FPGA Central

I've got a 'mystery signal' I recorded. I'm not sure how it is
encoded. I've got pics showing what the wave form looks like. I also
have a sample wave file.

http://www.cravenfamily.com/roller/p...signal_encoded

The few people I've talked to that are familiar with digital
communications didn't recognize the encoding. Any insight would be
appreciated.

Paul Vincent Craven
[email protected]

Hello Paul,

A big part of solving the mystery can be had by telling us how and where you
got this signal. It seems to basically have 4 levels but it looks corrupted
by an AGC. Is this the output of a demodulator? If so what kind of
demodulator is it? If this is from a radio, what frequency are you tuned to?
What bandwidth is your IF?

Clay S. Turner





"Paul Vincent Craven" <[email protected]> wrote in message
news:[email protected] m...
> I've got a 'mystery signal' I recorded. I'm not sure how it is
> encoded. I've got pics showing what the wave form looks like. I also
> have a sample wave file.
>
>
http://www.cravenfamily.com/roller/p...signal_encoded
>
> The few people I've talked to that are familiar with digital
> communications didn't recognize the encoding. Any insight would be
> appreciated.
>
> Paul Vincent Craven
> [email protected]

Paul Vincent Craven wrote:
>
> I've got a 'mystery signal' I recorded. I'm not sure how it is
> encoded. I've got pics showing what the wave form looks like. I also
> have a sample wave file.
>
> http://www.cravenfamily.com/roller/p...signal_encoded

Connecting to www.cravenfamily.com[66.139.77.72]:80... connected.
HTTP request sent, awaiting response... 500 Internal Server Error
21:52:21 ERROR 500: Internal Server Error.


Erik
--
+-----------------------------------------------------------+
Erik de Castro Lopo [email protected] (Yes it's valid)
+-----------------------------------------------------------+
"Some people don't want genitalia shoved down their throats."
-- Rex Mossop, Australian football commentator and morals crusader

> A big part of solving the mystery can be had by telling us how and where you
> got this signal. It seems to basically have 4 levels but it looks corrupted
> by an AGC. Is this the output of a demodulator? If so what kind of
> demodulator is it? If this is from a radio, what frequency are you tuned to?
> What bandwidth is your IF?

It was recorded via narrow-FM radio tuned in to about 450 MHz.
Recording was done via a Kenwood TM-D700A ham radio fed to the line-in
of my notebook. The signal is railroad related.

You are right, AGC could be corrupting it. How does one tell? What is
the easiest/cheapest way of getting around this problem?

The link for the signal:
http://www.cravenfamily.com/roller/p...signal_encoded
....unfortunately has been going up and down today. Sorry if anyone
tried it and couldn't reach the server.

Paul Vincent Craven

> A big part of solving the mystery can be had by telling us how and where you
> got this signal. It seems to basically have 4 levels but it looks corrupted
> by an AGC. Is this the output of a demodulator? If so what kind of
> demodulator is it? If this is from a radio, what frequency are you tuned to?
> What bandwidth is your IF?

It was recorded via narrow-FM radio tuned in to about 450 MHz.
Recording was done via a Kenwood TM-D700A ham radio fed to the line-in
of my notebook. The signal is railroad related.

You are right, AGC could be corrupting it. How does one tell? What is
the easiest/cheapest way of getting around this problem?

The link for the signal:
http://www.cravenfamily.com/roller/p...signal_encoded
....unfortunately has been going up and down today. Sorry if anyone
tried it and couldn't reach the server.

Paul Vincent Craven

Hello Paul,

Given your new input, I would guess then this is a 4 level fsk signal. Flex
paging is one example I've worked with quite a bit that uses this scheme
although this doesn't appear to be Flex. APOC and Mobitex also work
similarly.

