Active Radar Hardware

My active radar code has been working for some time, but I’m just now
getting around to thinking about real-world tests. The difficulty I’m
facing at the moment is how to isolate the transmitter and receiver
while still providing decent power on target. I’m looking for
suggestions.

Right now, using the RFX2400, I’m thinking something like:

Tx/Rx port ----> AMP ----------->.
|
circulator <----> Antenna
|
Rx2 port <---- LNA <-----------.

where ---- is coax.

I found a nice LNA with SMA connectors from MiniCircuits (25 dB gain,
1.0 dB noise figure), but the problem is that all of the 2.4 GHz
circulators I have found offer only ~20 dB of isolation. So, even with
the “AMP” gain set to 0 dB, the Tx power into the circulator will be
+13dBm.* Any attenuation to this signal via the circulator en route to
the receiver will be undone by the LNA, and +13 dBm > 0 dBm, which is
the safe level for the receiver.** I cannot turn down the transmit
power if I hope to be able to operate at useful ranges.

I have zero experience building a system like this. I’ll be thinking a
lot more about it, and looking for better parts, but I’ve learned to ask
the experts early :wink: Any ideas?

Thanks,
-Lee

*http://lists.gnu.org/archive/html/discuss-gnuradio/2006-06/msg00238.html
**http://lists.gnu.org/archive/html/discuss-gnuradio/2006-09/msg00291.html

P.S. The parts I’m considering can be found here:

http://www.minicircuits.com/cgi-bin/inquery.cgi?querystring=ZQL-2700MLNW&searchtype=modelfamily&x=23&y=10
http://www.ditom.com/product_info.php?products_id=30&osCsid=9339bea3608c32e01f46033a16b8255f

Hi

I have zero experience building a system like this. I’ll be thinking a
lot more about it, and looking for better parts, but I’ve learned to ask
the experts early :wink: Any ideas?

I used to write software for a radar. It used separate antennas for
transmit
and receive for several reasons though the RX and TX antennas were
located
right next to each other. This arrangement works very well. If you are
concerned about too much power entering the RX stage you could use a
switchable attenuator using a voltage signal and PIN diode. Actually,
if you
want any sensible readings and a signal that has good dynamic range at
the
ADC you will need a controllable attenuator that increases the RX gain
over
time (in a logarithmic way) as the return signal gets weaker with the
squared
square of the distance (R^4 !). You’ll always get overload signal for a
certain close-in range anyway. Not much you can do about that.

Regards
Jason Hecker

On Friday 29 September 2006 07:30, Jason Hecker wrote:

lot more about it, and looking for better parts, but I’ve learned to ask
the experts early :wink: Any ideas?

I used to write software for a radar. It used separate antennas for

I still do :slight_smile:

transmit and receive for several reasons though the RX and TX antennas were
located right next to each other. This arrangement works very well. If
you are concerned about too much power entering the RX stage you could use
a switchable attenuator using a voltage signal and PIN diode. Actually, if

We make MF & VHF systems and for some of our VHF systems we use a single
set
of antennas and a T/R switch (passive and active). However since our
frequency of operation is 2-3 orders of magnitude lower it’s probably
all
different…

you want any sensible readings and a signal that has good dynamic range at
the ADC you will need a controllable attenuator that increases the RX gain
over time (in a logarithmic way) as the return signal gets weaker with the
squared square of the distance (R^4 !). You’ll always get overload signal
for a certain close-in range anyway. Not much you can do about that.

For boundary layer/troposphere radars we halve our effective PRF and
transmit
a short low power pulse and listen with a low gain setting on even
pulses and
a high power long pulse with high gain on odd pulses. The sampling
configuration also alternates to sample low and high.

I guess the other possibility is being adaptive, ie adjust your
power/gain
based on what you can currently see.

Jason and Daniel -

Thanks for the feedback. I’ll look into PIN diodes. The reason I
wanted to use a circulator was because it doesn’t require a switching
signal to be generated from the USRP. And, while generating such a
signal is possible, it is difficult (if not impossible?) to ensure it is
timed properly with the transmit waveform – unless I do some FPGA
coding, which may be the necessary next step. (Again, something with
which I have no experience! But learning new things is fun, right?) I
think I will look into using the actual Tx waveform as the switching
signal.

Also, Jason, thanks for pointing the need for a variable attenuator to
condition the signal for ADC. Although I’ve been aware of this in the
past, I forgot about it in my application.

Thanks again fellas,
-Lee

We make MF & VHF systems and for some of our VHF systems we use a single
set of antennas and a T/R switch (passive and active). However since our
frequency of operation is 2-3 orders of magnitude lower it’s probably all
different…

The radar I worked on was a ~2.7GHz job. Since the receiver was spammed
for
the first several hundred metres in range due to large amounts of echo
from
ground clutter any such early returns were gated out. After that the
circuit
controlling the attenuator (PIN diodes) ramped the voltage so that the
overall gain of the system increased in time. This was to maximise or
normalise the S/N ratio at the ADC. When you are looking for metallic
bogie’s many 10’s of KM away you need all the return you can get even
with
the gain achieved from integrating multiple returns.

Anyway, the author of the grandfather post doesn’t sound like he has too
much
RF experience (who does?) Perhaps a circulator will do or even better,
an
RX/TX switch module (diplexers?) - though this might involve a hardware
hack
to get the switching pulse out in time. The thought did occur to me to
get
something like a Furuno boating radar head with an integrated separate
transmit and receive antenna. You’d get a nice narrow beam from one of
those.

On Saturday 30 September 2006 00:04, John Ackermann N8UR wrote:

But also realize that in addition to preventing physical damage to the
receiver, you also need to deal with “desense” (as we old repeater
builders call it) that results when the RX front end is subjected to a
strong input. It takes a finite amount of time for the front end
operating conditions to return to normal after being exposed to a big
signal, so that may impact your turn-around time.

If the Rx boards support it you could turn off the LO during
transmition, that
can improve recovery time.

(Although this is all second hand knowledge at VHF :slight_smile:

Lee, I’m not sure what power levels you’re running, but if you need more
isolation than a single circulator can provide, you can always stack
'em. Then, you could also put a pair of clamping diodes at the receiver
input to protect it from any system failure.

But also realize that in addition to preventing physical damage to the
receiver, you also need to deal with “desense” (as we old repeater
builders call it) that results when the RX front end is subjected to a
strong input. It takes a finite amount of time for the front end
operating conditions to return to normal after being exposed to a big
signal, so that may impact your turn-around time.

Remember that I’m in the neighborhood, and between me and my local ham
gang there’s not much RF test equipment we don’t have. We’ll be happy
to help out. In fact, we’d love to see a demo of your radar at one of
our meetings… :slight_smile:

John

This forum is not affiliated to the Ruby language, Ruby on Rails framework, nor any Ruby applications discussed here.

| Privacy Policy | Terms of Service | Remote Ruby Jobs