radar guns not fast enough [Archive]

XJet

Nov 12, 2007, 03:20 AM

Radar guns work by comparing the frequency of the reflected signal with that of the signal that was transmitted.

Due to the phenomenon known as doppler shift, the frequency reflected form an object moving towards the gun will have a higher frequency than that which was transmitted, and the frequency reflected from one moving away will have a lower frequency.

It is then quite trivial to convert that frequency into a speed (I can provide the math but it's not really essential to this post).

The best example of this doppler shift is the way the sound of a police, fire or ambulance siren (or even the sound of a model engine) seems to increase in pitch as it approaches and decrease in pitch as it moves away.

Anyway, there's no absolute reason (other than a decision made by the writer of the software) why you can't read speeds higher than the arbitrary maximum most guns provide -- it's just a case of dealing with a slightly higher frequency difference between the sent and reflected signals.

Where the big problem arises with low-powered hand-held guns is that comparatively small and very fast moving objects only return a sufficiently strong reflection for a very short period of time.

This means that the gun may not receive a stable reflected signal for long enough to establish a reliable or accurate figure.

Increasing the power of the gun would help, as would increasing the sensitivity of the receiver used.

However, doing either becomes an expensive option, once you start moving away from the relatively standard microwave components that these guns tend to use.

Another option is to increase the gain of the antenna (horn) used -- which has the same effect as upping the power *and* increasing the sensitivity.

But (wouldn't you know it) this also has a downside in that it reduces the illuminated area over which a reading can be taken (ie: it focuses the radar beam into a much smaller area) and makes the gun significantly larger.

By far and away the best way to increase the effectiveness of these simple radar guns is to increase the radar profile of the object you're measuing the spead of.

Relatively small triangular/pyramid radar reflectors are fairly simple devices but many still won't provide a reliable reflection (they often tend to "glint" the signal as a mirror does in sunlight -- ie: you get flashes of strong signal and periods of no reflection at all depending on orientation). With a fast-moving model constantly changing orientation in respect to the radar beam, reflectors become a bit of a hit-and-miss affair.

Perhaps the best way to achieve a higher radar profile/signature is to use a strip of metal tape (aluminum or copper) along the leading edge of the wing. This will act as a good reflector and (because of the convex profile) ensure that there's always some degree of signal being reflected.

But then you have the problem of the inbuilt maximum speed limitations provided by the software in the gun.

This maximum speed is probably set by a desire to minimize the gun's sensitivity to reflections and signals that fall outside the normal range of operation. By filtering out all the frequencies that are above or below the intended operational range, the receiver can be made to appear more sensitive -- thus improving its performance wthin the designed range of speeds.

This can be gotten-round by designing a gun that works over a speed range that doesn't start at zero. For example, you could design a gun that works just as well measuring speeds from 200mph-400mph as does a gun that works from 1mph-200mph.

I do some of this stuff for a living (one of my current projects is a proximity warning and collision-avoidance radar system for UAVs). This system actually uses much of the same technology and componentry as is found in radar guns.

Fortunately full-sized aircraft have a very strong radar signature which allows a great deal of range to be achieved with relatively low powers. Clocking models is actually a whole lot harder.

Acetronics

Nov 12, 2007, 07:22 AM

Hi, Ralph

The problem is quite simple : radar ranging need some surfaces to reflect micro-waves ... and a speed glider MUST NOT have large surfaces that could reflect those waves ...

just make your own conclusions with that !!!

Alain

village_idiot

Nov 12, 2007, 09:24 AM

XJet

Nov 12, 2007, 01:30 PM

What about a big reflector in the nose, cover it with fiberglass.

What if it didn't work on doppler, but instead compared the time of return of a first pulse to the time of return from a second pulse. More like an absolute ranging device. Are there hand held devices that work in this way?

The problem with that is that there may be many reflections returning to the radar gun and some may be larger than that from the model.

Remember that, in order to get an accurate reading, the radar ought to be as close to parallel as possible to the path of the object -- which means that there's a good chance "ground clutter" will get in the way in many cases. The ground is going to provide multiple reflections so the pulse that is sent out gets returned many times -- once from each object it encounters.

Working out what's static and what' smoving becomes non-trivial and prone to error.

That's why doppler shift is used for speed radars -- objects that don't move don't generate any "difference" frequency so are automatically ignored.

As I said before, if you want to increase the radar profile of your planes, put some metal foil/tape along the leading edge of the wing.

village_idiot

Nov 12, 2007, 01:47 PM

Ok, I hadn't thought about all the other sources of reflection.

jkettu

Nov 12, 2007, 02:54 PM

Does it have to be groundspeed? If not, you could measure the airspeed with a pitot-static system. Even if there is no room in the plane for a telemetry transmitter, you could just log the data and downloading it after flight.

Brandano

Nov 12, 2007, 05:47 PM

the radar gun should be able to differentiate between reflections from static objects nd moving objects. Essentially, it's the way it can detect the speed, by measuring the doppler shift of the radar signal. A good reflector doesn't need to be large, but has to be engineered to send the signal back to the source, the same way refractive tape works. one way to achieve this is to make a reflector that essentially is 3 reflective surfaces arranged to that they intersecate at 90 degrees to each other. Whatever direction you fire a ray at it it will be shot right back, like throwing a rubber ball in a corner of a room. you could arrange 3 disks of cardboard covered on both sides in foil and intersecating each other, embedded in any place whithin the wing or fuse, behind fiberglass. Carbon fiber and other conductive materials will shield any reflector, so that might be quite an issue.

Ralph Walton

Nov 12, 2007, 05:49 PM

Airspeed is not usefull for record attempts and is not in the scope of this dissussion. the are many other possible ways of recording speed including high speed camers, gps grid arrays placed on the ground etc etc.They have their own threads.

This discussion is about what can be done to improve radar guns.

Hi Xjet. you seem to have a good grasp of whats required, and I'm shure there are alot of guys out there, who would happily pay (a reasonable fee) to have their guns modified to read from 100 to 400 MPH instead of the std 0 to 300. If I buy a Pro Stalker or Prospeed gun are you reasonably confident you could do the mods, or do you know someone who could do them? Would one gun be prefered over another?

Now the tricky question. Do you have a ballpark cost of the mods? Are we talking hundreds or thousands?

Ralph Walton

Nov 12, 2007, 05:59 PM

I agree Brandano and others who have mentioned reflectors. A suitable reflector on the plane is going to help alot. It may be tricky, or as simple as what a fellow Kiwi discovered recently.

"We've found a bit of aluminium tape chordwise around the wings does wonders for the radar signature... Alex's Destiny now shows up more or less every lap at Sugarloaf, where it was maybe one in eight before the tape."[QUOTE]

Daemon

Nov 12, 2007, 07:27 PM

For whatever it's worth, the current record holding plane (100" Kinetic
at 309mph) was built with large flat strips of aluminum embedded in
the layup of the wing, flat side facing forward. It was built specifically to
catch radar, and yet is still one of the hardest planes in this class to pick up.
We started getting no-reads in the low 200's (where we have no problem
with even some smaller planes) and got sporadic reads
as it approached 300, and only a couple over 300.

A bigger plane (126" Thundertaker) was much easier to read all through
the mid 200's and then pretty much disappeared over 300.
Got one reading at 306, another at 303, and as it went faster, no more.

Corner reflectors have been discussed at great length in the DS forum.
The problem with the corner reflectors is that we just don't have any
place to up em. The nose of these gliders are usually filled with lead,
by necessity and taper to a very small diameter, and the wings are simply
too thin. Strips of metal taped to the outside of the wing (chaf) has
been tried but at nowhere near the speeds we're talking.

Honestly though what's driving this is not the sporadic performance
in the mid-high 200's. The Stalker Pro is pretty solid there. What's
got everyone working on this problem hard now is that we've simply maxed
out what the guns are rated for, let along capable of. The Staker Pro is
rated to 300mph and reads to about 305. The Pro Speed (which is
a less reliable gun in the 200's than the Stalker Pro), has a spec sheet
that says it's good to 310, and some manuals say it'll work to 320
but I don't believe it. Besides there only seems to be *one* single Pro
Speed gun that can pick up planes at this speed and the rest are worthless.
I think the current crop of record planes are capable of going faster than
320mph anyway, so it's time to find a way to modify the existing guns, or
replace them entirely.

The manufacturers of the existing guns obviously want to know how big
the market is for this sort of thing before they will commit to any
modifications or new products, and of course it's very small. Unless
they happen to have their own products in the pipeline that will
fullfill our needs I don't see much help coming from the manufactures.

ian

nauga

Nov 12, 2007, 08:37 PM

The manufacturers of the existing guns obviously want to know how big the market is for this sort of thing before they will commit to any modifications or new products, and of course it's very small. Unless
they happen to have their own products in the pipeline that will
fullfill our needs I don't see much help coming from the manufactures. Raytheon, TI, and Northrop-Grumman all have radars more than capable of processing DS returns ;)
What are the manufacturers' responses when approached? It's not uncommon to see stuff used in competitions produced in small lots for either promotional consideration (i.e. a sponsorship, take lots of pictures) or just industry goodwill. Also sounds like an ideal project for an EE master's thesis with some support from a manufacturer. No one is close to exceeding radar limits, only unit limits. Maybe you could make it attractive to a gun maker, either from an intellectual property standpoint, an advertising standpoint, or a goodwill standpoint.

