Print

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

[beanieboy182]

cheers
they work like a charm
it was my stupid router and typing in the login1.txt and login2.txt and saving them as it didnt save them as unix or whatever it was supposed to be

but as it works now i can see why my line isn't so great
but really i have little to no chance of finding the spike at around 2000 kHz
well without buying something expensive lol

[burakkucat]

but really i have little to no chance of finding the spike at around 2000 kHz
well without buying something expensive

But it doesn't have to expensive! Please take a look at the following links --

http://www.ebay.co.uk/itm/370607013140
http://dangerousprototypes.com/2012/03/20/introducing-rtl-sdr-a-20-sdr/
http://sdr.osmocom.org/trac/wiki/rtl-sdr

Under £12 to buy the hardware (using eBay) and then just build the Open Source software for whatever OS you are using. 

[asbokid]

The gurus of SDR have patched the RTL-SDR driver to allow an antenna to be attached directly to the RTL2832 [1].  The Realtek IC is found in low-cost DVB-T USB sticks. [2][3]   This hack bypasses the Elonics e4000 DVB-T/FM tuner that is found on these sticks. Bypassing gains direct access to RF between 0 - 30MHz, the frequency range used by xDSL.    Before this patch, recovering signals from these low frequencies involved an up-converter like these [4][5]

This patch could turn the ten buck RTL-based DVB-T dongle into an excellent tool for tracking RF noise ingressing an xDSL broadband service.

cheers, a

[1] http://www.realtek.com.tw/products/productsView.aspx?Langid=1&PNid=22&PFid=35&Level=4&Conn=3&ProdID=257
[2] http://groups.google.com/forum/#!topic/ultra-cheap-sdr/cG988R1D8uE
[3] http://cgit.osmocom.org/cgit/rtl-sdr/commit/?h=steve-m/direct_sampling&id=4b49512fccbba68670f86942cc457be0c87d6871
[4] http://george-smart.co.uk/wiki/FunCube_Upconverter
[5] http://www.ct1ffu.com/site/index.php?option=com_content&view=article&id=178&Itemid=104

[burakkucat]

Just between us, b*cat confesses to have ordered one of the appropriate type DVB-T USB sticks [1] from eBay. When it delivered to The Cattery, from Hong Kong, there will be yet another plaything for his amusement. 

[1] http://www.ebay.co.uk/itm/140817268570

[beanieboy182]

well when i get paid i shall invest in that little dongle and antenna
but then i might need some help setting it up and working out how to listen to it and walk around whilst listening for noise or viewing it with software or whatever ? :S

[asbokid]

This forum boasts its own expert on RF filter design <ahem!>  Hopefully he (or she?!) might find a spare moment to cast his (or her) expert eye over the filtering used on the up-converters for these dongles.   Maybe there are ways to improve their performance?

The up-converter allows the DVB-T dongle to be tuned to xDSL frequencies, right down to DC.  Which would make them a great ultra-low cost addition to Walter's legendary Wheelbarrow of Broadband Resuscitation Tools.

At the moment, and even though the up-converters are very simple in design, they still command a crazy price (£50+).   Great Uncle Ebenezer bin Asbo, however, has been studying the up-converters with a view to building something similar.  In terms of the components  - all of which are SMD, even the coils -  the boards should cost almost nothing to make.

Relevant links will be added below..

cheers, a

[burakkucat]

The up-converter allows the DVB-T dongle to be tuned to xDSL frequencies, right down to DC.  Which would make them a great ultra-low cost addition to Walter's legendary Wheelbarrow of Broadband Resuscitation Tools.

Great minds think alike! 

I was considering the possibility of Walter using an up-converter, a DVB-T dongle, a Raspberry Pi, an LCD display and a directional aerial. 

[asbokid]

A team at Northwestern University has released an open source Wavelet-based Peak Detection algorithm for Mass Spectrometry Analysis. [1][2] It can be re-purposed for this comparatively trivial task of RF ingress noise detection on DSL lines.

We start with qln.txt, a two column text file of QLN tonemap data from a Broadcom chipset modem. It is just a capture from the modem shell command xdslcmd info --QLN.

Code: [Select]

$ head -20 qln.txt
"index"         "QLN"
   0            -113.0000
   1            -123.0000
   2            -123.0000
   3            -123.0000
   4            -123.0000
   5            -123.0000
   6            -123.0000
   7            -123.0000
   8            -123.0000
   9            -123.0000
   10           -123.0000
   11           -123.0000
   12           -123.0000
   13           -123.0000
   14           -123.0000
   15           -123.0000
   16           -122.0000
   17           -122.0000
   18           -122.0000

We begin by loading R, the "free software environment for statistical computing and graphics" [3]. Then we install Northwestern's Wavelet-based Peak Detection library. [2].

