I just pulled apart a HotWheels radar gun and what did I find? An ATMega88!
For those who do not know what this toy is:
What does this all mean? Not much really, except if you want to know where AVRs are used - here is a high volume app.[/url]
Pictures would be a cool thing to post. Do you have a digital camera? I would like to know what technology is used to measure speed.
I bet its ultrasonic
Imagecraft compiler user
Oh I am SOOOO hacking that thing! :twisted:
Happy electric car owner / builder
Nope! Its not ultrasonic - it is actually radar. It uses a DRO (dielectric resonant oscillator) I'm told. So its using the old doppler effect technique methinks.
There's some discussion and pictures on this forum.
Nope! Its not ultrasonic - it is actually radar.
Before the thread falls by the wayside can I ask if you had a radar project in mind? I was hoping to use a DSP and allow the gun to work in my car. So far other cars are invisible when I'm driving unless they're moving in the same direction as me and about +10mph faster. I figure a DSP would help me it ignore terrain echos and just track cars.
That's the problem with RADAR! It detects anything that moves! Due to the beam spread, it will pick up the road, the power poles, the mailboxes and maybe a car! I'm not too sure a DSP can help you here as the usual radar gun works by the doppler effect as opposed to you aircraft tracking radar that sends a pulse and looks at the return, doppler effect mixes the return signal with the send signal and you get a difference frequency. You might be able to pick out items moving at different speeds by using DSP. Best be going to the technical library and doing some reading!
Personally, I'd use a laser rangefinder - they're a bit pricey though (>$1000) but you get what you pays for.
How can they get away with making a radar gun and the whole liability of radar exposure, etc..?
exposure can be neglected if radiated power is low enough, atleast thats the idea of it, however, i wont recommend to point it directly against your brain for to long. I guess radiation from your "SMS messenger device" should concern you far more than this toy, afaik. ;) Actual any device radiating power of some extent can be a health danger, not only devices named to be a "radar". :twisted:
"The fool wonders, the wise man asks"
I've been looking at that Hot Wheels gun, and also that thread from AUStech. There's a lot of interesting potential in this toy ...
First, a note about the radiation exposure question: this thing is something like 1 uW, and the duty cycle is only about 13%, IIRC, so there's absolutely minimal exposure. You'll get more radiation exposure from walking through one of the old supermarket door openers.
BUT ... (and this is why I joined this forum), like I said, it's got a lot of interesting potential.
This radar appears to be pulsed, in order to obtain the low duty-cycle numbers. And that can lead to a lot of fun and games. Because of the pulsing action (god that sounds dirty! :D ) you SHOULD also be able to derive range information, with the addition of appropriate circuits.
Also, through the use of some filters that could easily be done in firmware - they're simple, really, just comparing the returns of each pulse to the two or three previous - you can cancel out radar returns from objects in a certain speed range. Usually this is used to cancel out terrain or stationary objects, but since y'all were mentioning mounting it in a car, you could feed an own-speed input, and maybe some vector information from accelerometers, and use that to cancel out objects which can be identified as zero, leaving only moving objects, irrelevant of the doppler.
(Which leads to the reason I'm here ... I'm looking to see whether AVR or Parallax MCU's are going to be better for the needs of my project.)
This could have a lot of interesting potential in robotics projects, model rocketry, R/C aircraft, etc.
Yeah there are definately some nice rocketry & model air applications for it. During another test, we clocked a potato gun (which I may or may not be affiliated with) muzzle velocity at 176mph. Supposedly, any speed between 10 and 200mph is visible to this model.
What I was hoping to do is use a high-speed DSP to run a discrete fourier transform on the echo. An input telling the DSP the car's present speed (like JesseCuster suggested) should allow the program to look at the frequency domain --likely in the form of a table--and ignore the terrian echos. After that, looking for peaks in the table data should indicate something of interest.
My major obstacle for now is that I've never worked with a DSP, so I'm not sure of what to shop for. We're also probably looking at high-speed A/D conversion.
Speed and I/O requirements of this sort of thing have had me looking at the Parallax Propellor series - 8 MCU's in one chip, with scads of I/O pins ... but they're EXPENSIVE, so I've been looking at some of the other options out there, like the AVR and PIC stuff.
It's going to be a couple months before I have the spare cash to throw into this project, so I've got time to do the research on what's going to be best.