Nauga,
who doesn't like inventing wheels

XJet

Nov 12, 2007, 09:51 PM

Just how much spare room do you have inside your gliders?

A GPS-based solution would be another option and it's now possible to get micro-sized GPS receivers that are only about 12mm (1/2") square. These could be mated with a suitably designed microcontroller to log the speed, altitude and heading data for display later once the glider had landed.

Very accurate, probably possible in a box about 5/8"x5/8"x1/2" that would weigh about 1/2-3/4 oz and costing around $200 or so (with ground-based reader).

Is that going to be practical?

Another (more expensive) option ($$$) is laser.

The final (and perhaps viable) is to use a K-band radar rather than an X-band one.

The shorter wavelength of the K-band means that your model will automatically have a higher profile (more reflective and larger as a ratio of the wavelength) -- but unfortunately this is counterbalanced (to a degree) by the higher attenuation of K-band signals through the air. But (in this wonderful wold of swings and roundabouts) the free-space losses can be compensated for by the use of a more efficient antenna.

I'll crunch some numbers and see if it's worthwhile going to a higher-frequency radar.

Kiwi Jono

Nov 12, 2007, 10:54 PM

GPS modules need to be looked at further (would fit in most gliders) but I think there may be an issue with update rate?
Have a look at some of the DS videos - these gliders "flick" around a tight circuit (40+ Gs) and rarely fly in a straight line for any significant time. Also the position of each circuit varies quite a bit (depending on how the boundary layer is shifting, how much turbulence there is and the skill of the pilot). This also creates problems with hand held laser (but maybe a scanning system might pick it up?) because it is very difficult to keep laser pointed at the glider all the time (hard enough with a radar).

XJet

Nov 12, 2007, 11:38 PM

The standard update rate for civilian GPS is 1Hz (once per second) but that doesn't affect the accuracy of its speed reading (although it will be the average over that second).

In our UAVs we use GPS systems that operate at 5Hz which would provide a better indication of maximum speeds if those speeds were unly sustained for a very short period of time (0.2 seconds or longer). Unfortunately these faster GPS units are a little larger -- we use ones that are about 25mm (1" square).

nauga

Nov 12, 2007, 11:52 PM

In our UAVs we use GPS systems that operate at 5Hz which would provide a better indication of maximum speeds if those speeds were unly sustained for a very short period of time (0.2 seconds or longer). What acceleration (steady state and vibation) are these units spec'd to? The ones I'm familiar with are spec'd pretty far below what's required for a high-speed DS lap.

Nauga,
smashed flat

Ralph Walton

Nov 12, 2007, 11:53 PM

Breaking news. I recieved this reply on the radar issue this afternoon it ties in with this thread seamlessly. I have also posted the relevent part in the DS forum.

Your email sounds interesting. I haven't heard of this sort of speed record being attempted before, but I can certainly understand the aerodymanics of it.
I fly gliders, hang gliders, paragliders and even on the odd occasion, some RC stuff.

These days I'd sooner be in the air flying above the RC gliders at Otford than on the ground looking up at them

Flying on the downwind side of the hill.... oh yeah.... that'll increase the absolute speed realitve to the ground. The air is rough as guts back their will nasty rotor and lots of air rushing downwards. With an RC glider, the worst that can happen is it goes splat on the ground. When ya fly a hang glider back their, the thought of going splat on the ground haunts you for a long time afterwards.

It's not normally a problem that anybody max's out a radar, it's amusing. There is actually a very simple solution to your radar problem. The circuit is called a pre-scaler. They actually make pre-scalers for some very fast circuits that run to >10GHz and come on a single chip. But in your case, the dopler shift for 300mph is still only a very low frequency as far as electronics are concerned and it can also be done with a single chip and a few other cheap components.
The circuit your looking for is a "divide by 2". That's exactly what it does. For every two pulses it has on it's input, it outputs one pulse. This just wires in between the radar mixer/detector circuit and the counter input. It will give you a radar reading of half the true speed.
The limitation isn't the radar (microwave frequency) its the CPU just not programed to count higher dopler frequencies.
So all you have to do is put a switch in to read true or half. Flick the switch and the radar will read up too 600mph, but will display 0-300 representing double the speed.
If the display reads 200, then the true speed is 400.

The circuit is very easy to build and radar guns are easy to modify. The biggest problem will be finding the space inside the chassis. If you have circuit diagrams or some pictures it might help.
If you notice any on ebay, point one out and I'll try and buy it. Modify it and then if anybody wants it, I know I won't have a problem selling it

There is however a better way. You can buy GPS modules very cheaply and dataloggers.
They can talk to the GPS modules and log their position and speed in 3D space.
When you recover the plane, just download the data and you have a winner.
This is great because it can give you a 3D profile of the flight and is excellent for claiming records. It's very hard to forge GPS logs. It's not impossible, but I thought I'd attempt to forge a GPS log just to see if it could be done and then used against speeding fines etc.
What a tangled web we weave when we first try to deceive.
A forged GPS log looks like the difference between a CGI cartoon character and a real actor.
It's easy to spot the difference.

Though they aren't suitable for your application, the software and hardware is ideal for your needs. Hang gliders use a Vario (vertical airspeed indicatior) You might have heard them beeping away if hang gliders are flying nearby. They connect them to their GPS's or they now have built in GPS's and at the end of the day, you connect them to your computer and they play back a 3D video of your flight, in some cases you can upload the 3D terrain maps etc. It's very cool software.

Suunto have watches. They would fit into a model glider easy. The current models aren't up to the stage that the varios are, but it won't be long before their GPS watches combine their altitude data and do the same thing, if they can't already.
Nasty price though... $500 is a cheap watch.

I'll find the files I have from my suunto watch... it is only 1 dimensional.. but it's interesting to look at.

Ash

Daemon

Nov 13, 2007, 12:40 AM

What acceleration (steady state and vibation) are these units spec'd to? The ones I'm familiar with are spec'd pretty far below what's required for a high-speed DS lap.

We have discussed GPS options fairly extensively in the DS forum.
I'll be evaluating the Etek EB-85A 5Hz GPS unit tied to a Dragon OSD (On
screen display) and a 2.4Ghz video downlink for realtime display of GPS
telemetry. We don't know what any of this can handle in terms of G-loading
or whether the velocity averaging will puke on the fact that the plane is
always turning. Or whether the GPS receiver will play nice with 100%
carbon planes. Also all GPS units I know of, only show speed over
ground (horizontal plane) and we fly a tilted circuit, so only top and
bottom ends of the circuit are parallel with horizon.

The OP of this thread wants to keep it specific to radar though.
If a pre-scaler will do the job, then I'm sure there'll be plenty of
volunteers ready to test it. One thing I'm concerned about is that
these guns send out a coded pulse so that they can identify their own
signal from the background noise and other guns (we've run
3 to 5 guns at once and they all pick out their own returns). If
the return frequency is scaled down by a factor of two, I wonder
if it'll still recognize its own pulse.

ian

Andrew McGregor

Nov 13, 2007, 01:28 AM

Hi..

I'm the one who came up with the chordwise alloy tape idea, after reading a paper on the microwaves 101 site (IIRC) on radar reflections and stealth technology. In short, both foamies and carbon skinned composite models are nearly perfect stealth airplanes. And, interestingly, most of the radar reflection from a fullsize aluminium plane comes from a) the sharp edges on the side AWAY from the radar and b) discontinuities in the skin, particularly door openings.

So, the carbon skin will not conduct radar energy to the far edge of the surface; military stealth planes use the same material for precisely this reason. Simple solution... put some aluminium (copper or any other metal would work basically as well) on the skin. Obviously that would best be done while constructing a composite model, but self adhesive tape works and doesn't disturb the aerodynamics to any significant degree.

I suspect most of the radar reflection on some carbon DS models without tape will actually be coming from the radio antenna! I think that explains why Alex's Destiny was so hard to pick up, as he uses Spektrum and has his antennas inside the fuse.

So, it has helped our radar (not sure who actually owns the gun...) pick up the local collection of DS models quite a bit, especially at Sugarloaf where we suspect the radar's electronics suffer from interference from the TV and comms tower that's only a couple of hundred meters away.

Ralph Walton

Nov 13, 2007, 01:30 AM

Ian
I agree about the GPS info that Ash sent to me It was interesting that he threw it in,I believe soon we will have working onboard GPS or similar, but its not in the scope of this discussion.

I'm waiting on a reply from him to see if he's interested in modding a gun to read higher speeds, increasing its "power" so theres not so may missed passes, and joining in directly to this forum.

Ralph Walton

Nov 13, 2007, 01:40 AM

Hi Andrew, sounds like the tape was a great success. Has anyone else tried it?

Ash asked for photos of the inside of the common guns used. If anyone can post them here I can pass them along to him. Also circuit diagrams?

XJet

Nov 13, 2007, 01:44 AM

Ron, the biggest problem with using an audio prescaler will be that the gun is still facing real problems picking up enough reflected signal to provide a "fix" on the glider.

As for the GPS solution -- there are two ways that you can get a speed measurement.

The first is the "speed" figure returned by the GPS itself. This is determined by analysing the doppler shift in the signals received from the various satellites used so is pretty accurate but may undergo some degradation when the major vector is vertical.

The second is to do simple trig on the 3D coordinate stream the GPS reports. This allows the true point to point speed to be calculated with some accuracy. The only limitation of this is the comparitively low resolution of GPS in the Z-axis (altitude).

