MultimonNG for POCSAG under Linux

[Adam Nielsen]

Hi all,

I just discovered MultimonNG[1], an extensively updated version of the
original Multimon which now works on 64-bit systems.

I was able to get this working very easily with my RTL SDR, using the rtl_fm
program distributed as part of Osmocom's librtlsdr.

If anyone is interested, here is the command I used to decode the POCSAG
transmission on 148.629MHz and display the output:

rtl_fm -f 148639000 -s 22050 - | ./multimonNG -a POCSAG512 -a POCSAG1200 -f
alpha -t raw /dev/stdin

This tunes to a local POCSAG transmitter in Brisbane, Australia, and starts
decoding in both POCSAG512 and POCSAG1200 modes (this station broadcasts each
alternately.)

It produces output like this:

POCSAG1200-: Alpha: 15:05 28/10 ALARM - UPS THEATRE ALARM
POCSAG1200-: Alpha: 15:06 28/10 ALARM - MEDICAL/LAB GAS<NUL><NUL>
POCSAG512-: Alpha: ScadaC API Communication Failed
POCSAG1200-: Alpha: ESCL#2234:ESC NO. 2233 CONFIRMED BY MARCUS W<NUL>
POCSAG1200-: Alpha: PLS ADVISE COMMS WHAT TIME YOU COMMENCED YOUR MEAL.
THX<NUL><NUL>

It seems to work quite well, and uses very little CPU (~10% when decoding.)
If you don't know whether there are any POCSAG stations in your area, download
the test file on the MultimonNG page and have a listen to the .wav file.
POCSAG has quite a distinctive sound, so you will most likely recognise it if
you have stumbled across it before in HDSDR or similar.

Cheers,
Adam.

[1] http://dekar.wc3edit.net/2012/05/24/multimonng/

[KD9GN]

Hi Adam,

Thanks for the info .. I updated my rtl-sdr and installed multimonNG and
have it working on CentOS 6.3. I have a signal close enough I can use
my ezCap 668 with the tiny little whip antenna that came with it and it
seems to be working well.

Thanks,

Dave - KD9GN

[limaunion]

Hi Dave, thanks for the info!

But there is something that I don't understand, for example:

$ rtl_fm -f 186285000 -s 22050 - | ./multimonNG -a POCSAG512 -a POCSAG1200 -f alpha -t raw /dev/stdin
Found 1 device(s):
  0:  Realtek, RTL2838UHIDIR, SN:

Using device 0: ezcap USB 2.0 DVB-T/DAB/FM dongle
multimonNG  (C) 1996/1997 by Tom Sailer HB9JNX/AE4WA
            (C) 2012 by Elias Oenal
available demodulators: POCSAG512 POCSAG1200 POCSAG2400 EAS UFSK1200 CLIPFSK AFSK1200 AFSK2400 AFSK2400_2 AFSK2400_3 HAPN4800 FSK9600 DTMF ZVEI SCOPE
Enabled demodulators: POCSAG512 POCSAG1200
Found Fitipower FC0013 tuner
Oversampling input by: 46x.
Oversampling output by: 1x.
Buffer size: 8.08ms
Tuned to 186538575 Hz.
Sampling at 1014300 Hz.
Output at 22050 Hz.
Exact sample rate is: 1014300.020041 Hz
Tuner gain set to automatic.

Why if I set the freq. 186285000 the output shows 'Tuned to 186538575 Hz.' ? any idea? thanks.
LU

[limaunion]

one more, is there any possibilty to hear the audio while running rtf_fm + multimonNG?

[KD9GN]

I just started using rtl_fm and multimonNG last night so I really don't
know much about it. As far as listening to audio and running multimonNG
at the same time, I am not sure if you can do that or not.

I use SDRSharp and PDW 3.1 on my windows box and you can listen to the
audio while watching it decode.

> <http://dekar.wc3edit.net/2012/05/24/multimonng/>
>
> --
> You received this message because you are subscribed to the Google
> Groups "Ultra Cheap SDR" group.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/ultra-cheap-sdr/-/sQ3eKtt51lAJ.
> To post to this group, send email to ultra-c...@googlegroups.com.
> To unsubscribe from this group, send email to
> ultra-cheap-s...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/ultra-cheap-sdr?hl=en.

