Homing in on the mystery of Pulsy McGrooder

A little while ago, I posted a message talking about some “mysterious”
pulse-like interference I’ve been receiving
with my Gnu Radio/USRP/DBS_RX setup, using my radio-astronomy receiver
software. I affectionately
dubbed this signal “Pulsy McGrooder”. Ok, so I’m a bit strange.

Tonight, I did some hunting around in frequency, looking for where this
signal was strongest. I found that I could
hear it across the entire passband of my front-end filter, which
covers 995-1600Mhz. I confirmed that it dropped
off sharply at the edges of the filters passband (good thing, since I
paid $75.00 for that filter!!!).

I found that the signal was strongest near 1350Mhz. I used peak-hold
in my spectral display, and found that it was
a fairly broad feature, roughly 20dB out of the noise on peak-hold.
With averaging turned on, you couldn’t
see it clearly, but peak-hold worked like a champ.

So, I went looking for 1350Mhz as some kind of “magic” frequency.
Turns out it’s a very popular airport
radar frequency, although I don’t think that it’s used from passing
aircraft, but rather as part of the
ATC radar network.

I looked up that frequency range in Industry Canadas frequency
allocation search tool, and found that NAV CANADA
had several radars in Ontario in that frequency range, including two
at the Ottawa airport–one right at 1350Mhz.
But, Ottawa is 60Km away from me, I would have thought the signal
would be too weak for me to see.

Not so, as it turns out. I ran a path-loss calculation, using the Tx
Power data in the Industry Canada spectrum
database. That radar in Ottawa is at +45dBW (+75dBm). I assumed a
20dB gain transmit antenna, 60km
free-space path, and 30dBi gain for my receiving antenna. Plugging
that all in to a path loss calculator, I find
that the radar signal is arriving at my feed at approximately
-11dBm. Which is just a smoking strong
signal, in anybodys book!

So, how is this getting into the passband of the receive chain? I have
the DBS_RX configured to low-pass-filter
the I and Q to 2Mhz, but the signal levels from that radar getting
into the DBS_RX must be quite high–I have
about 35dB of gain in front of the DBS_RX.

Given how strong this signal is here, 60Km away, I imagine that Ku-band
satellite users near the airport must
suffer quite a bit of interference, even though good RG6 has about
80dB of isolation!

Is anybody aware of 1350Mhz being used from aircraft themselves, rather
than as passive reflectors?

giggle

Marcus … that’s an aircraft transponder. Except for some vintage
1930’s aircraft that don’t have an electrical system - every plane /
helicopter has one. Hit Google or Yahoo! for “Mode-C” and “Mode-S”
transponder.

The ground radar “pings” the aircraft’s transponder (interrogation). The
transponder sends back a reply. Some replies carry a 12 bit ID code plus
altitude. Others might have actual GPS coordinates.

-rick

Here’s a link on Mode A & C transponders (does not include newer mode-S)

http://www.airsport-corp.com/modec.htm

-rick

David I. Emery wrote:

The transponders are 1090/1030 mhz and not 1350. 1350 is just
radar.

There are double pulses that I’m seeing, with variable timing between
the main pulse
and the sub-pulse. The other 1350Mhz radar is much further away from
me, but
perhaps the “sub pulses” I’m seeing are coming from the other radar
station, and
they’re drifting in and out of phase with respect to one another. The
sub pulses
are weaker than the main pulses by quite a bit.

-11 dbm sounds rather loud.

At 1350, the pulses are arriving 35-40dB out of the noise. But I agree
that a simplistic
line-of-sight propagation model isn’t ideal. But even if this thing
were coming in at
-40dBm, that’s still plenty-strong to screw me up!

On Mon, Sep 11, 2006 at 11:30:47PM -0500, Rick P. wrote:

giggle

Marcus … that’s an aircraft transponder. Except for some vintage
1930’s aircraft that don’t have an electrical system - every plane /
helicopter has one. Hit Google or Yahoo! for “Mode-C” and “Mode-S”
transponder.

The ground radar “pings” the aircraft’s transponder (interrogation). The
transponder sends back a reply. Some replies carry a 12 bit ID code plus
altitude. Others might have actual GPS coordinates.

The transponders are 1090/1030 mhz and not 1350.   1350 is just

radar.