The EB85 that others have referred to is an "okay" GPS receiver but the earlier firmware revision does get a little confused and lags when subjected to very rapid changes in direction in any of the three planes. A later firmware update corrects this behavior though.

It is also relatively small and lightweight, especially the version with the integrated ceramic patch antenna.

Mike the Snake

Nov 13, 2007, 07:08 AM

This is exciting news! I knew there was something that could be done frequency/harmonics-wise.

The missed reading thing I bet is a voltage issue. As speeds increase, the time is less, and I'd think a more powerful signal would be the ticket.

With this "divide by 2" circuit, and increasing readings with more voltage (or just batteries hot off the charger) and reflector strips and our problem may be solved!

Ralph Walton

Nov 13, 2007, 08:19 AM

We need to keep our eyes on the prize, and work together toward a gun that can be pointed in the general direction, get readings 95% of laps, and read to 600 MPH.

That should keep you guys happy for a while. Hey, I'm happy about it and I don't even own a gun yet. Its high on my list of things to buy though!

Mike the Snake

Nov 13, 2007, 09:03 AM

Two separate issues exist. Higher speed reading capability, and catching those readings consistently.

The main issue to address is the speed capability.

The missed readings I think will be part of the game, as time and angles above 300 are so crucial, but I have seen both brands of radar guns (Stalker and Prospeed) just start giving blank readings pass after pass for no apparent reason.

It has been noted that people's guns start missing readings when the battery gets low.

My guess is that the strongest signal is required to catch the fast passes.

It may be that the extra voltage from guns "hot off the charger", with the extra flash charge voltage is making that difference between a gun that reads, and one that gives more non-readings.

Maybe voltage affects the width of the beam?

In any case, this is a huge deal if the "divide by 2" circuit works.

How quickly could this be explored?

Andrew McGregor

Nov 13, 2007, 04:12 PM

Hi Andrew, sounds like the tape was a great success. Has anyone else tried it?

It's been used on four or five models, but only with the one gun to my knowledge.

Ralph Walton

Nov 13, 2007, 06:19 PM

How quickly could this be explored?

I've asked Ash how he wants to handle this project. Divide by two conversions, missed pass issues, communincation, when, etc etc.

I will let you know ASAP.

Richard S

Nov 14, 2007, 12:27 PM

is it possible to make a small constant frequency emmitter that sits in the model?

Brandano

Nov 14, 2007, 03:30 PM

That's also another interesting approach. Add an emitter to the plane, then measure the doppler shift with the gun. Doesn't really need to be a radar signal either, a piezo emitter with a well known frequency and a microphone would work fine, once you filter out all the other spurious noise. Maybe an ultrasonic range finder like on older Kodak cameras could be hacked for the purpose? this sistem is already used to estimate the speed of prop driven planes when taped, btw.

Andrew McGregor

Nov 14, 2007, 04:16 PM

Actually, the Doppler software used for power planes doesn't even require that you know the frequency in advance, just that it's pretty stable... and it's very tolerant of background noise too. Just an alarm buzzer on an RC switch should do it.

indtech

Nov 14, 2007, 04:49 PM

Lost Plane Alarm?

Brandano

Nov 14, 2007, 05:21 PM

I regularly hear Piaggio P180 Avanti planes flying overhead (my workplace is near an airport where the Italian Air Force ones are based) and the doppler shift is quite impressive. 738 km/h maximum cruise speed on a prop driven plane... wow! Though probably most of the effect is due to the odd noise of the props behind the wing trailing edge.

Daemon

Nov 15, 2007, 04:06 AM

We've also discussed using Audio Doppler in DS forum. I've looked
at the system the ducted fan guys use, and they rely on the plane
flying a level path with fan turning contant rpms to set the frequency.
They've got a little piece of software that'll show the doppler shift
visually. Click on either side of the curve and it calculates speed automatically.
While DSing we almost never fly straight, and we're flying record speeds
in 60-80mph winds so need to put the microphone somewhere it can reliably
hear the plane. A very loud constant frequency tone would be best.
Lost model alarm is not loud enough. Can barely hear it when the plane
is not moving, let alone going 300+mph, turning circles over
700 feet in diameter every 5 seconds.

As for putting a constant frequency radio emitter in the plane.
That's also been discussed and some folks actually talked about
building it. Basically separating the guts of a radar gun so
the emitter is in the plane, and tje receiver on the ground. But it'd
need a good sync signal, because it'd still send a coded pulse
to identify it's own signal from the noise and Tx and Rx need to be in
sync to identify the pulse. They figured they could develop it for some
significant amount $10-20k or something, but there'd basically be only one
of them in existence. It'd solve the power problem because the signal
would be 4 times stronger, but I don't know if it'd solve the speed problem.

ian

Brandano

Nov 15, 2007, 05:43 AM

I still think that you can get almost the same effect as a radar gun using the emitter and receiver from an ultrasonic rangefinder. It can be extremely loud without affecting the onlookers, and while it's highly directional you'd normally want to measure the speed when the plane is coming head-on. You'd have to use just the transducers, but replace the circuitry to read the doppler shift of a known fixed frequency rather than timing the ping as the original does.
Ok, other system: 2 or more cameras, a powerful laptop, and software to stereoscopically track the flight path and integrate the speed. You'd need the software to be able to pinpoint the plane against the sky's background, and that's easy to do with most motion detector algorithms. Add a bright IR LED to it and you can even have a specific reference point on the airframe. It would still be cheaper than 20k, though bulkier than a radar gun

Ralph Walton

Nov 15, 2007, 05:53 AM

News from Ash.

< As well as asking him how he wants to handle his ideas, I sent him a link to youtube with some good DS action on it.
Heres his reply:>

Wow, I was thinking that video looked sped up.

I don't mind, your welcome to post any of my blurb anywhere. Everybody just calls me 'trash'. < Ashes' email is trash0@hotmail.com >

I'll dig up some good examples of divder circuits. There are lots of designs, most are very simple.

I'm away from home during the week and I try and log in remotely when I can.

Radar target physics are reasonably easy to understand. Problem with gliders is that their low profile makes them very steathy. The kind of designs that the military actually seek for stealth aircraft comes naturally to your aircraft.

Square shapes present the largest radar signature. A good radar target is a corner reflector.
Radiosondes carry them. they look like this http://images.google.com.au/images?svnum=10&um=1&hl=en&q=radar+reflectors&btnG=Search+Images
Very un-aerodynamic. I've build many balsawood gliders over the years, it would be possible to build something highly reflective inside the wing, but I'm assuming a lot of these planes might be out of the box.

The skin of the aircraft will also be reflecting a lot of the K band signal, and most of that will be away from the radar gun because of the shape of the aircraft. Weight is obviously an issue, so I guess that leaves it up to coating the leading edge of the wings and or the nose in foil. If the nose itself is not very reflective to radar, then building a corner reflector target into it would be a step.
I'd test the glider on a wire sled.
Tie fishing line between the goal posts on a football field and have somebody bungy the glider down the field with the radar at the end.

The other way to get a speed reading is to put a signal source on the glider and measure the dopler from it. But once again, it's adding weight to the glider. I can think of some very simple and light methods of doing it, but it means a lot more complexity on the ground to detect it.

One was to use a flashing LED with precise timing, and measure and time the pulses. But the precission required is rather tight. Lidar uses this same method, and I'm wonder how available they are on ebay? A quick search didn't find anything. Reflective paint would make that work well.

There isn't much you can do about the power issue of radar guns. They're already using reasonable power gunn diodes. Ah, I did just have an idea !

You basically can't change the gunn diode, but you can increase the gain of the antenna system.
It's a lot more clumbsy, but I think it will do the job nicely.

I can't remember what dish antennas are commonly available in NZ for satellite/payt tv, but they will increase the gain of your radar (the punch if you like).

http://home-and-garden.webshots.com/album/183973395WcCdiw?start=0

This is my page of misc technical goodies. The "wokked wabbit" is the kind of thing you'll be thinking of. An offset dish satellite antenna will provide you with a lot more gain, but you'll have to be a lot more precise with the direction you point the antenna.
Ideally you'd set up a tube to act as a gun sight so that you knew where to point the antenna.
You could hold the dish in your hand, or mount it on a speak tripod.

<question I sent to Ash>

I need to ask you what direction do you want these ideas of yours to go from here? I guess the first question is, would you be interested in converting guns for the DS community?

I could do it, and though I'm damn broke at the moment (thanks to the mrs) I'd prefer to teach others how to do it so they can spread the technique amoungst themselves and develop it further.

Modifying the radar guns should not be very difficult. It's just a matter of selecting the appropriate part of the circuit to insert the divider circuit.

< So there you go gentlemen, feel free to contact Trash direct. I have asked him if he can join RC groups and participate directly, but have not had an answer to that question. Interesting that he wants to pretty much do a "how to" on the divider circuits, and give advice on getting better "reception" (fewer missed passes) Watch this space!-Ralph>

Andrew McGregor

Nov 15, 2007, 06:29 AM

Ohhhh... yeah, high gain antennas might be a thing to try.

As it happens, I have a couple of commercial high gain parabolic wifi antennas lying around. I'd have to check if their grids will work in the radar band, but we might be onto something...

Mike the Snake

Nov 15, 2007, 07:31 AM

Exciting.

Reading higher is the main issue, so the divider circuit would be most important.