[Adam Nielsen]

> Why if I set the freq. 186285000 the output shows 'Tuned to 186538575
> Hz.' ? any idea? thanks.

I think this is because of the centre spike in the E4000. It is
designed to tune to a slightly different frequency and then decode the
broadcast off-centre, so any centre spike doesn't interfere with the signal.

If you want to hear the audio while decoding, you can use multiple pipes
(only tested in Bash) similar to this:

$ rtl_fm -l 0 -f 148637000 -s 22050 - | tee >(./multimonNG -a POCSAG512
-a POCSAG1200 -f alpha -t raw /dev/stdin) >(play -t raw -r 22050 -e
signed-integer -b 16 -c1 -V1 -q -) > /dev/null

There is a slight delay for me, but I still hear the audio as the
message is being printed on the screen.

Note that I have added the -l option in the command above which reduces
the squelch level. This is needed to receive signals considered too
weak by the default value (-l 150).

Cheers,
Adam.

[fw8...@googlemail.com]

Hello,
thanks for this group!
I've installed the packages on my rpi and have Cinergy DVB-T Stick. 
Then I've set the frequency. But multimon says everytime: 

pi@raspberrypi ~ $ rtl_fm -N -o 4 -A lut -s 22050 -C -f 172.680M - | multimon-ng -t raw -a POCSAG512 -a POCSAG1200 -a POCSAG2400 -f alpha /dev/stdin
POCSAG512 POCSAG1200 POCSAG2400 EAS UFSK1200 CLIPFSK AFSK1200 AFSK2400 AFSK2400_2 AFSK2400_3 HAPN4800 FSK9600 DTMF ZVEI1 ZVEI2 ZVEI3 DZVEI PZVEI EEA EIA CCIR MORSE_CW DUMPCSV SCOPE
multimon-ng (C) 1996/1997 by Tom Sailer HB9JNX/AE4WA
(C) 2012/2013 by Elias Oenal
available demodulators: POCSAG512 POCSAG1200 POCSAG2400 EAS UFSK1200 CLIPFSK AFSK1200 AFSK2400 AFSK2400_2 AFSK2400_3 HAPN4800 FSK9600 DTMF ZVEI1 ZVEI2 ZVEI3 DZVEI PZVEI EEA EIA CCIR MORSE_CW DUMPCSV SCOPE
Enabled demodulators: POCSAG512 POCSAG1200 POCSAG2400
Found 1 device(s):
0: Realtek, RTL2838UHIDIR, SN: 00000001

Using device 0: Terratec Cinergy T Stick RC (Rev.3)
Found Elonics E4000 tuner
Oversampling input by: 12x.
Oversampling output by: 4x.
Buffer size: 7.74ms
Tuned to 172944600 Hz.
Sampling at 1058400 Hz.
Output at 22050 Hz.
Exact sample rate is: 1058400.010094 Hz

Tuner gain set to automatic.

POCSAG1200: Address: 123456 Function: 2
POCSAG1200: Address: CORRUPT Function: CORRUPT
----------------------------------------------------------------------------------
What's wrong? Corrupt isn't good. I can't read the messages. Sometimes i can read one part of a word, but then it says CORRUPT...
Is it normaly that my set frequency (172.680 MHz) Tuned to 172944600 Hz??
Should i change the configuration?

Thanks for your help .
Felix

[Adam Nielsen]

> POCSAG1200: Address: 123456 Function: 2
> POCSAG1200: Address: CORRUPT Function: CORRUPT
> ----------------------------------------------------------------------------------
> What's wrong? Corrupt isn't good. I can't read the messages.
> Sometimes i can read one part of a word, but then it says CORRUPT...

I haven't seen this before so my guess is either "CORRUPT" is the
POCSAG message being broadcast, or perhaps your signal is too weak to
decode. If you listen to the frequency in SDR# or gqrx, can you clearly
hear the POCSAG signal with no interference or noise?