Also FWIW, TACAN and JTIDS systems operate between 960 and 1240

mhz with pulsed emissions as well as radars.

But Marcus's description of signals every 5 seconds sounds exactly

like radar and I bet if he has ID’d the right radar as the culprit he
can go visit it and watch it turning around at exactly the interburst
interval he sees. That might in fact be a good test if he cannot find
a
better method of verifying that is his radar as various actual radars
tend
to use different antenna RPMs.

I might also add that unless propagation is strictly line of sight

one needs more sophisticated path models to determine expected signal
power.
-11 dbm sounds rather loud.


Dave Emery N1PRE, [email protected] DIE Consulting, Weston,
Mass 02493
"An empty zombie mind with a forlorn barely readable weatherbeaten
‘For Rent’ sign still vainly flapping outside on the weed encrusted pole

  • in
    celebration of what could have been, but wasn’t and is not to be now
    either."

Marcus L. wrote:

I looked up that frequency range in Industry Canadas frequency
allocation search tool, and found that NAV CANADA
had several radars in Ontario in that frequency range, including two at
the Ottawa airport–one right at 1350Mhz.
But, Ottawa is 60Km away from me, I would have thought the signal would
be too weak for me to see.

Hello Marcus!

Could you make a recording of this signal?
AFAIK radars usually use Chirp signals, they correlate good.

Patrick

Engineers motto: cheap, good, fast: choose any two
Patrick S.
Student of Telematik, Techn. University Graz, Austria

  I might also add that unless propagation is strictly line of sight

one needs more sophisticated path models to determine expected signal power.
-11 dbm sounds rather loud.

If your current path loss calculator doesn’t accomodate for terrain,
you might give SPLAT! (http://www.qsl.net/kd2bd/splat.html) a try! You
can download Digital Elevation Map (DEM) data from NASA for your
region of interest to account for the effects of terrain. The current
version doesn’t account for things like antenna gains, but you can do
it manually. The new version will have antenna gain compensation and
Fresnel zone calculations, among other things.

Good luck!

Adam

On Tue, Sep 12, 2006 at 08:01:51AM -0400, Marcus L. wrote:

station, and
they’re drifting in and out of phase with respect to one another. The
sub pulses
are weaker than the main pulses by quite a bit.

I very much doubt that the "other radar" is responsible for the

secondary pulses you see unless they arrive at a distinctly different
time in the 5 second (I think you said earlier) rotation cycle of the
radar. I presume from your description that each main pulse is either
proceeded or followed (by some short (us) interval that is unclear from
what you have written) by a weaker secondary pulse.

Almost certainly two independent radars would operate at

slightly different PRF’s (depending on how old they are that might be by
a percent or two or for more modern gear by the difference in typical
crystal oscillators in the two time bases, perhaps 10-100 ppm in usual
cases). This would result in a steady phase change between the pulses
each pulse - if you can measure it you’d presumably find it basically
linear in microseconds per pulse.

But because their antenna rotations are unlikely to be

accurately synchronized and the chance of their beams both pointing at
you at the same time slim it would be unlikely to find the pulses of one
present at the same time in the five second antenna rotation cycle as
the other at least for any extended period. And very unlikely that the
relative strength of the strong pulse and weak pulse would remain the
same as the two antennas swept past you from different distances.
Likely one would peak before the other and this could obviously result
in one being stronger for a while and then the other becoming the strong
one.

It does sound to me like one of two other explanations is at

work here. Either you are seeing actual echos (reflections off aircraft
or ground based scatterers) which would of course always follow the
main pulse by some time interval that would depend on the additional
path length. Or the other explanation is that the radar deliberately
radiates a secondary pulse before or after the primary one (sometimes
from a different feed on the antenna).

Actual real live echoes from aircraft strike me as quite

possible to pretty probable at least unless you have a direct line of
sight to the radar and can actually see it in the distance and therefore
see such walloping huge signal from it that any echoes are too weak by
comparison to be visible within the dynamic range of your gear.

