a motor to spin (distribute) the oil, check that as it starts up at the same time.
whats the noise like on an am radio?
|
|
|
the simple test to see if a ups would help on an adsl router is run an old usb one from a laptop while testing.
edit: but you are always better removing noise at its source
Edited by ggremlin (Tue 23-Nov-10 11:28:21)
|
|
|
as I understand it, in a conventional oil boiler, the oil is ignited by a spark, this spark continues until heat is detected.
it may be that over the years the spark gap has been eroded, and requires adjustment. or the heat detection need cleaning.
this may not stop the noise all together, but would make it shorter so the router may not resync.
there is also a motor to spin (distribute) the oil, check that as it starts up at the same time.
whats the noise like on an am radio?
Good point about an eroded spark gap. I will also check using an AM radio, my usual way of checking for thunderstorms within range!
Andrew
|
|
|
Andrew said that it doesn't happen every time, so I'd guess that it's not the ignition per se, but whether it's the transformer or a fan there's an inductive load suddenly being switched in- do that at the wrong moment in the mains cycle and there'll be quite a spike.
A filter in the supply to the router might do it... I run mine from the UPS, but that might be an expensive cure if there isn't one already there!
I tried a filter in the router supply last night but it didn't make any difference. I already have a ring ferrite on the supply lead from the PSU to the router. That's been there for over a year just as a precaution and I had one in stock anyway
Andrew
|
|
|
Yeah remember that, but forgotten all the ways of estimating what is the best size/type of ferrite core for suppressing noise.
As a note: It may not happen all the time, as while a spark is often very similar, sometimes it may just arc a little differently across the gap and this is when the noise occurs.
Could try just wearing lots of jumpers.
|
|
The author of the above post is a thinkbroadband staff member. It may not constitute an official statement on behalf of thinkbroadband.
|
|
|
I can recommend these 'power conditioners': http://www.amazon.co.uk/LINDY-Mains-Conditioner-Powe...
NB. As others have said it does sound as if the boiler probably has a faulty component...
|
|
|
Could try just wearing lots of jumpers.
And save on the cost of oil! We do have a wood-burning stove as well
Andrew
|
|
|
Thanks but I'd have to put a 13 amp plug on the boiler!!
Andrew
|
|
|
Or you could just plug your router into it!
|
|
|
Or you could just plug your router into it! 
I have a single 13A mains filter (looks like a two way adapter with only one socket) that I have already tried on the router mains feed. No difference
Andrew
|
|
|
I've an inline supression device from RS on my boiler, seems get rid of the nasty stuff. Will have a look for the RS stock code when I get home.
Dave
|
|
|
Thanks, please let me know what it is!
Andrew
|
|
|
Here you go, RS part no 3256543 Link hopefully that will work, always seem to have problems with links to RS. There is a higher current version.
Got rid of the rubbish on the mains from the fan and pump from my boiler, also stuck a cap on the room thermostat just for good measure  .
Dave
|
|
|
Thanks very much for that and yes, the link worked OK. I will order one.
However, my boiler has been running for nearly four hours now and not a trace of a spike/pulse on Routerstats and no re-syncs  That is with the router plugged into the old house wiring and not the new to which the boiler is connected suggesting that the problem is transmitted through the wiring. Can't really understand why as all wiring ultimately goes back to the same incoming main but via a separate set of circuit breakers. The boiler circuit is new but the original wiring is 35 years old!
Andrew
|
|
|
Could be something simple like a dodgy connection on the new wiring, might be worth reterminating and checking the connections.
Dave
|
|
|
True, but I think I will contact the gentleman that installed and certified the new wiring, first!!
Andrew
|
|
|
Thanks but I'd have to put a 13 amp plug on the boiler!! 
Nothing technically wrong with this.
But always a danger of being switched off accidently !
|
|
|
In this weather you'd certainly know if it was switched off!!!
Andrew
|
|
|
Use an unswitched socket!
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
M H C
taurus excreta cerebrum vincit
|