> Is it normaly that my set frequency (172.680 MHz) Tuned to 172944600
> Hz?? Should i change the configuration?

Yes this is normal, the stick is using offset tuning so it will end up
on a slightly different frequency than the signal you are decoding.

Cheers,
Adam.

[fw8...@googlemail.com]

Hey Adam,

Thanks for your answer!

If I use SDR# on my PC, the decoding is okay.

and if i use a scanner with discriminator on my other PC with POC32, the decoding is good too. 

There's no or not much noise and I could read the full messages. The next sender is only 500 meters away, so the signal must be okay.

I will decode messages with my raspberry pi. Have you some other ideas to decode the messages?

Maybe a alternative to multimon? I've put a screenshot to this post. 

Is there a simple way to get the soundsignal out of the earphone-socket?

Thanks Felix 

[David J Taylor]

Hey Adam,
Thanks for your answer!

If I use SDR# on my PC, the decoding is okay.
and if i use a scanner with discriminator on my other PC with POC32, the
decoding is good too.
There's no or not much noise and I could read the full messages. The next
sender is only 500 meters away, so the signal must be okay.

I will decode messages with my raspberry pi. Have you some other ideas to
decode the messages?
Maybe a alternative to multimon? I've put a screenshot to this post.

Is there a simple way to get the soundsignal out of the earphone-socket?
Thanks Felix

==================================

Felix,

I tried the commands in your screenshot but, although I didn't get an error
message immediately after the "Tuner gain set to automatic" message, the
command prompt returned. Any ideas?

To get the sound out, try piping the rtl_fm output to something like:

... | aplay -t raw -r 32000 -c 1 -f S16_LE

Doubtless you'll need slightly different parameters. The level from the 3
mm socket is rather low, though.

Thanks,
David
--
SatSignal Software - Quality software written to your requirements
Web: http://www.satsignal.eu
Email: david-...@blueyonder.co.uk

[David J Taylor]

Felix,

I tried the commands in your screenshot but, although I didn't get an error
message immediately after the "Tuner gain set to automatic" message, the
command prompt returned. Any ideas?

[]
===========================

I fixed my problem by assing a space between the "s" and "r" qualifiers and
their arguments, and remembering to add "-" in front of the "r".

-r 22050 -s 88200

You may have omitted the spaces in your command.

Cheers,

[Adam Nielsen]

> There's no or not much noise and I could read the full messages. The
> next sender is only 500 meters away, so the signal must be okay.
>
> I will decode messages with my raspberry pi. Have you some other
> ideas to decode the messages?
> Maybe a alternative to multimon? I've put a screenshot to this post.

I wonder whether the problem is related to the Pi being big-endian and
multimon perhaps only designed for little-endian? Can you try recording
the audio on the good PC, and pipe it to multimon on the Pi? Or the
reverse, record the FM audio on the Pi and send it to a working decoder
on another PC?

That might reveal whether the problem lies with the signal getting to
the Pi or with multimon.

Cheers,
Adam.

[David J Taylor]

From: Adam Nielsen

I wonder whether the problem is related to the Pi being big-endian and
multimon perhaps only designed for little-endian? Can you try recording
the audio on the good PC, and pipe it to multimon on the Pi? Or the
reverse, record the FM audio on the Pi and send it to a working decoder
on another PC?

That might reveal whether the problem lies with the signal getting to
the Pi or with multimon.

Cheers,
Adam.

==============================

Adam,

In my own, admittedly very, brief tests I was seeing some clear text (with
no errors) and, once every so often, a plain text "test/sequence" message.

THIS IS A TEST PERIODIC PAGE SEQUENTIAL NUMBER 7412<EOT><EOT>

It suggests to me that the some messages that I see here might be encrypted,
which isn't surprising, or that baud rate or frequency could be slightly
off. Address and Function values were often corrupt as well. I did check
that the Pi's CPU wasn't being maxed out, and it was not. This on 153.325
MHz, UK.

[jdow]

Have you considered the possibility of overload?

{^_^} Joanne/W6MKU

[David J Taylor]

Have you considered the possibility of overload?