Likely the "direct" path involves Fresnel scattering and lots of

attenuation by foliage and the like and is many tens of db more lossy
than the free space path loss equation would suggest. And likely
reflections from passing planes and perhaps even certain fixed mutually
visible scatterers like towers or mountains involve direct unobstructed
paths from the radar to the target (airplane) and from the target
(airplane) to you. These would have much less attenuation by close to
the ground propagation effects than the nominally direct path, and
while a reflective target often scatters energy more or less in all
directions (thus the fourth power law of returned signal in the classic
radar equation) this energy can still be considerable relative to a
trans horizon direct path if little or no energy is lost by close to the
ground absorbers and scatterers in the reflected path.

And if you see the delay between main pulse and secondary moving

with time, well flying airplanes do move with time too…

What you have of course is a simple bistatic radar system.

The other possibility would be more likely if the delay between

the runt pulse and the main one was constant. Some radar systems,
especially those intended to operate with active transponders on the
target, transmit a secondary pulse using an antenna with a different
pattern. Typically these secondary antennas have a broader beam
pattern than the primary beam so the secondary pulse is stronger
relative to the primary pulse except when the primary antenna is pointed
right at the target. Used with a transponder that can distinguish
between primary and secondary pulses this allows the transponder to only
reply when it sees the primary pulse stronger than the secondary pulse
indicating the antenna is pointed right at it. This of course is
important if the transponder (eg aircraft) is near the antenna (with low
free space path loss) as it may see enough signal from antenna sidelobes
and reflections to reply when the antenna is at azimuths pointing well
away from it. And obviously this appears to the radar to be a target at
another azimuth.

The air traffic control transponders in use today on 1030/1090

mhz work this way with a secondary pulse from a secondary antenna used
to ensure the transponder only replies to the main lobe of the antenna.

At 1350, the pulses are arriving 35-40dB out of the noise. But I agree
that a simplistic
line-of-sight propagation model isn’t ideal. But even if this thing
were coming in at
-40dBm, that’s still plenty-strong to screw me up!

No question.

Of course you could track the rotation of the radar and blank

your receiver during the time it sweeps past you (this of course might
not work for reflected signals from passing aircraft very well as they
could appear at other azimuths and times in the radar rotation cycle
depending on where the airplane is).


Dave Emery N1PRE, [email protected] DIE Consulting, Weston,
Mass 02493
"An empty zombie mind with a forlorn barely readable weatherbeaten
‘For Rent’ sign still vainly flapping outside on the weed encrusted pole

  • in
    celebration of what could have been, but wasn’t and is not to be now
    either."

Patrick S. wrote:

Hello Marcus!

Could you make a recording of this signal?
AFAIK radars usually use Chirp signals, they correlate good.

Patrick
I can make a recording of the 20Khz or so of post-detector signal.

David I. Emery wrote:

a percent or two or for more modern gear by the difference in typical
crystal oscillators in the two time bases, perhaps 10-100 ppm in usual
cases). This would result in a steady phase change between the pulses
each pulse - if you can measure it you’d presumably find it basically
linear in microseconds per pulse.

The timing difference between the “main” and “sub pulse” varies between
a fraction of
a second, and about 1.5 seconds.

one.
possible to pretty probable at least unless you have a direct line of
sight to the radar and can actually see it in the distance and therefore
see such walloping huge signal from it that any echoes are too weak by
comparison to be visible within the dynamic range of your gear.

While I’m probably seeing direct and reflected pulses, the inter-pulse
timings that
I’m seeing preclude aircraft echoes as the source of the sub-pulse.
Unless they’re
flying somewhere around the orbit of the moon :wink:

directions (thus the fourth power law of returned signal in the classic
radar equation) this energy can still be considerable relative to a
trans horizon direct path if little or no energy is lost by close to the
ground absorbers and scatterers in the reflected path.

And if you see the delay between main pulse and secondary moving
with time, well flying airplanes do move with time too…

See above. I haven’t done “deep inspection” on the pulses to look for
microsecond-scale
differences in timing. The main and “runt” pulses have timing
differences varying
on the order of a second or so.

right at the target. Used with a transponder that can distinguish
mhz work this way with a secondary pulse from a secondary antenna used
to ensure the transponder only replies to the main lobe of the antenna.

Very interesting information. Information that I’m sure someone on
this
list could use to make yet-another interesting application for Gnu
Radio!
That’s what makes the synergy of projects like Gnu Radio so much fun.
Open sharing of information can lead to entirely new avenues.