Code: [Select]

$ R

R version 2.15.1 (2012-06-22) -- "Roasted Marshmallows"
Copyright (C) 2012 The R Foundation for Statistical Computing
ISBN 3-900051-07-0
Platform: x86_64-pc-linux-gnu (64-bit)

> source("http://bioconductor.org/biocLite.R")
BiocInstaller version 1.4.7, ?biocLite for help
> biocLite("MassSpecWavelet")
BioC_mirror: http://bioconductor.org
Using R version 2.15, BiocInstaller version 1.4.7.
Installing package(s) 'MassSpecWavelet'
..
* DONE (MassSpecWavelet)

The PeakDet library is loaded, and we read in our QLN dataset before calling the peak-det functions from the library:

Code: [Select]

> library(MassSpecWavelet)
Loading required package: waveslim
> qlnDAT <- read.table("qln.txt", colClasses = c(rep("NULL",1), rep("numeric",1)), header=TRUE)
> qlnMAT <- data.matrix(qlnDAT)
> wCoefs <- cwt(qlnMAT, scales=seq(1,31,1), wavelet='mexh')
> localMax <- getLocalMaximumCWT(wCoefs)
> ridgeList <- getRidge(localMax)
> majorPeakInfo <- identifyMajorPeaks(qlnMAT, nearbyPeak=TRUE, ridgeList, wCoefs, SNR.Th=1)
> peakIndex <- majorPeakInfo$peakIndex

Which discovers the following Peak Information..

Eight noise peaks in total, using those thresholds. Discovered at DMTs #84  #141  #162  #249  #268  #283  #339  #377,  each of a given intensity.

Code: [Select]

> majorPeakInfo
$peakIndex
 1_84 1_141 1_162 1_249 1_268 1_283 1_339 1_377
   84   141   162   249   268   283   339   377

$peakValue
       1_2       1_30       1_45       1_51       1_66       1_84      1_100
 5.9741241 14.6055427  3.4114940  0.0000000  2.2530691 23.2881416  1.9204691
     1_112      1_120      1_141      1_162      1_200      1_212      1_226
 0.0000000  2.6028832 19.9828253 75.8928215  0.7603451  0.0000000  0.0000000
     1_249      1_268      1_283      1_301      1_323      1_339      1_367
14.3018559 52.1779085 10.8258617  0.0000000  0.0000000 12.3137493  1.2914229
     1_377      1_399      1_409      1_417      1_429      1_442      1_457
 6.7045729  1.1645874  0.0000000  0.0000000  1.4541205  0.0000000  4.2575986
     1_477      1_485
 1.5450179  0.0000000

$peakCenterIndex
  1_2  1_30  1_45  1_51  1_66  1_84 1_100 1_112 1_120 1_141 1_162 1_200 1_212
    3    28    48    51    70    60    96   112   117   140   159   198   212
1_226 1_249 1_268 1_283 1_301 1_323 1_339 1_367 1_377 1_399 1_409 1_417 1_429
  226   249   270   284   301   323   339   367   389   396   409   417   427
1_442 1_457 1_477 1_485
  442   468   478   485

$peakSNR
       1_2       1_30       1_45       1_51       1_66       1_84      1_100
 1.4794073  3.6031302  0.8416022  0.0000000  0.5558233  5.7450933  0.4737722
     1_112      1_120      1_141      1_162      1_200      1_212      1_226
 0.0000000  0.6421211  4.9296847 18.7224618  0.1875742  0.0000000  0.0000000
     1_249      1_268      1_283      1_301      1_323      1_339      1_367
 3.5282118 12.8720857  2.6706977  0.0000000  0.0000000  3.0377537  0.3185890
     1_377      1_399      1_409      1_417      1_429      1_442      1_457
 1.6539919  0.2872992  0.0000000  0.0000000  0.3587258  0.0000000  1.0503329
     1_477      1_485
 0.3811499  0.0000000

$peakScale
  1_2  1_30  1_45  1_51  1_66  1_84 1_100 1_112 1_120 1_141 1_162 1_200 1_212
    5     5     5     0     6    30     5     0     5     5    30     5     0
1_226 1_249 1_268 1_283 1_301 1_323 1_339 1_367 1_377 1_399 1_409 1_417 1_429
    0     5    25     5     0     0     5     5    20     5     0     0     5
1_442 1_457 1_477 1_485
    0    15     8     0

$potentialPeakIndex
 1_84 1_100 1_112 1_120 1_141 1_162 1_200 1_212 1_226 1_249 1_268 1_283 1_301
   84   100   112   120   141   162   200   212   226   249   268   283   301
1_323 1_339 1_367 1_377 1_399 1_409 1_417 1_429
  323   339   367   377   399   409   417   429

$allPeakIndex
  1_2  1_30  1_45  1_51  1_66  1_84 1_100 1_112 1_120 1_141 1_162 1_200 1_212
    2    30    45    51    66    84   100   112   120   141   162   200   212
1_226 1_249 1_268 1_283 1_301 1_323 1_339 1_367 1_377 1_399 1_409 1_417 1_429
  226   249   268   283   301   323   339   367   377   399   409   417   429
1_442 1_457 1_477 1_485
  442   457   477   485

And that data can be plotted with the following command. In the graphs, the noise peaks above our threshold are highlighted in red.