{^_^} Joanne/W6MKU

====================================

Yes, that was my first thought as well. I was watching the CPU usage and
the multimon-ng usage depended on the number of decoders activated, using 10
% with one decoder (PCOSAG1200), while the rtl_fm usage was constant at
around 22-23 %.

Cheers,

[jdow]

On 2014/01/05 00:16, David J Taylor wrote:
> Have you considered the possibility of overload?
>
> {^_^} Joanne/W6MKU
> ====================================
>
> Yes, that was my first thought as well. I was watching the CPU usage and the
> multimon-ng usage depended on the number of decoders activated, using 10 % with
> one decoder (PCOSAG1200), while the rtl_fm usage was constant at around 22-23 %.
>
> Cheers,
> David

Signal strength overload. That could explain the problems with the really
close by station, you know.

{^_^} Joanne/W6MKU

[David J Taylor]

Signal strength overload. That could explain the problems with the really
close by station, you know.

{^_^} Joanne/W6MKU

=====================================

Ah, I thought you meant CPU overload, as I have seen that before. HI!

I don't believe that's the problem here, as I'm using a relatively small
indoor antenna. Not that there's any indication from rtl_fm as to what the
signal strength is, though. What we both seem to be seeing, though, is some
text comes out OK, in complete sentences, and other test does not.

But I will re-test with a smaller antenna and see what happens. Very
similar results. Sometimes there are perhaps 10 or 20 characters correct,
followed by a <CORRUPT> indicator, perhaps suggesting baud rate is slightly
off? I don't know whether the decoder is adaptive of not. The dongle has
been permanently powered via a hub, so should have warmed up, and I have
measured its frequency error and applied that to rtl_fm. Checking a few PPM
either way won't do any harm, though.

Thanks for triggering some thoughts!

[jdow]

On 2014/01/05 03:39, David J Taylor wrote:
> Signal strength overload. That could explain the problems with the really
> close by station, you know.
>
> {^_^} Joanne/W6MKU
> =====================================
>
> Ah, I thought you meant CPU overload, as I have seen that before. HI!
>
> I don't believe that's the problem here, as I'm using a relatively small indoor
> antenna. Not that there's any indication from rtl_fm as to what the signal
> strength is, though. What we both seem to be seeing, though, is some text comes
> out OK, in complete sentences, and other test does not.
>
> But I will re-test with a smaller antenna and see what happens. Very similar
> results. Sometimes there are perhaps 10 or 20 characters correct, followed by a
> <CORRUPT> indicator, perhaps suggesting baud rate is slightly off? I don't know
> whether the decoder is adaptive of not. The dongle has been permanently powered
> via a hub, so should have warmed up, and I have measured its frequency error and
> applied that to rtl_fm. Checking a few PPM either way won't do any harm, though.
>
> Thanks for triggering some thoughts!
>
> David

You can monitor the frequency using SDRSharp (if Windows) or LinRad (if
Linux) and get a feel for the strength of the signals. Do remember that
the dynamic range of these little beasties is somewhat limited.

{^_^} Joanne/W6MKU

[David J Taylor]

You can monitor the frequency using SDRSharp (if Windows) or LinRad (if
Linux) and get a feel for the strength of the signals. Do remember that
the dynamic range of these little beasties is somewhat limited.

{^_^} Joanne/W6MKU

================================================

Thanks, Joanne. Eventually got Linrad to compile on the Raspberry Pi (which
takes some time) but when I run it I get:

svgalib:Cannot open /dev/mem

Likely it needs privilege to run - yes, now I get allocated a virtual
console. So given that to run rtl_fm I use the command:

rtl_fm -N -f 153.325M -p +108 -o 4 -l 10 -A fast -r 22050 -s 88200 -C -D

how would I then use LinRad to determine the signal strength? I'm running
from a text-only command prompt.