(BTW: the experiments using the Hot Wheels radar will be more of a proof-of-concept experiment than anything else, with my final project being a lightweight portable radar, with TWS, STT, and all the other neat little stuff. Hoping for a final product in the 3-5 kW range.)
3-5 kilowatt??? Is that accurate?
Mike Coles http://blips.net
Yup. Put through a 256-element phased array, that's less than 20 watts per antenna.
'course, that is several years down the road ...
First step is mechanically steered small system, to get the processing and programming, then a ~1 kw electronically steered to get the phase shifting circuitry, then do the big guy.
3-5 kW isn't that much, really - most small craft navigation radars are in that range, and you can get them relatively inexpensively. I'll probably wind up using a magnetron from a junked Furuno. The antenna design and control systems for the antenna and the tracking system will be the difficult parts for the big system.
I want to check up on something before I start taking this one apart:
As I understand it, the speed measurement is done by mixing the echoed signal (amplified) with the original K-band tone. This gives you the sum and difference frequencies. Then, that signal is run through a low-pass filter to remove the sum element and leave behind the beat frequency. Finally, by looking at this frequency (and maybe phase), I should be able to calculate the target's speed. Does this sound about right?
First, it's not K-band, it's X. 10.5 gHz. (Or at least mine is ... )
And I'm not sure it's even that complicated. First, I'm not sure that it has any amplification on the receiver, and also because it's transmitting at a stable known frequency, and to my knowledge there's nothing encoded in that transmission, I don't think you need to mess with beat frequencies. All you need to do, really, is measure the frequency of the returned signal, and run that through the following equation:
ΔF = (2 * s) / λ, where delta F is the change in frequency between F1 (transmitted freq) and F2 (received freq), and lambda is the wavelength of the signal. S = the closing speed of the target. (Negative values of S indicate that the relative motion of the target is in the general direction of away from you.)
So, if you know that the gun transmits at 10.525E9 Hz, and you've received back a signal at 85 Hz higher, and you know that the wavelength is .1 (and I don't know for a fact what the wavelength is) meters ... you're left with:
85 = (2*S) / .1
which works out to S = 4.5 m/s closing speed.
Of course, this only represents relative speed ... Unless the target is heading STRAIGHT at you or away from you, you can't use it to determine absolute speed of the target without knowing your own course and speed, and knowing the range to your target as well.
By beat frequency I just meant the actual 85Hz like in your last example. The process was described as being similar to coherent AM detection. ...Mix the transmitter's stable frequency with the shifted echo to yeild the difference frequency. At 40GHz, I'm anticipating a ~1kHz shift for 10mph targets and ~20kHz for 200mph objects. That's a lot lower than I was expecting earlier so a high-end AVR or 8051 may be an option. 40GHz is just a guess at the moment, an email to Bushnell should get me an exact frequency.
Anyway the first thing I should probably do it open the case and see if I can locate this beat frequency on any of the PCB traces. What other ICs did you find in the HotWheels radar?
Oh, that's right - you have the bushnell unit ... sorry! I was still thinking in terms of the Hot Wheels unit.
I haven't opened mine up yet, actually. BUT if you check out that link for the AUStech thread, there's photos in there of the internal components.
Also, some people have been buying Gunn oscillators and transplanting them into the hot wheels unit. So long as you get the 10 gHz units, not the 24's, they'll interface correctly with the hot wheels controller, so long as you provide the correct external power.
Has anyone made any progress on these? I have to admit, I got distracted with another AVR project and haven't put much time in yet. The resources on these low-end radars are pretty sparce.
Deal with beat frequencyes is not too difficult, and can be done with any 'high end AVR', no need for more powerfull uC.
To deal with any RF band directly in the range of MHz (for not to mention GHz) is another world, that is much better to stay far away.
I always had much respect to the guys doing RF, since I was at the university. Design a PCB for 1GHz is more an art (ok, easier in this SMD world of today) than a science. Try yourself to do a circuit with the new Atmel's ZigBee transceiver (2.4GHz), probably it would not work at the first try (probably even if you copy the layout...).
"Common sense is the least common of the senses" Anonymous.
* Bump. I dismantled the assembly this weekend and luckily discovered the system is built from 4 interchangeble modules and not just 1 big board. The stages are clearly marked: Trigger board, Pre-amp, Filter board, and Digit (readout) Board. All connected through small ribbon cable.
Depending on how much I find out about the boards within the next day, I might start planning the DSP immidiately.
© 2020 Microchip Technology Inc.