Code: [Select]

> plotPeak(qlnMAT, peakIndex, main=paste("identified QLN peaks"))

Which produces this graph:

(click for full size)

We can also do a baseline correction to visually examine the SNR of each noise peak:

Code: [Select]

> plotRange <- c(1,512)
> peakSNR <- majorPeakInfo$peakSNR
> allPeakIndex <- majorPeakInfo$allPeakIndex
> selInd <- which(allPeakIndex >= plotRange[1] & allPeakIndex < plotRange[2])
> plot(allPeakIndex[selInd], peakSNR[selInd], type='h', xlab='DMT Index', ylab='Peak noise to background ratio')
> points(peakIndex, peakSNR[names(peakIndex)], type='h', col='red')
> peakIndex <- majorPeakInfo$peakIndex
> title('CWT Baseline correction - SNR of peaks - majorPeaks in red')


(click for full size)

Hopefully, it is possible to build these Peak Detection functions as a static library, and to use a C language binding to call them from own code. [4]

The next stage is re-sampling(?) the DSL data to align it with the 9kHz channel spacing of AM broadcast radio.

An idea mentioned earlier is to apply a cluster detection algorithm to those noise peak frequencies, for geolocation purposes. [5]

Perhaps the tool could become a plug-in for a DSL line-monitoring application like the one Roseway is developing  [6]

cheers, a

[1] http://www2.in.tu-clausthal.de/~hammer/lectures/biomlsem/sechs.pdf
[2] http://www.bioconductor.org/packages/release/bioc/html/MassSpecWavelet.html
[3] http://www.r-project.org/
[4] http://stackoverflow.com/questions/7457635/calling-r-function-from-c
[5] http://forum.kitz.co.uk/index.php/topic,11424.msg220565.html#msg220565
[6] http://forum.kitz.co.uk/index.php/topic,11478.0.html

[burakkucat]

Very nice. 

Someone was obviously wide awake whilst b*cat was sleeping! 

[roseway]

Perhaps the tool could become a plug-in for a DSL line-monitoring application like the one Roseway is developing

I think that may be too much for my poor old brain, but I can certainly save the QLN data (and any other useful data) as text files for other programs to use.

[asbokid]

The gurus of SDR have patched the RTL-SDR driver to allow an antenna to be attached directly to the RTL2832 [1].  The Realtek IC is found in low-cost DVB-T USB sticks. [2][3]   This hack bypasses the Elonics e4000 DVB-T/FM tuner that is found on these sticks. Bypassing gains direct access to RF between 0 - 30MHz, the frequency range used by xDSL.    Before this patch, recovering signals from these low frequencies involved an up-converter like these [4][5]

This patch could turn the ten buck RTL-based DVB-T dongle into an excellent tool for tracking RF noise ingressing an xDSL broadband service.

cheers, a

Just between us, b*cat confesses to have ordered one of the appropriate type DVB-T USB sticks [6] from eBay. When it delivered to The Cattery, from Hong Kong, there will be yet another plaything for his amusement. 

An enjoyable evening was spent atop the highest local peak playing with one of these dongles. It even tuned to Mrs Humphreys' (illegal) baby monitor transmitting on 195.250MHz.  Either the tuner has a faulty squelch or that poor child badly needs burping!

The datasheet and reference schematic for the best performing tuner found in these DVB-T dongles, the Elonics E4000 multiband tuner, have finally "come out", courtesy of Ben Ryan - he's the man to hang!  [7]

For those who do not want to sign up for a silly BGW "Live.Com" account, simply to download those datasheets from Ben's Live.Com page, they can sign up for a Google Corp. account to download the files from here instead [8]

Tragically, Elonics, the Edinburgh-based fabless chipmaker, has gone bust. So while the release of documentation on the E4000 is obviously great news, unfortunately, no one is making the tuner ICs any more.   But hopefully an enterprise will soon snap up Elonics' IPR from the Official Receiver.

cheers, a

[1] http://www.realtek.com.tw/products/productsView.aspx?Langid=1&PNid=22&PFid=35&Level=4&Conn=3&ProdID=257
[2] http://groups.google.com/forum/#!topic/ultra-cheap-sdr/cG988R1D8uE
[3] http://cgit.osmocom.org/cgit/rtl-sdr/commit/?h=steve-m/direct_sampling&id=4b49512fccbba68670f86942cc457be0c87d6871
[4] http://george-smart.co.uk/wiki/FunCube_Upconverter
[5] http://www.ct1ffu.com/site/index.php?option=com_content&view=article&id=178&Itemid=104
[6] http://www.ebay.co.uk/itm/140817268570
[7] http://groups.google.com/d/msg/ultra-cheap-sdr/zXhBfiYdxf4/L2r108nnyXQJ
[8] http://docs.google.com/folder/d/0B6wW18mYskvBSnJQeVl0YmVDVzA/edit

EDIT: Links adjusted

[burakkucat]

  There is nothing to report, from The Cattery, on that front. My dongle is still in transit from the orient.