Thanks,
David GM8ARV

[jdow]

On 2014/01/05 08:20, David J Taylor wrote:
> You can monitor the frequency using SDRSharp (if Windows) or LinRad (if
> Linux) and get a feel for the strength of the signals. Do remember that
> the dynamic range of these little beasties is somewhat limited.
>
> {^_^} Joanne/W6MKU
> ================================================
>
> Thanks, Joanne. Eventually got Linrad to compile on the Raspberry Pi (which
> takes some time) but when I run it I get:
>
> svgalib:Cannot open /dev/mem
>
> Likely it needs privilege to run - yes, now I get allocated a virtual console.
> So given that to run rtl_fm I use the command:
>
> rtl_fm -N -f 153.325M -p +108 -o 4 -l 10 -A fast -r 22050 -s 88200 -C -D
>
> how would I then use LinRad to determine the signal strength? I'm running from
> a text-only command prompt.
>
> Thanks,
> David GM8ARV

LinRad requires a GUI.

In that case you are sort of out of luck as best I can tell. Simply try
telling the dongle to run a lower gain. If that makes the stronger signal
receive properly you have overload-itis. Without specification it seems
you get AGC. So that makes overload a little less likely. But it does not
mean it won't happen.

I'd consult the source code to see exactly what -s and -r do. The numbers
seem low, especially if one of them is trying to tell the dongle what sample
rate to use. The help output seems to indicate -s is the output sample
rate. But, then, what is -r and what is the dongle sample rate?

And is multimon designed to work with deemphasis on? The sample command
suggests it is not. Have you experimented?

And have you brought some GUI based appliance up on a laptop and carried
it out to where the RPi is and hooked it up to watch signal levels for
awhile?

rtl_fm should be able to parse command-line input to change parameters on
the fly. It would make this sort of debugging much easier, especially if
it sent the command responses out the debut output.

{^_^} Joanne/W6MKU

[David J Taylor]

LinRad requires a GUI.

In that case you are sort of out of luck as best I can tell. Simply try
telling the dongle to run a lower gain. If that makes the stronger signal
receive properly you have overload-itis. Without specification it seems
you get AGC. So that makes overload a little less likely. But it does not
mean it won't happen.

I'd consult the source code to see exactly what -s and -r do. The numbers
seem low, especially if one of them is trying to tell the dongle what sample
rate to use. The help output seems to indicate -s is the output sample
rate. But, then, what is -r and what is the dongle sample rate?

And is multimon designed to work with deemphasis on? The sample command
suggests it is not. Have you experimented?

And have you brought some GUI based appliance up on a laptop and carried
it out to where the RPi is and hooked it up to watch signal levels for
awhile?

rtl_fm should be able to parse command-line input to change parameters on
the fly. It would make this sort of debugging much easier, especially if
it sent the command responses out the debut output.

{^_^} Joanne/W6MKU

======================================================

Thanks for the comments and suggestions, Joanne. I don't know how o use
LinRad to see the gain, and I don't think there's a graphics server even
installed on that Raspberry Pi, certainly not running.

Parameters - with rtl_fm, -s sets the sample rate, and -r the output rate.
Not quite sure why the example I copied had -r as 22050 and -s as 88200.
Looks like the bandwidth for this 1200 baud signal needs to be around 3 KHz.
Removing the -s 88200 seemed to make no difference. Noted that I have the
de-emphasis set with the -D parameter, but testing without seemed to give
even worse decoding. rtl_fm says the sampling rate is 1.056 MHz.

I had a play with the gain settings. At +18 dB there was no decoding at
all. From +24 to +48 dB decoding had similar errors, so for now I have set
the gain to +30 dB. The decoding was similar when the gain was allowed to
be automatic.

73,

[jdow]

Lug a laptop that supports a program that speaks RTLSDR to the base of the
antenna feed and monitor with that.

{^_^}

[David J Taylor]

Lug a laptop that supports a program that speaks RTLSDR to the base of the
antenna feed and monitor with that.

{^_^}

===========================

Easier just to plug the antenna into the spectrum analyser, to be honest!
May do that after tea.