We DO get readings, I think many blank readings are over the guns max, and time and angle make it more hit and miss.

Increasing the gain sounds like a good idea as well.

Lidar guns are available but only go to 200 I think.

The vids I've seen of lidar also take precise aiming and don't give instantaneous readings.

indtech

Nov 15, 2007, 10:11 AM

Note that a radar gun only reads the speed component coming directly at the gun. The plane must be flying directly at (or away from) the gun.

http://www.radarguns.com/radar-and-cosine-effect.html

XJet

Nov 15, 2007, 10:12 PM

Using a fixed-frequency radio source on the plane is not going to provide anything like the accuracy you need.

The doppler shift at 10GHz is still only an audio frequency (few KHz or so) and the drift of *any* transmitter on 10GHz far exceeds this figure so your reading would be smaller than the margin of error caused by that drift.

I mentioned using higher-gain antennas - but this very rapidly becomes a law of diminishing returns and once you've gone to about 17dB on K or X-band, the antennas/dishes become wildly unwieldy and have a very narrow appature through which the plane can be flown and picked up.

Daemon

Nov 15, 2007, 11:27 PM

The doppler shift at 10GHz is still only an audio frequency (few KHz or so) and the drift of *any* transmitter on 10GHz far exceeds this figure so your reading would be smaller than the margin of error caused by that drift.

It's no different than what radar guns already do. Measuring the
frequency shift, and ensuring minimal drift is obviously
not the problem, as these guns are highly accurate, when they have
time to get a read. Picking your own signal out of the noise is the hard
part, and thus the need for highly accurate time sync.
That's not a problem when Tx and Rx are in the same case. Much harder
when Tx and Rx are separated (as are all Rf timing schemes).

For audio doppler (because the wave propagation speed is so much
lower) the doppler shift % is much much higher. For instance
at 300mph, the doppler shift of an audio frequency is 62.5% so
a 3000_Hz tone becomes a 4875 Hz tone. Doing an FFT, measuring
the shift visually and converting to an appropriate speed on the fly
is the tricky bit.

ian

trash0

Nov 16, 2007, 08:36 AM

Ok, I've dragged my lazy arse onto this thread for some serious radar talk.
I'll just pick up on this page rather than read back through the previous pages, else I'll never shut up.

Using a fixed frequency source is actually quite feasible. If you know the frequency of the source, then you know that the frequency is relative to the speed. This is all good and well for demonstrating dopler shift on a large scale. It's just not practical on a small RC scale.
A VHF transmitter though simple, isn't high enough in frequency to offer a good enough resolution of dopler shift. Higher frequency oscillators like a DRO are small and light, but they have a lousy phase noise. They drift all over the place. You require very high precission for your reference oscillator. So it really needs to be a crystal locked oscillator with a multiplier chain. I could build one into a the size of a cigarette packet, but I consider it too bulky and too much weight for this exercise. Besides, they require some serious elentronic enginerring to design and build them, it just isn't worth the effort. The receiver also needs to be of very high quality and a normal radar gun is not good enough.

But Radar guns themselves use DRO's or Gunn diodes and have lousy phase noise.
They drift in frequency and do everything I just said above which is bad. But it doesn't matter. The gunn diode can drift in frequency because the dopler frequency is always going to be relative to the source frequency and so the IF is always going to be the same frequency for the same speed. The differences in frequency drift at the X or K band end of the radar only manifest themselves as tiny differences in dopler frequency at the IF end. The instability of the radar unit only varies in fractions of a kph.

I've been talking to ralph about the frequency divider for radar guns. There is a whole range of suitable chips, but this was a nice one I found in the datasheets. You should be able to buy it from your local Jaycar / Dick Smith / Radio Shack / Tandy.
If you can't find them, then there are other chips which do the exact same thing.
A 4017 is another chip, different pins, but easy to use.

Here's the blurb
it's a CMOS 4042 chip right now. It's called a D-type flip flop.
http://pdf1.alldatasheet.co.kr/datasheet-pdf/view/11961/ONSEMI/MC14042.html
This chip is probably as simple as it's going to get. You won't need any external components and it will run on whatever voltage you have in the gun.

Pin 16 is +V
Pin 8 is 0V
Pin 6 is connected to pin 8
Pin 3 is connected to pin 4
Pin 5 is the signal input
Pin 2 is the divided signal output.

That's it.... nothing more to it.
The modification to the radar gun involves cutting a track that is carrying the dopler IF, and wiring it out to a switch and this chip.

The trick comes with knowing where to cut the track. There are several places where it can be done. But with a circuit diagram, I can tell you the best place to insert the divider.

trash0

Nov 16, 2007, 08:52 AM

I did have another thought tonight about the radar guns and their design and intended use. Which is for law enforcement of motor vehicles. Duh!
Any iffy data is not going to be acceptable for the CPU. It's either all or nothing.
A fraction of a second for a weak target. You don't want that sort of data getting into court to oppose speeding fines.

So for the purpose of measuring speed records of very small targets above the range of the CPU, I just thought, maybe there's something better. And there is.

Many years ago we used to steal the automatic door modules off shopping center doors. There was a rash of vandalism as the local CBer's nicked the modules and used them to build their own microwave links between each other. These days you can buy these modules on ebay dirt cheap or recover them from old radar detectors and alarm modules. Some of you might have seen the modification I did to the hotwheels radar gun using one of these modules. Police radar guns pretty much use the same equipment.

The idea I had was to take an old module and either point it at your target aircraft, or better still, mount the module on a small dish and point it at the target area.
Power the module with the required 7 volts and connect the mixer unit to a tape recorder (I suppose that's an MP3 recorder these days). I feel old !
You wont know the speed instantly, but you can play the tape back anytime and listen to the dopler shift at will.
Connect this audio up to you computer and use a frequency counter program or a spectrograph program to work out the frequency and thus the speed.
If you had a laptop connected directly to te module, then it could record the audio while at the same time display the maximum speed on the screen.

It is also possible just to put an audio socket in the side of your radar gun and pipe the dople signal out to a tape recorder or computer in the same way.

As long as you know the gunn modules frequency, you can calculate the dopler speed on the computer. Some good DSP programs should be able to pull even a bad signal out of the noise for the purpose of confirming a record.
Best of all, you can send the audio file of the dopler shift to anybody as confirmation.

On the subject of dish antennas and beamwidths. It is easy to widen the beamwidth of a dish by defocusing the feedhorn. It is also possible to use a larger apature. But it becomes an issue to lug a 2m dish up a hill.
It might sound like a dish antenna does not offer enough beamwidth, but when you consider it is very similar to a standard hand held spot light, well it's kind of easy to point a dish in an area where you expect the plane to pass through.

Wifi antennas aren't good enough for X or K band. An old satellite paytv antenna is perfect. A small offset feed dish works a treat. It's smal, light and offers just enough and not too much gain. The offset angle presents a challange for some people.

I wonder where the maximum velocity occurs in the path of the plane and is the radar operator square on to the glider when making this measurement ?
A couple of metres around the loop (90 degrees) would make all the difference between zero and full dopler shift.

Ralph Walton

Nov 16, 2007, 09:39 AM

Thanks for joining in Trash!

Ok guys, no response the first time, Trash would like some photos of the insides of or better still, circuit diagrams, of popular guns, that people want to mod.

Mike the Snake

Nov 16, 2007, 09:44 AM

2 main issue exist I think.

Most importantly, we are flying faster than the guns can read. That would be the one thing I'd want changed on my gun.

Next is the missed readings at high speeds, which I think is much less of an issue.

At high speeds, the angles and time are totally crucial. My gun picked up more fast passes when aimed perfectly, as have other guns I've used.

I've seen some guns have issues, and after moving just a few feet to another position they then worked fine.

The environment isn't the best for laptops. Wind and dust.

Ultimately, the guns just as they are (fit into a backpack) that read higher readings would make me happy.

How do the sensitivity settings work? I've read that most set theirs at the highest sensitivity. Does this mean a more focused beam?

Would it be easier if Trash had a gun to work on?

daboz

Nov 16, 2007, 10:21 AM

thanks Trash, and one more point. Weight is not as much as a factor as you might think. Some of our planes weigh about 7lbs with less than an 80" wing span. We add led to them for ballast to make them even heavier.

Thanks!

XJet

Nov 16, 2007, 12:51 PM

Trash0, there's a major problem with those gunn diode radar modules, most of them saturate the mixer diode because there's just too much leakage from the osc. This tends to desensitize them significantly.

We did quite a bit of experimentation with these modules (looking for a cheap, small option for our UAV proximity/collision-avoidance radar system).

Even with a 17dB horn they were unable to provide much range, even when dealing with something as "big" as a 172.

The other problem is that any mismatch between the gunn module and the antenna (qutie likely in the amateur setup being considered for a DIY speed gun) significantly increases the level of mixer saturation and further degrades the s/n ratio of the whole setup.

"Defocusing the beam" from a dish is a win/lose situation. As the beam gets wider, the strength of the RF field that bathes the model becomes less -- so the strength of any reflection is correspondingly less.

In theory, the *best* antenna is one that focuses the transmitted beam into a tight coherent beam so that as much as possible of that energy is reflected back to the receiver.

One thing we had good success with was using separate transmitter and mixer modules (with their own horn antennas). This allowed the level of mixer injection to be more accurately controlled and allowed the mixer bias to be set for maximum S/N ratio.