For now, I have a request in to NAV Canada to get more technical
information
about their 1350Mhz radar in Ottawa. In fact, I used to know a ham
radio guy
who worked their back in the 1980s–if he still does, I might be able
to get some
help getting rid of this thing.

I plan to acquire a narrower filter than I’m currently using. I
estimate that the
new filter has about 55dB of rejection at 1350. That should help
some, although the
sideband components from such a radar are also something to think
about.

I might also build an impulse-noise-blanker block for Gnu Radio,
specifically
for dealing with this type of thing. Unless someone has already done
this :slight_smile:

Another odd feature of this pulse noise is that it prefers to appear at
night, although
it also appears during the day sometimes, too. My theory is that
during the day,
my sidelobe noise is dominated by the Sun, with the radar pulses not
being able
to compete. Whereas at night, there’s no Sun noise for the radar
pulses to
compete with.

My other theory is that perhaps the ERP of the radar is pumped up at
night for some
reason, or simply that propagation is better at night.

Marcus L. Mail: Dept 1A12, M/S: 04352P16
Security Standards Advisor Phone: (ESN) 393-9145 +1 613 763 9145
Strategic Standards
Nortel Networks [email protected]

David I. Emery wrote:

duration typical pulsed radar pulses.

And now it is obvious to me you were referring to the inter
burst interval and not the inter pulse interval.

Yes. Not understanding that these “pulses” have underlying structure
would lead
me to simply call them “pulses” (which, from the point of view of
someone
who expects events to unfold over periods of minutes, they are :slight_smile: ).

and the radar but capable of reflecting a significant amount of energy
by you. And yes, moving aircraft would result in the time interval
or other reflectors not directly between you and the radar is to

They tend to filter them pretty well in the radar in order to
keep them from causing problems out of band.

Any idea what the regulations say about this? For example, if
out-of-band stuff is
only 60dB down 100Mhz away, there’s still something like +5 to +10dBm
leaving the radar antenna. Although, given a minimum of 130dB of
path-loss
between it and me, the “direct” signal seems decreasingly like a
culprit, but rather
some mixing product between the relatively-strong main signal at
1350Mhz and
something else, causing it to appear in my IF passband.

Sun noise is pretty damn weak compared to your estimate of -40
dbm for the radar. But obviously such a phenomenon as you describe
(sun noise in the sidelobes) should show up as huge diurnal variation in
background directly correlated with the movement of the sun across the
sky. This should be pretty obvious from your data independent of the
radars.

I’m easily able to detect sources down to 1000Jy (where 1Jy =
1.0e-26W/M**2/Hz), and
the Sun crossing the main beam creates a response equal to 140000Jy.
There is diurnal
variation in the baseline–caused both by the Sun, and the galactic
plane. Even when
the radar pulses are a problem, I can detect, for example, Cygnus A,
which peaks at
about 1500Jy.

My main concern is that the strong radar pulses will disturb the
baseline so much that
there’s no way I can reach my ultimate system sensitivity–which
should be roughly
20-30Jy.

Propagation of radar energy trans horizon undergoes lots of
weather and time related changes due to tropospheric ducting and bending
phenomena - which can cause spurious “angel” echoes on weather radars…

And indeed this is sometimes worse at night.

Indeed, on another forum, somebody else noted that sometimes L-band
radar
propagates better at night, which might explain the
statistical-preference for
this stuff showing up at night. It has shown up during the day, but
only
about 1/3rd as often.


Marcus L. Mail: Dept 1A12, M/S: 04352P16
Security Standards Advisor Phone: (ESN) 393-9145 +1 613 763 9145
Strategic Standards
Nortel Networks [email protected]

On Tuesday 12 September 2006 14:11, Marcus L. wrote:

Patrick S. wrote:

Hello Marcus!

Could you make a recording of this signal?
AFAIK radars usually use Chirp signals, they correlate good.

Patrick

I can make a recording of the 20Khz or so of post-detector signal.

Hmm, sounds like a candidate for pulsar detection testing. After all,
the
pulsar looks just like a radar, with a rotating continuous beam…

Lamar Owen
Director of Information Technology
Pisgah Astronomical Research Institute
1 PARI Drive
Rosman, NC 28772
(828)862-5554
www.pari.edu

On Wed, Sep 13, 2006 at 10:05:56AM -0400, Marcus L. wrote:

The timing difference between the “main” and “sub pulse” varies between
a fraction of
a second, and about 1.5 seconds.