[jdow]

On 2014/01/06 09:58, David J Taylor wrote:
> Lug a laptop that supports a program that speaks RTLSDR to the base of the
> antenna feed and monitor with that.
>
> {^_^}
> ===========================
>
> Easier just to plug the antenna into the spectrum analyser, to be honest! May do
> that after tea.
>
> 73,
> David GM8ARV

But but but but but That's Cheating!

{O,o} Ack Plblbbltptb! <- totally crazed

[David J Taylor]

From: jdow

But but but but but That's Cheating!

{O,o} Ack Plblbbltptb! <- totally crazed

=======================================

It was actually quite enlightening - just how to measure the signal level.
I've not yet learned enough about the Rigol DSA815 to be sure I'm making the
measurement correctly. Setting the SA to zero span, 30 KHz bandwidth, I see
levels of about -48 dBm. With similar settings, the local FM radio (from 40
km away) is -65 dBm. So the pagers really are quite strong, and I think
they are on a mast just a few hundred metres away. This with an indoor
antenna - in the shack.

Some years back I was able to borrow an HP analyser from my then work, and
on a loft antenna it showed levels on 137 MHz of around -35 dBm from QRM to
satellites:

http://www.satsignal.eu/wxsat/pager-probs-new-antenna.htm

[jdow]

On 2014/01/06 11:33, David J Taylor wrote:
> From: jdow
>
> But but but but but That's Cheating!
>
> {O,o} Ack Plblbbltptb! <- totally crazed
> =======================================
>
> It was actually quite enlightening - just how to measure the signal level. I've
> not yet learned enough about the Rigol DSA815 to be sure I'm making the
> measurement correctly. Setting the SA to zero span, 30 KHz bandwidth, I see
> levels of about -48 dBm. With similar settings, the local FM radio (from 40 km
> away) is -65 dBm. So the pagers really are quite strong, and I think they are
> on a mast just a few hundred metres away. This with an indoor antenna - in the
> shack.
>
> Some years back I was able to borrow an HP analyser from my then work, and on a
> loft antenna it showed levels on 137 MHz of around -35 dBm from QRM to satellites:
>
> http://www.satsignal.eu/wxsat/pager-probs-new-antenna.htm
>
> 73,
> David GM8ARV

Those numbers suggest you might get away with simply using a smaller
antenna or reducing the dongle's RF gain setting if either signal is of
interest. The larger one is probably well out of the dongle's dynamic
range.

{^_^} Joanne/W6MKU

[David J Taylor]

From: jdow

Those numbers suggest you might get away with simply using a smaller
antenna or reducing the dongle's RF gain setting if either signal is of
interest. The larger one is probably well out of the dongle's dynamic
range.

{^_^} Joanne/W6MKU

=========================================

Thanks, Joanne. Yes, I already reported that I could reduce the RF gain to
a level where decoding stopped. The antenna is the TV antenna which came
with a stick, some 130 mm in length.

However, I don't feel that signal strength would fully explain why some
messages decode to plain text perfectly, and yet others are garbled, as
others have reported.

[fw8...@googlemail.com]

Hi @all!

Thanks for your help and your ideas!

This group is really great :).

But I've still the <corrupt> problem...

I've tried a few antennas like a 2m-Diamond-antenna, a TV-antenna, the antenna of the DVB-T-Stick and I have no differences.

Also I've tried to power the stick over a powered USB-hub ... no improvement.

The idea with the space between -r 22050 -s 88200 ... no improvement.

Now I've a speaker at my raspberry and hear the signal with noise and not in good quality..

Should I change my frequency? Or gain? 

Our pager-frequency is 172,680MHz, the Modulation of Pocsag is 4kHz (1200 Baud).

With best greetings from Germany

Felix

[Jesse Burt]

Hey guys/gals,

Those having troubles decoding - have you tried looking at the signal on the scope (enable the SCOPE "demodulator" in multimonNG with ' -a SCOPE ') just to see what multimonNG is seeing? I'm not sure how tolerant multimonNG is of clipping, or square waves that aren't so square (it might sound clear to your ears, but maybe there's some extra noise in there that it can't cope with).

Just as an example, with waves that look like the attached PNG, it decodes flawlessly on my machine.