Just a few thoughts based on my own observations and experience.

trash0

Nov 17, 2007, 06:38 AM

I haven't ever seen the diodes saturate, but I haven't gone looking for it in the past. Two cavity units will have much less mixing signal, but it's more a case of the noise figure of the diode and the gain of the RX system (none).
It looks like you'e already discovered that these kinds of modules exist.

Circulators and Magic-T's will offer you isolation that is worth the effort.

I don't use gunn modules for communications anymore, GaAsFETs offer the required performance. I haven't seen inside any of the ex-police radar guns, they may already use a DRO and FET. Improving their sensitivity is still going to be a task either way.

Converting an old satellite LNC (LNB) could offer the performance your looking for.
The trick is to start with a Low band (Kuu) LNC and re-cut the front end bandpass stage or broad band it,remove the DRO and pipe in a signal from your source gunn diode and remove the L band IF section and replace it with an wide band audio/ultrasonic amplifier. There's some issues with doing it and my biggest concern is that most of these ideas are beyond the reach of the average hobbiest.

The good news is that there is simple options for testing modules and designs that is cheap and simple and anybody can do them. If members can get their hands on gunn modules, then they can experiment with what works without ever actualy measuring the speed. Powering a module and using a crystal ear phone on the mixer diode will work. It's a microwave crystal radio dopler unit. :)
Find methods of recording the dopler audio and then playing it back later for evaluation.

Using any dopler radar for collision avoidance is not my idea of a good idea. I have no idea how big a 172 is.
I'd be interested to hear more about your experiments with it.

Matching the antenna to the module is very simple, a single tuning screw is usually more than enough to get a good match. In some horns I've seen 3 tuning screws, but a single screw is all that is required.

Yes, defocusing the beam offers less gain, but it is still better than the horn antenna on the radar module. It's a ballance between too wide and too narrow. I don't have an answer for this yet. It is something that will have to be done through trial and error.

I'm always supprised how well some of the really silly suggestions can work. The pizza box antenna was originally a bit of a joke. I had to stop laughing when people reported back how well they worked. The theory was sound, but I never expected it to work.

I'm a little stumpped how to get a small model aircraft to present a larger radar signature without effecting aerodynamics. Microwave reflectors under the fuselage in the nose cone and behind the leading edge of the wings are my choice.

I didn't realise how heavy these planes were, so it sounds like there is plenty of room for small GPS loggers or other more exotic circuits.
GPS is rather slow respond, so the logs may not show the aircraft moving any significant distance or speed for the pattern being flown. Airspeed would be easy to measure, but relative ground speed is obviously what you want.

I'm sure the is a simple solution for this problem somewhere.

Andrew McGregor

Nov 17, 2007, 04:00 PM

A 172 refers to a Cessna 172 light plane... aluminium airframe with an 11 meter wingspan. An absolutely massive radar target compared to a DS glider, even if you don't take into account the stealthing of the glider.

TheAdrian

Nov 18, 2007, 06:29 AM

Various ideas:

Dopper radar:

The easiest way might be to modify an existing doppler radar unit to tap off the audio signal after the detector. You could then analyse it using a PC sound card and suitable software. You could record the audio and analyse it later for developing the software. Once you have the software working you could take a laptop up the slope with you and do the analysis in real time.

You could mount the radar on pole so it is on a short tangent to the fastest portion of the DS circle. Send the audio down a cable so the operator can be a safe distance away.

If you really can't get a big enough return echo for your radar you could fit an active transponder to your model. It could receive a reference signal and reboadcast it modified by some fixed N/M ratio in a phase locked loop. Your sensor unit could use a similar phase locked loop to generate its local oscillator. This would incidentally eliminate leakage from from the radar transmitter directly to the receiver and so get round any swamping issues.

GPS:

One presumes GPS software is optimised towards computing accureate postion. In your case you don't need the precise position so you could optimise the software towards quick computation of velocity.

An advantage of measuring the velocity vector is that then you can compute the actual speed, rather than the component in the direction of you sensor. So you can legitimately claim a slightly higher speed record.

D.S.

Nov 19, 2007, 04:26 PM

Stalker Pro

D.S.

Nov 19, 2007, 04:52 PM

Sorry the photos aren't great. I took this one outside in the direct sun. :o

Dean

XJet

Nov 19, 2007, 09:16 PM

I see they're using separate transmitter and receiver horns (as I suggested in a previous post)-- possibly to avoid the saturation and resulting lack of sensitivity I've observed with traditional Gunn-diode based doppler modules.

Pretty old design though -- most of those chips are 14 years old :-)

Ralph Walton

Nov 20, 2007, 08:00 AM

Thank you Dean. It's nice to see that at least 1 person is interested in a faster gun!

I've often wondered whats inside a stalker pro.

I presume one horn is a (catcher) for the signals bounced from the glider. Would there be any advantage having a horn with a larger opening?

Lets say the "catcher" horn was twice as long, it would have 4x the open area at the front. Would it catch 4x the return signal and therefore miss fewer passes?

Can anyone show the insides of a prospeed?

Dynamic_G

Nov 21, 2007, 02:58 PM

The question is, Dean, "will it work when you put back together again"?

D.S.

Nov 21, 2007, 04:26 PM

The question is, Dean, "will it work when you put back together again"?
"Relax, all right? My old man is a television repairman, he's got this ultimate set of tools. I can fix it." :D

http://www.imdb.com/title/tt0083929/quotes

There really isn't much to putting it back together, just plug the circuit boards back in and reconnect a few things. It's still in pieces now, I know that as soon as I put it all back together somebody is going to ask for another close up photo or something.

Dean

trash0

Nov 22, 2007, 09:03 AM

If you really can't get a big enough return echo for your radar you could fit an active transponder to your model. It could receive a reference signal and reboadcast it modified by some fixed N/M ratio in a phase locked loop. Your sensor unit could use a similar phase locked loop to generate its local oscillator. This would incidentally eliminate leakage from from the radar transmitter directly to the receiver and so get round any swamping issues.
Oh, my head hurts after that post :)
Ok... fitting an active transponder requires accuracy. This increases complexity, cost and weight. Though it is possible, it is not something I consider practical for readers of this forum.
If you would like more information on the subject, then search for "low phase noise microwave oscillators" and "crystal oscillators with frequency multiplication". I cannot emphisise how important frequency and phase stability are for such an oscillator.

Rebroadcast the signal. I would normally put the suggestion into the silly ideas basket, but even silly ideas do have the ability to come bouncing back.
In reality, it isn't silly at all and there is a microwave circuit which does just this. It's even used for calibrating radars, and forms the basis for some radar jammers.
There are two kinds. One is called a "Mark Generator"
It is nothing more than the mixer diode in the waveguide. The diode is injected with a signal, lets say 10MHz. In incoming 10.5GHz signal is mixed with the 10MHz signal and reflected back out the waveguide to give a 10.5GHz signal with 10.498GHz and 10.501GHz side bands.

A Magic-T waveguide mixer can return the sidebands without the center frequency. This is a technique used to stop a local osciallator swamping the mixer diodes. When used with in a radar detector, it goes to some way to making it undetectable. (in theory). When used with a radar, it can determine direction of movement.

The best circuit for the job is known as a "microwave boomerang".
It consists of an antennas, a circulator, a double ballanced mixer, a local oscillator. Sometimes filters and amplifiers are used in the circuit.
It does a similar thing to the above two components, though it gives a little more control over the signal.
It will shift the frequency by the local oscillator's frequency.
If the oscillator is 100MHz, then a 10.5GHz signal will be shifted to 10.6GHz, plus any dopler shift. It could also be set to send back 10.4GHz plus dopler shift.

It doesn't address any swamping problems, it just shifts the RX frequency.
It really isn't an issue to change the construction of the detection circuits so they use a reduced LO power. It's still not really a problem the LO power, because most mixers require a large signal in order to work. Rather it's a case of the RX signal being very small. Ideally a low noise pre-amp would be a better choice. But there is more than one way to skin a cat.
K Band low noise amps are expensive. Ka Band satellite LNC's (are ~20GHz) and are a cheap source at ~$300. Doh !

One presumes GPS software is optimised towards computing accureate postion. In your case you don't need the precise position so you could optimise the software towards quick computation of velocity.[quote]
Precission is part of the picture, especially on a geographically small scale that RC models fly. GPS NMEA communications tends to be slow 9600bps.
I don't know enough about how quickly you can communicate with the modules. If somebody is keen on PIC programing and serial communications, the GPS modules are very small and cheap.
The PIC can interogate the GPS at regular intervals, but there is no way of calculating how far the glider has moved as it passes through these two points. The software assumes a straight line (shortest path) where in reality the path is curved, longer and actually faster.
The Pic software of read and capture is simple, knowing how to interpret it is hard.

[quote]An advantage of measuring the velocity vector is that then you can compute the actual speed, rather than the component in the direction of you sensor. So you can legitimately claim a slightly higher speed record.
Your full of interesting ideas. Consider capturing 3 points of the flight. The shortest path between all three is two straight lines. But, the quickest path is a single line, and one which pulls the least amount of average G's through the total length of the path.
The maths is too complex for a PIC chip to do inless you like doing 8 bit calculus. Argh ! Instead we capture the three points, or more correctly, the pic captures as many points as quickly as possible for the duration of the flight. They are downloaded to the PC and then the calculus applied to any/all three points in the series. It will become obvious where the highest speed occurs.