The way pagers and other receivers get around the AGC "breathing" problem is
they lock the AGC at the end of the preample. Notice the "comma" pattern at
the beginning. This is likely used to set the AGC and level slicer and to
generate a reference for a bit clock. After this is probably some kind of
barker code which provides for a block sync. After that your guess is as
good as mine. The data may be scrambled for whitening and burst error
resistance. Also the 4 levels are likely to be grey coded, so if the level
slicer is a little off, you only blow one bit per symbol instead of two. You
may try talking to some people in the Railroad industry as they may be able
to give you a name for the protocol. Also look at the vendors for Railroad
related equipment. GE and Harris come to mind. See if they offer any info.

I hope this helps.

Clay





"Paul Vincent Craven" <[email protected]> wrote in message
news:[email protected] m...
> > A big part of solving the mystery can be had by telling us how and where
you
> > got this signal. It seems to basically have 4 levels but it looks
corrupted
> > by an AGC. Is this the output of a demodulator? If so what kind of
> > demodulator is it? If this is from a radio, what frequency are you tuned
to?
> > What bandwidth is your IF?
>
> It was recorded via narrow-FM radio tuned in to about 450 MHz.
> Recording was done via a Kenwood TM-D700A ham radio fed to the line-in
> of my notebook. The signal is railroad related.
>
> You are right, AGC could be corrupting it. How does one tell? What is
> the easiest/cheapest way of getting around this problem?
>
> The link for the signal:
>
http://www.cravenfamily.com/roller/p...signal_encoded
> ...unfortunately has been going up and down today. Sorry if anyone
> tried it and couldn't reach the server.
>
> Paul Vincent Craven

This could also be a Pi/4 DQPSK signal, which looks like 4-level FSK
when discriminated. I think some mobile radio systems use this kind of
modulation.

John

"Clay Turner" <[email protected]> wrote in message news:<vLAUc.16096$[email protected]> ...
> Hello Paul,
>
> Given your new input, I would guess then this is a 4 level fsk signal. Flex
> paging is one example I've worked with quite a bit that uses this scheme
> although this doesn't appear to be Flex. APOC and Mobitex also work
> similarly.
>
> The way pagers and other receivers get around the AGC "breathing" problem is
> they lock the AGC at the end of the preample. Notice the "comma" pattern at
> the beginning. This is likely used to set the AGC and level slicer and to
> generate a reference for a bit clock. After this is probably some kind of
> barker code which provides for a block sync. After that your guess is as
> good as mine. The data may be scrambled for whitening and burst error
> resistance. Also the 4 levels are likely to be grey coded, so if the level
> slicer is a little off, you only blow one bit per symbol instead of two. You
> may try talking to some people in the Railroad industry as they may be able
> to give you a name for the protocol. Also look at the vendors for Railroad
> related equipment. GE and Harris come to mind. See if they offer any info.
>
> I hope this helps.
>
> Clay
>
>
>
>
>
> "Paul Vincent Craven" <[email protected]> wrote in message
> news:[email protected] m...
> > > A big part of solving the mystery can be had by telling us how and where
> you
> > > got this signal. It seems to basically have 4 levels but it looks
> corrupted
> > > by an AGC. Is this the output of a demodulator? If so what kind of
> > > demodulator is it? If this is from a radio, what frequency are you tuned
> to?
> > > What bandwidth is your IF?
> >
> > It was recorded via narrow-FM radio tuned in to about 450 MHz.
> > Recording was done via a Kenwood TM-D700A ham radio fed to the line-in
> > of my notebook. The signal is railroad related.
> >
> > You are right, AGC could be corrupting it. How does one tell? What is
> > the easiest/cheapest way of getting around this problem?
> >
> > The link for the signal:
> >
> http://www.cravenfamily.com/roller/p...signal_encoded
> > ...unfortunately has been going up and down today. Sorry if anyone
> > tried it and couldn't reach the server.
> >
> > Paul Vincent Craven

I'll look into that. I'm not too familiar with that form of encoding.

I posted some more wave images to my site. These include an overlay of
the packet I received one second later. This helps show the packets
are regular in nature, and it also helps point out what parts are
probably just distortion.

http://www.cravenfamily.com/roller/p...ore_wave_files