Cheers,

Jesse

[David J Taylor]

Cheers,
Jesse
====================================

Thanks for that suggestion, Jesse. I've now configured the Windows Xming
program to talk to the Raspberry Pi which gives me a much better view of
what's going on. Sadly, my waveforms are nothing like as clean as yours,
but at least I can see some differences when playing with the frequency
offset, confirming its value, but no differences with gain.

As a matter of interest, what command line are you using with rtl_fm?

Looking at the waveform suggested that some of the signals are
multi-frequency, not just two frequencies as I expected for pure POCSAG1200,
so likely they are not, in fact, POCSAG 1200 but some other form of data
modulation. This is in the UK around 153 MHz - whether that's the practice
(more than just POCSAG on one frequency) in other countries I don't know.

Cheers,
David

[Jesse Burt]

David,

On Thursday, January 9, 2014 5:24:06 AM UTC-5, David Taylor, GM8ARV wrote:

As a matter of interest, what command line are you using with rtl_fm?

I used the command line from the top level post in this thread (after having a great deal of trouble trying to get decoding working too, using a completely different method: ggrx to tune in and just running multimon separately - it just wasn't getting any audio that way... blast). Just modified to include the 2400bps rate, the scope, and my local frequency:

rtl_fm -f 152592300 -s 22050 - | ./multimonNG -a POCSAG512 -a POCSAG1200 -a POCSAG2400 -a SCOPE -f alpha -t raw /dev/stdin

Looking at the waveform suggested that some of the signals are
multi-frequency, not just two frequencies as I expected for pure POCSAG1200,
so likely they are not, in fact, POCSAG 1200 but some other form of data
modulation.  This is in the UK around 153 MHz - whether that's the practice
(more than just POCSAG on one frequency) in other countries I don't know.

I get some like in my first screenshot, then some that look like this:

It seems the pager data gets spat out soon after the data that looks like my first screenshot though, so not sure what this part is.

FWIW, mine is on the same band as yours, too, right around 153MHz.

Cheers,
Jesse

[jdow]

That signal may be FLEX four level at 3200 or 6400 bps.

http://en.wikipedia.org/wiki/FLEX_%28protocol%29

{^_^} Joanne/W6MKU

On 2014/01/09 02:49, Jesse Burt wrote:
> David,
>
>
> On Thursday, January 9, 2014 5:24:06 AM UTC-5, David Taylor, GM8ARV wrote:
>
> As a matter of interest, what command line are you using with rtl_fm?
>
>
> I used the command line from the top level post in this thread (after having a
> great deal of trouble trying to get decoding working too, using a completely
> different method: ggrx to tune in and just running multimon separately - it just
> wasn't getting any audio that way... blast). Just modified to include the
> 2400bps rate, the scope, and my local frequency:
>
> |

> rtl_fm -f 152592300-s 22050-|./multimonNG -a POCSAG512 -a POCSAG1200 -a

> POCSAG2400 -a SCOPE -f alpha -t raw /dev/stdin
> |
>
>
> Looking at the waveform suggested that some of the signals are
> multi-frequency, not just two frequencies as I expected for pure POCSAG1200,
> so likely they are not, in fact, POCSAG 1200 but some other form of data
> modulation. This is in the UK around 153 MHz - whether that's the practice
> (more than just POCSAG on one frequency) in other countries I don't know.
>
> I get some like in my first screenshot, then some that look like this:
>

> <https://lh5.googleusercontent.com/-XXQupsLPWic/Us59agmUQHI/AAAAAAAAD6o/c8XSGXi-3Zw/s1600/mmonNG-scope2.png>

>
>
> It seems the pager data gets spat out soon after the data that looks like my
> first screenshot though, so not sure what this part is.
> FWIW, mine is on the same band as yours, too, right around 153MHz.
>
> Cheers,
> Jesse
>

> --
> You received this message because you are subscribed to the Google Groups "Ultra
> Cheap SDR" group.
> To unsubscribe from this group and stop receiving emails from it, send an email
> to ultra-cheap-s...@googlegroups.com.
> To post to this group, send email to ultra-c...@googlegroups.com.
> Visit this group at http://groups.google.com/group/ultra-cheap-sdr.
> For more options, visit https://groups.google.com/groups/opt_out.