An air speed indicator keeps popping into my head, but I have to keep reminding myself it's not about air speed.

Ok.... sorry guys for lagging behind, I often work in remote locations.

Thanks for the pics DS. If possible, can we have some more ? :)
My requests are; A picture down the waveguide, the tiny circuit board attached to the top waveguide. This appears to be the gunn diode cavity.

Interesting to note Xjet, that coax joining the two waveguides together.
I wonder what the coupling loss is ? A photo down the waveguide will give us an idea.

The bottom cavity, the one with the gold coax connector appears to be the mixer. I'm interested in some photos of the metal box on the coax with 5 wires and the coax passing through it.
The circuit board where the gold connector (SMB) also connects.
But, there other side with the tracks on it. I want to see where the center pin connects to.
It looks like the TL082 op amp is the first audio stage, the output of this chip is the ideal place to take an audio sample from and possibly insert the frequency divider.
A disturbing sight is the Analog to digital convertor. There is also an 8 channel multiplexer (the 4051). It's late at night and my brain can't work out what they're using the chip for.
I can't quite make out the two small LM??? 8 pin chips. They're also probably op amps. This daughter board is the important one where the first processing of the dopler signal is done.

There also seem to be a lot of other components missing. Looks like this radar gun has a bigger brother with more complexity.

I've been thinking in terms of X band 10.525, but this radar is K band, so I'll try and stay consistant with K band 24.150GHz. There isn't much difference anyway.
Though I was just entertaining the thought of Ka Band radar 38GHz and how well it reflects off small surfaces.

D.S.

Nov 22, 2007, 11:09 AM

A picture down the waveguide

I already have a photo of that but I just didn't post it before, I will get the others you asked for when I get going this morning.

I have talked to the tech / repair guy at Stalker several times and he would probably be very glad to answer any questions. He has been in this field for 50 years and even though he is supposed to be semi retired he enjoys it so much that he doesn't want to stop. He told me that only the software would need to be changed to read higher speeds with the Stalker Pro gun. Unfortunately, he doesn't do software and there is practically no chance that we could get the Stalker software people to help us with that. I don't know if it is even possible to solve the acquisition problem with just a software change.

I have heard that they are working on a replacement for the Stalker Pro sports radar gun that may be capable of reading speeds as much as twice as high as the current model. Nothing official from the company yet, but that's what I have heard. The current Stalker Pro is an old design and the new gun will be quite different. I have heard that it will probably look like Stalker II. The Stalker II gun has a round case like the Prospeed gun and I would think that means that it will have a single horn rather than the dual horn design used on the current model. I have not heard if the new gun will have an improved ability to detect difficult targets like our planes so it might not be any better for our application whenever it does finally become available anyway.

Stalker does currently offer other models with a few more features like the Stalker ATS (http://www.stalkerradar.com/sports_ats.shtml) and that probably does explain the missing components. The ATS has some cool features and I was told that they will covert a Stalker Pro gun to the ATS version by replacing the boards and the rear display and control panel. They charge $1,000.00 for the conversion. OUCH!

Dean

The Stalker Pro
http://www.stalkerradar.com/sports_pro.shtml

The replacement for the current Stalker Pro gun is probably going to look like this
http://www.stalkerradar.com/sports_ats.shtml

http://www.stalkerradar.com/images/stalkerII_banner.jpg

Andrew McGregor

Nov 22, 2007, 03:47 PM

Well, I just bought one of these: http://www.jaycar.co.nz/productView.asp?ID=KC5441

Reading through the notes, it appears the speed limits of this design are to do with the analog filters in the front end... easy to change. And, I can find out if a mod is working because the stock limits are low enough that the local pilots already regularly exceed them... so if I can read them when they're going fast, it worked. I'm going to use a better antenna, too... coffee cans might be cheap, but they're not particularly good.

Only trouble is, it runs at 2.45GHz, which may play havoc with Spektrum radios. Only one way to tell, though.

D.S.

Nov 22, 2007, 05:27 PM

http://www.rcgroups.com/forums/showthread.php?t=205460

The kit to build this one looks similar but costs about half as much.

http://www.ramseyelectronics.com/cgi-bin/commerce.exe?preadd=action&key=SG7

Dean

http://static.rcgroups.com/forums/attachments/1/7/1/9/9/a207914-85-radar%20gun%20pan.jpg

D.S.

Nov 22, 2007, 11:05 PM

A few more photos for trashO.

I've been thinking in terms of X band 10.525, but this radar is K band, so I'll try and stay consistant with K band 24.150GHz. There isn't much difference anyway.
Though I was just entertaining the thought of Ka Band radar 38GHz and how well it reflects off small surfaces.

The literature says that the Stalker Pro is a Ka gun. :confused:

Dean

Ralph Walton

Nov 23, 2007, 01:53 AM

Watching with interest.

Can anyone post similar pictures of the innards of a prospeed gun?

I'm interested to see what Trash thinks of them, if they have any potential, worth buying, etc.

I'm still fascinated that one prospeed gun ( or the driver! :) ) seems to be alot better than the others.

trash0

Nov 23, 2007, 10:10 PM

He told me that only the software would need to be changed to read higher speeds with the Stalker Pro gun. Unfortunately, he doesn't do software and there is practically no chance that we could get the Stalker software people to help us with that. [quote]
I can do software, I can even decompile software and reverse engineer it.
Being a bad satellite pirate for many years, I've written lots of machine code.
The problem is knowing what the processor is. I may have to learn another CPU architecture, then it's a matter of getting the firmware out of the gun/ROM. It's painful, but it can be done.
The end result would be not modifying anything in the gun, just swap out the ROM (like a play station chip) and it's done.

[quote]I don't know if it is even possible to solve the acquisition problem with just a software change. normally I would say no, but now that I know it has a A2D converter onboard, it certainly is possible. There are plenty of cool DSP tricks that can be done even with a simple CPU.

Interesting that these units are also marketed for civil use.
So talking to the Stalker customer support should help.

Yeah, I just noticed that say Ka Band in blrub.

I'm still really curious about what is in the metal box attached to the coax.
I'm slowly working through the circuit board. I'll reverse engineer some of the front end circuitry and pick off a monitor point that can be connected to a computer sound card. I'll also post a spectrograph program that can be used to test the tap off point.

I'm sorry I'm a little slow to respond on posts at the moment because I work away from home during the week, and don't have much time for myself when I'm at home on the weekends. Just busy before xmas.

Ralph Walton

Nov 29, 2007, 09:13 AM

Hey Andrew. Clocked anything with your radar yet?

Andrew McGregor

Nov 29, 2007, 05:00 PM

Nope, I've been real busy with work and it's still a box of parts. It's going to have to wait till I get back from Canada in about a fortnight before I can even start to assemble it.

Ralph Walton

Dec 02, 2007, 09:25 AM

D.S.

Dec 03, 2007, 10:41 AM

Has anyone got a prospeed they can photograph the innards of? This will help Trash0 to give advice and opinion as to what can be done to improve them.

I'm still on the lookout for a Stalker Pro gun, so if anyone has one for sale , please PM me
Better ask both of those questions over in the DS forum. That's where the people that have them hang out. This thread seems to have gone cold so they probably aren't reading this one any more.

I took my Prospeed (which I don't have anymore) apart to replace the clear display window. It's very easy to do, just remove the two nuts at the back and pull on the aluminum ring around the antenna dome at the front. There is a rubber O-ring between the aluminum and the metal can so you need to give it a good tug but it will come out all together as one piece, nothing falls apart. :)

Dean

Mike the Snake

Dec 04, 2007, 09:31 AM

I am watching both threads with great interest.

Jantar2A

Dec 07, 2007, 11:34 AM

yeh, what mike said.

N

Kito

Dec 07, 2007, 01:10 PM

setup a transmitter and measure the speed as increase of amount of time necessary for the reciever to recieve the freq.
as it moves in a direction it will take more or less time

Ralph Walton

Dec 10, 2007, 02:16 AM

Just a reminder to keep this thread on track

Many other methods of speed measurement such as gps are being discussed but can we please just keep this discussion to improvements to
radar guns?

Kito

Dec 10, 2007, 07:48 AM

Just a reminder to keep this thread on track

I figured the op meant Radar Guns arent fast enough, What else can i do

Andrew McGregor

Dec 10, 2007, 12:04 PM

setup a transmitter and measure the speed as increase of amount of time necessary for the reciever to recieve the freq.
as it moves in a direction it will take more or less time

That's been discussed; it's too hard to make the oscillator stable enough to do that… if I understand you right, which I'm not sure I do since you didn't specify it well enough.

Ralph Walton

Dec 12, 2007, 09:32 PM

Had an email from Trash0

He's started on a test jig to test divider circuits. His plan is to make circuit boards for sale "for a few dollars" to those who want to mod their guns to read higher speeds.

He was thinking of picking up a stalker pro "cheap" so he could play with it. When he found out what they cost :eek: he changed his mind.

Just to repeat the intent of this thread, when the guns can read higher, Trash0 is going to help with the many missed passes issue.

And once again, yes I'm shure there are alternative methods of measuring DS speeds (gps etc) but this thread is only concerned with improving present radar guns :)

colbourne

Dec 17, 2007, 09:59 PM

Is there any point in modifying the $20 Hot Wheels gun for this purpose.
It would need the divide circuit and to be made more powerful as explained on the Austech thread by using another gunn diode.
After these mods is it still a piece of junk or could it be used for measuring dynamic soaring speeds.