OK, you confused me (awfully easy to do these days as I sink

into senility) because I think of radar related events on a microsecond
and not a second scale. Obviously more or less the only thing happening
on a second scale is antenna rotation. And I believe when you speak of
“pulses” you mean BURSTS of pulses. PRF of radars tends to be in
100-500 hz area typically in this kind of application so almost
certainly what you are seeing is clusters or bursts of tens or more
probably hundreds of pulses - not individual couple of microsecond or so
duration typical pulsed radar pulses.

And now it is obvious to me you were referring to the inter

burst interval and not the inter pulse interval.

All of which leads me to comment that either the two radars you

think might be responsible are remarkably well synchronized in rotation
speed (which is possible I guess but I tend to think unlikely), or your
observations are not over a long enough period to see them drift more
randomly out of phase, or you are seeing something in the antenna
pattern of one SINGLE radar that results in multiple lobes pointing
toward you. This could be features in its antenna pattern (such a
broad secondary beam and a narrow primary beam) or due to reflections
off some distant but highly reflective object not exactly between you
and the radar but capable of reflecting a significant amount of energy
your way when illuminated by the radar at peak intensity.

Remember that the reflections off aircraft I mentioned in my

last comments would be expected to show up not only at the azimuth of
the radar when the radar is pointed right at you but when the radar is
pointed in other directions - eg at the aircraft, illuminating that
particular aircraft well with its beam. This would imply that aircraft
echoes would appear on other azimuths as seen by you than the main beam
of the radar and more importantly at DIFFERENT times in the seconds
timescale rotation of the radar than when the main lobe sweeps directly
by you. And yes, moving aircraft would result in the time interval
between the peak of the reflected energy from the aircraft and the main
lobe of beam sweeping past you changing over seconds or minutes.

But because they are reflections from the beam of that one radar

they would always relate in timing to the rotation of that antenna and
the timing of the main direct path lobe as observed by you. This is
not of course true of a second radar.

One possible way to distinguish between energy from a second

radar on approximately the same frequency and reflections off aircraft
or other reflectors not directly between you and the radar is to
determine if the PRF of the secondary burst energy is exactly the same
as the PRF of the primary lobe as it sweeps by you. There is a pretty
good chance that the “other” radar would not use exactly the same PRF
(no obvious reason to) and if it didn’t and that is what you were
seeing you’d see two distinctly different “buzz” frequencies which ought
to be obvious with a suitable measurement (eg a FFT of the low pass
filtered detected video).

I plan to acquire a narrower filter than I’m currently using. I
estimate that the
new filter has about 55dB of rejection at 1350. That should help
some, although the
sideband components from such a radar are also something to think about.

They tend to filter them pretty well in the radar in order to

keep them from causing problems out of band.

Another odd feature of this pulse noise is that it prefers to appear at
night, although
it also appears during the day sometimes, too. My theory is that
during the day,
my sidelobe noise is dominated by the Sun, with the radar pulses not
being able
to compete. Whereas at night, there’s no Sun noise for the radar
pulses to
compete with.

Sun noise is pretty damn weak compared to your estimate of -40

dbm for the radar. But obviously such a phenomenon as you describe
(sun noise in the sidelobes) should show up as huge diurnal variation in
background directly correlated with the movement of the sun across the
sky. This should be pretty obvious from your data independent of the
radars.

My other theory is that perhaps the ERP of the radar is pumped up at
night for some
reason, or simply that propagation is better at night.

Propagation of radar energy trans horizon undergoes lots of

weather and time related changes due to tropospheric ducting and bending
phenomena - which can cause spurious “angel” echoes on weather radars…

And indeed this is sometimes worse at night.


Dave Emery N1PRE, [email protected] DIE Consulting, Weston,
Mass 02493
"An empty zombie mind with a forlorn barely readable weatherbeaten
‘For Rent’ sign still vainly flapping outside on the weed encrusted pole

  • in
    celebration of what could have been, but wasn’t and is not to be now
    either."

Lamar Owen wrote

Hmm, sounds like a candidate for pulsar detection testing. After all, the
pulsar looks just like a radar, with a rotating continuous beam…

The pulsar receiver software was how I found this thing in the first
place…