[David J Taylor]

That signal may be FLEX four level at 3200 or 6400 bps.

http://en.wikipedia.org/wiki/FLEX_%28protocol%29

{^_^} Joanne/W6MKU

=================================

Yes, Joanne, that looks possible. The only decoder I could find quickly
requires gnu-radio, which is a little too much for the modest Raspberry Pi.
Interesting to know, though.

73,
David GM8ARV

[David J Taylor]

David,

I used the command line from the top level post in this thread (after having
a great deal of trouble trying to get decoding working too, using a
completely different method: ggrx to tune in and just running multimon
separately - it just wasn't getting any audio that way... blast). Just
modified to include the 2400bps rate, the scope, and my local frequency:

rtl_fm -f 152592300 -s 22050 - | ./multimonNG -a POCSAG512 -a POCSAG1200 -a
POCSAG2400 -a SCOPE -f alpha -t raw /dev/stdin

I get some like in my first screenshot, then some that look like this:

It seems the pager data gets spat out soon after the data that looks like my
first screenshot though, so not sure what this part is.
FWIW, mine is on the same band as yours, too, right around 153MHz.

Cheers,
Jesse

================================

Jesse,

Thanks for that. The multi-level waveform looks just like mine. The
command line I have been using is:

rtl_fm -N -f 153.325M -p +108 -g 36 -o 4 -l 10 -A fast -r 22050 -C -D - |
multimon-ng -t raw -a POCSAG1200 -f alpha /dev/stdin

(all on a single line). So you can see that I set the frequency to nominal
and then use the -p +108 to set the parts per million offset, the "-o 4"
sets the output oversampling and the waveforms did not look as good without,
the "-A fast" sets the best trig function lookup although I didn't see a lot
of different in CPU between the three options, the -l 10 is a squelch level,
and I haven't checked that setting, "-r" is supposed to be the same as
"-s", -C I could not find documented, and -D applies de-emphasis, but only
on wideband FM, which we aren't using.

I'd do more debugging with the SCOPE option, but my main PC crashed when I
was using that with X over Xming, likely completely unrelated, but not an
encouragement! Perhaps I can find an X-server for the iPad?

Thanks again for your comments and interest.

[jdow]

On 2014/01/09 06:47, David J Taylor wrote:
> Perhaps I can find an X-server for the iPad?

Not a good idea at this moment. They have just discovered a very serious
vulnerability that has been in the X system for the last 23 years. Wait
until you are sure the version you try to put on your iPhone has that
problem fixed.

{o.o} Joanne

[Thomas Noel]

The vulnerability requires previously trying to install a type of bit-map font that has not been commonly used since early '90s, and it is not present in current 10.9 Mac OS or IOS 7. There is a reason it has not even been noticed in 23 years.
In any case it is just another escalation of privilege vulnerability. Don't operate any important computer as ROOT.

Thomas Noel
KF7RSF

[David J Taylor]

The vulnerability requires previously trying to install a type of bit-map
font that has not been commonly used since early '90s, and it is not present
in current 10.9 Mac OS or IOS 7. There is a reason it has not even been
noticed in 23 years.
In any case it is just another escalation of privilege vulnerability. Don't
operate any important computer as ROOT.

Thomas Noel
KF7RSF

==========================================

Thanks for the information. I have to laugh a little since about 70% of my
work on the Raspberry Pi requires using "sudo" in front of commands!

73,
David GM8ARV

[David J Taylor]

From: jdow

Not a good idea at this moment. They have just discovered a very serious
vulnerability that has been in the X system for the last 23 years. Wait
until you are sure the version you try to put on your iPhone has that
problem fixed.

{o.o} Joanne

========================================

Joanne,

I didn't find a free X server in any case, so I went with VNC which seems a
little slower than X on the PC, but still very usable. I'm using a iPad
BTW, I think tryong to run a high-resolution display on an iPhone would be
tiresome at best.

73,
David GM8ARV