Another question slightly related :

Dick Smiths were selling a very cheap GPS for AU$69 which was designed to be used by skiers. Sorry I cant tell you the model name .
I wanted to know what the maximum speed it could measure is, and whether it would be suitable for model use.

D.S.

Dec 17, 2007, 10:27 PM

Did you not notice that the main guy (Trash) over in the Austech Hot Wheels gun mod thread is the same guy trying to help us here? :rolleyes:

Ralph is going to be upset because you mentioned GPS in his thread. :D

Dean

colbourne

Dec 18, 2007, 12:24 AM

Yes I did notice and it seems like Trash definately knows his stuff. We are lucky that he is interested in this field.
Has he used a Hot Wheels gun with a glider yet successfully and if so up to what speed ?

Sorry about the mention of the GPS , oops I have done it again :)

Ralph Walton

Dec 18, 2007, 05:43 AM

GPS again! :rolleyes: :rolleyes: !

Trash was just playing with the hotwheels gun for fun. He has never previously had anything to do with DS.

The tread in the Austec forum was how I met him. He obviously has forgotten more about radar than most of us will ever know. ;)

I approached him about the gun problems associated with DS and he was attracted to the challenge. I'm always impressed how willing, successfull people are, to donate their time and effort to problems that often have no gain for them. He's doing it mostly for the intellectual curiosity. ( he was pretty impressed with some of the links to DS videos I sent him too)

Has he used a Hot Wheels gun with a glider yet successfully and if so up to what speed ?

Not as far as I know. He has picked up the odd jumbo jet on landing approach over his house though. Thats a pretty serious toy! :cool:

Ralph Walton

Dec 27, 2007, 03:14 AM

Got info from Trash0 on the "divide by 2" circuit. :) He posted it in the DS forum , but I've moved it here to try to keep this gun mod info in one place. I'm not an electronics whizz, but it doesn't look too difficult to me.

There's a guy Ezza, from Australia, whose offerred his Pro Sport for Trash0 to mod, once he's pushed his PB up a bit more. I believe the Pro Sport is similar inside to a Pro Stalker. Not shure if the important bits are identical for modding, but if they are it would solve the last few bits of the puzzle of the "divide by 2" circuit connection to the gun.

I quickly whiped up a circuit that anybody can build that does a simple divide by 2 function to halve your input so your radar will read 0-300, but will measure 0-600 proportionally.

The trick is now to install it in the gun. Know exactly which wire to cut to pick up the dopler signal.

http://home-and-garden.webshots.com...048941318kWvAtp

trash0

Dec 28, 2007, 02:41 AM

That's been discussed; it's too hard to make the oscillator stable enough to do that… if I understand you right, which I'm not sure I do since you didn't specify it well enough.
I'm quite happy to discuss the problems with doing speed measurement that way. If you want to start another thread, just send me a message with the link and poke me every now and then with an email to remind me to read it.

I don't log into this forum much and a work away from home a lot.

Did you not notice that the main guy (Trash) over in the Austech Hot Wheels gun mod thread is the same guy trying to help us here? :rolleyes:
That's me ! And yes, you're all DOOOMED ! :)

I was just about to grab the hotwheels gun of the shelf and put the test circuit in it. Anybody can build the circuit, it's just a 4013 chip.
There are still some issues I'm concerned about, and each radar gun has suttle differences which I'm sure will anoy me later.

I've not used the hotwheels gun for RC planes. My main excuse for doing the original modification was for motorcycles. (and annoying the cops).

Ralph has already posted the link to the photos I did the other night.

It's about time I also put up a quick DIY microwave mark generator so that you can generate your own phantom radar signals on a bench. There might be some time for that tonight.

trash0

Dec 29, 2007, 07:35 AM

It dawned on me the other day that I don't have a simple piece of test equipment that is handy to have for dopler radar. A Mark Generator or "microwave boomerang" !
They're a very smple to make, well, normally. It took me 30 second to hook up some test equipment and create one. But not everybody has access to some fancy test equipment. So I figured I better build one from scratch. Normaly in communications, we'd just build a crystal oscillator connect it up and away it would go. It's 15 minutes work and it's done. So I figured it' going to be a cinche to build a pissy audio VFO and connect it up. Much to my disgust, I came to realise that VFO audio sinewave generators don't come in a "simple" circuit. I scoured my reference books for design information or at best an example circuit, be the more I looked, the more complex they seemed to be.

I finally settled on a design after testing a few, which I decided was the "simplest" involving the minimal amount of components and the least hassle to get working.
This circuit is the basis, though I've modified it a little to suit our needs.
http://www.redcircuits.com/Page13.htm

The circuit is simple to operate. Turn it on, point the radar at the boomerang, and dial up the frequency you want. That's it.
In case you're all wondering, yes this is how how radar jammers work. Though a good radar jammer is a little smarter. If you'e thinking of sitting this circuit on the dash of your car and setting to to 50kph, remember that the frequency being returned to the radar gun will be your speed + the 50kph on top of that. DOH !

http://home-and-garden.webshots.com/photo/2177834860048941318HaFEyG

You can see in this series of photos what the components all look like, and the hotwheels gun responding to the boomerang set at 60kph.

The module I'm using is an X band module, and a K band module is more ideally suited to the stalker guns, but the X band modules will work well enough.

If you can't find a module, the 1N21 and 1N23 series diodes are very easy to find and they are dirt cheap. Typically <$2. The waveguide is very easy to make. Either with a few blank pieces of PCB or Tin can soldered together. One end of the diode is in contact with the waveguide and the other insulated and connected to the circuit.
The dimesions of the waveguide are normally important in relation to the frequency, but for such a simple applications, beggars can't be choosers and it will work anyway.

So there you have it folks... another useful piece of junk for your toolkit.
If nothing, you can use it to check your radar's accuracy.

Ralph Walton

Dec 30, 2007, 12:30 AM

Posted the link to the "divide by 2 " curcuit again. The one above doesn't seem to work. Might be coz its in a a quotation?

http://home-and-garden.webshots.com/photo/2071633230048941318kWvAtp

Trash0. Your "very simple" circuits make me feel very dumb! That sine wave generator is a bit out of my league. Hopefully I don't need one.

Waiting for my Pro Stalker to turn up.

trash0

Dec 30, 2007, 06:24 AM

I'm sure I can make a simpler version using square waves. But the problem is it's either going to be a fixed frequency sine wave output or if it's varible square waves, the harmonics cause problems. Square waves cause the radar to see multiple signals.
If say it is set for 10kph, it will also be seeing smaller signals for 20,30,40,50,60kph.
The radar should lock onto the stronger 10kph signal, but I can't be 100% sure of the results. We used to use square waves as outright jamming circuits.

The sinewave generator isn't really that hard to build. You can see from the photos of it with the light bulb attached. It took me about 15 minutes to build it once I'd done a few hours of R&D. I can simplify it so that anybody can build one just by following the pictures like the speed double circuit.

I'm working on a digital pre-amp circuit in case the stalkers need it. The hotwheels gun does.

Reminds me, there was another post I saw today on another forum asking about the hotwheels gun and it's top speed. I tested it last night and it topped out at 1999kph !

All of a sudden I can feel the rush of people buying hotwheels radar guns and stripping their local shopping centers of automatic door modules.

As another side experiment, I'll be setting up a module outside and running cables back to the computer. Making a height and scale comparison to large aircraft flying overhead and those of module aircraft. I'm going to run the DSP software on it, and try some special modulation schemes just to see how it plays out.

Ezza

Dec 31, 2007, 09:44 PM

Got info from Trash0 on the "divide by 2" circuit. :) He posted it in the DS forum , but I've moved it here to try to keep this gun mod info in one place. I'm not an electronics whizz, but it doesn't look too difficult to me.

There's a guy Ezza, from Australia, whose offerred his Pro Sport for Trash0 to mod, once he's pushed his PB up a bit more. I believe the Pro Sport is similar inside to a Pro Stalker. Not shure if the important bits are identical for modding, but if they are it would solve the last few bits of the puzzle of the "divide by 2" circuit connection to the gun.

I am glad to report that I managed to get my glider up to 250mph, which tops out my radar gun.

The gun is a Stalker Sport which is one model below the Pro. I am happy to send it to Trash0 so that he can try this experiment.

I have sent an email to Trash0, but he is probably out of contact for the holidays.

Really looking forward to see if this modification can work. :)

Eric

trash0

Jan 03, 2008, 04:02 AM

I haven't seen any emails yet.
I'm sure we'll get a chance to meet up sooner or later. I'm hoping that my work will be in your area in the coming weeks.

It will only take a couple of minutes to look inside your gun and work out the best place to modify it. It's only 1 wire that we need to find.

Ezza

Jan 03, 2008, 05:57 AM

Great news Trash0!!

I'll PM you my details, seeing as the email did not work.

Eric

Ralph Walton

Jan 03, 2008, 06:44 AM

Hi Ezza
Congratulations on your great speeds! (mutter mutter blasted Aussies beating us Kiwis again :rolleyes: ) You've convinced me I need an Opus. :)

I'm glad to hear you've caught up with Trash0 and going to get your gun modded. The sport will be better for the limit being removed.

Hey Trash,
I hope you'll be able to get some good photos of the mod process. Your how to on the "divide by 2" was awesome, and easy enough for a novice to follow.

Is there anything else that needs to be looked at inside the gun while you've got it open?
I was thinking about phase 2 of this project, about getting fewer missed passes. I was wondering if the gun will need any sort of socket or similar fitted, to connect an external aerial / dish etc ?

BTW, did I ask you if video casette tape would be any good for a reflector on say the leading edge of the wings?

trash0

Jan 04, 2008, 09:23 AM

I don't think video tape would be a good reflector. It's mostly plastic. The chromium oxide probably doesn't make for a good reflector.
In the satellite biz, we use Titanium Oxide paint. I personally think it makes no difference.
I'm not confident of dish antennas being suitable for this kind of work since the target is small and moving fast. The beamwidth is narrow and hard to have the target move through the beam.

When we've got the mod worked out, it should be nothing more than a 3.5mm audio socket in the side of the gun. When the circuit is disconnected, the gun operates normally, when the external board plugged in, it interupts the dopler signal, divides it by two and returns it to the gun. The circuit could be placedinternal, but I'm not confident of the amount of space in there, unless SMD is used, which are very hard for newbies to solder.
Anyhow... the focus for now is still on me getting close to a stalker.

I'm still puzzled by the metal box in the RX line of the photos DS did.
I should be able to reverse engineer enough of the control circuit to work out what role it plays. If we're lucky, it's just an amp in which case, the socket on the other side of the box is where we need to tap off.

From the photos, my current guess is Pin1 of the TL082 chip is where we'll tap off.

Jantar2A

Jan 04, 2008, 10:01 AM

I'll be posting pic's of the gut's of my stalker pro tonight for ya to look at

Nick

Jantar2A

Jan 04, 2008, 10:41 PM

Hi Trash,
here are some pic's of the gut's to my stalker pro radar gun,,so far you can just see the 2 main boards inside,,I didn't go any further at this point because it looks like what we are looking for is here???

Just from my guess it would be located on the 2nd board,,the one with the gold rf plug on it. but what do I know

Nick

trash0

Jan 06, 2008, 06:37 AM

Yes. I've tried reverse engineering the circuit as best I can from the pictures. That doesn't say much, but the small TL082 in the corner is the one we want.

What I really want to know is what is inside the metal box on the coax, or at least what it does. It's obviously an amp, but then, so is the chip (two actually)

If somebody has an oscilloscope and one of these guns, then the waveforms as these points are of interest.
espeically pins 1,2,3

I'm still working on the hotwheels gun on the bench here with the divider circiut. trying to make it foolproof.

superchuckles

Jan 06, 2008, 07:24 AM

there are obvious ways you can increase reflected signal - using adhesive spray very sparingly with aluminum foil will add only a tiny amount of weight but will increase your profile considerably. when trying to increase profile, work with what's most efficient - if you only want to use x# square inches of foil, don't waste it on rounded surfaces - flat surfaces (sides of fuselage, rudder, bottom of main wing) provide the best reflections. when i was in the air force, there were special paints that we used on raydomes that allowed the microwaves to pass right thru, but ordinary over the counter flat black (which looked the exact same) would block (not absorb, it would reflect) about 80% of the outgoing signal, rendering the radar all but useless... but in your case, if you used it as a primer (& sanded about half or more off before painting to reduce the considerable bulk flat black adds) it could be a solution that may work for you. if you want a metal paint that will put the thinnest of all possible coats, & gives you a primer, and is cheap & plentiful (will also galvanize steel parts from rusting) is to use cold galvanizing primer that you can buy from any hardware store. before you go thinking it's like any normal grey primer, this stuff goes on so thin it often takes 3 or 4 coats just to get it to cover & not be "see-thru". it has a high concentration of zinc in it, so it's an ideal way to primer, and put a thin yet very reflective coating on your plane all in one move. when/if you do this though, keep in mind try moving your radio antenna as far away from the fuselage as possible, since it's entirely possible it could affect the ground plane of the radio.

one other thing you may want to do for maximum readings (something that in california used to get people off from tickets if they knew) that radar doesn't work the same "everywhere". by law, an area had to be zoned for the usage of radar every 5 years - if it weren't, they couldn't legally use it as a means of calculating speed of a vehicle. you see, the fewer background items you have, fewer parked cars, fewer trees, buildings, chain link fences, etc there are, the less spurious reflection you will get that will affect performance. radar reflects really well of of jimbo bringin' over the brewskeys too - & since a 160 or more pound man reflects probably several dozen times the image, even if you're not pointing right at him.... if he's in the field of the collector antenna & he's closer, it's his 3mph that will be picked up, or no speed will be, as it will be trying to switch back & forth between jimbo & the suddenly appearing then dissappearing target that flew past jimbo.

you can make all the difference with a tiny amount of metal on flat surfaces, and/or coating with a metallic primer that nearly every hardware store carries (cold zinc, cold galvanizing primer, depends on manufacturer, but rustoleum is probably the most common for cold galvanizing primer), and if possible, eliminating as many background reflections as possible. (if you were at the salt flats pointing away from everyone with your plane, you could easily track it for speed i'm sure).

btw - increasing power on the emitter, though fairly easy to do, will exponentially reduce available battery power, while at the same time pumping said power in the form of microwave energy into anyone & everything that can see the open part of the feedhorn. if after all what i've suggested fails, then i can all but guarantee that your gun needs to be re calibrated or have it's components checked out by a licensed tech.

Ralph Walton

Jan 06, 2008, 07:51 AM

Interesting. The Zinc spray sounds like it would be worth experimenting with.

Trash0. I can't help you with an oscilliscope, but maybee some of the members could get a gun to one or vice versa?

My stalker pro is in the country now (thanks Spencer!), but it hasn't been delivered yet. (nothing moving in my neighbourhood will be safe when i I get it :) )

Andrew McGregor

Jan 06, 2008, 07:57 AM

Sadly, I'm a fairly long way from the Mount, or I'd bring a 'scope around…

Anyone in Christchurch reading this? Jono?

superchuckles

Jan 06, 2008, 05:25 PM

if ya have any kind of difficulty getting the zinc primer at a hardware store, try a plumbing suppply or store that carries a lot of plumbing stuff (by that i don't mean the finishing touches like a home depot do it yourself place, though they'll probably have it too - i mean the big plumbing places that have all the rough in piping & whatnot). guaranteed they'll have it if anyone there will (i only now realized you guys are on the opposite side of the planet from me, but am still confident they'd have it since zinc isn't like lead in as far as lead being poisonous). i know it lays a fairly good layer of zinc on - for example - i had been doing a lot of custom frame work for drag racing a number of years ago, & i had a piece of 1/4" thick plate steel (roughly 7mm) that was about half a meter wide, & a meter long - shiny clean steel plate. it was a scrap piece more or less & i had a can of cold galvanize zinc primer so i sprayed only one side of the plate & left the other side bare. for 3 years i left the plate out in front of my garage in the california rain & sun & the only place it developed any rust was in a few of the deepest grooves around the edges where i'd cut it with an acetylene torch originally -- & even that was just a light dusting which would come off with a wire brush in one or two scrapes. that's right, no rust on either side.... the bare side was still just as metallic (though people handling it all the time did discolor the metal after a while -- people were always in disbelief that it could sit year after year without rusting). i don't know what ever happened to it... one day i noticed someone had swiped it, but it had proven it's point to me that it indeed did have enough zinc in it to actually inhibit rust thru galvanization. i guess i was most surprised because it seemed so little paint came out with each coat that it was hard to accept that it carried enough metal with it to be effective.... but it was very much so.

if you've never seen or handled zinc by itself, it looks, feels, & acts very much like lead - main differences are of course it's not hazardous, & it had a different electrical potential from it's ground state (which is why it will corrode before the steel will as long as it's in direct electrical contact with the steel)..... anyway, just think of it as spraying an almost transparent thin layer of lead - you know that's gotta reflect some signal.

D.S.

Jan 06, 2008, 06:26 PM

That stuff is pretty easy to find on the net if you do a search. I buy my steel for projects from IMS (http://www.imsmetals.com/framedcart.htm) and I have purchased spray on zinc coating from them. The product they carry is called SPRAY-GALV and it's only $4.25 per can. Like superchuck said, it takes many light coats before it looks like you are getting anything on the surface.

Dean

Ralph Walton

Jan 08, 2008, 08:58 AM

No problem getting cold galv in NZ. when i'm in the position to have missed passes, (my radar gun is in transit) I will give it a go.

Hey Ezza Have you managed to get your stalker sport to Trash0?. I think that will give this project the great leap forewad that it needs and deserves.

I see theres a 375mph claimed gun on the DS forum

Trash0. I was asking about the VHS Tape for the following reason,

A few months back, a man was flying a kite about a mile from the local airport. The largest planes we get here are 20 or 30 seaters. He decided to use tape from a VHS video cassette for the tail to stabilise the kite.

He had a couple of hunderd feet of doubled over tape and was flying quite blissfully unaware that the local airport ad picked up a radar signal the size of a JUMBO JET in his vicinity .

They quickly realised that the "jumbo jet " wasn't moving, and called the local police to investigate.

They easily found the source of the massive radar signal and ordered the man to get the kite down. I don't know if he was charged or just warned.

Another question for you Trash0. I see you can get much higher rated gunn diodes than the 25 mW ones in the pro stalker. Would a say 100 mW one work better for our purposes? Looked like they were about $45 US where I saw them.

If you put more grunt into them , how high a wattage is still safe?