b) Forget how dirty the device will be, and the distance car x floor
doesn't change.
c) The device pulses a fast narrow light beam 45¡ towards the direction
of movement (ahead of the car).
d) An opto sensor is also pointing to 45¡ (same spot of the emitter),
will be receiving the light pulse bounced at the floor.
Question: It will have a doppler effect in the light beam frequency
received related to the car speed?
It will be possible to use a PIC to calculate this effect?
(Of course sound waves could also be used)
> Question: It will have a doppler effect in the light beam frequency
> received related to the car speed?
Yes, but not that large.
> It will be possible to use a PIC to calculate this effect?
> (Of course sound waves could also be used)
Think microwave. 10 or 24 GHz is commonly used in these apps, 1mW is TONS of
power.
Easy interface too, two wires. DC in, doppler out. (Homodyne receiver)
Wagner, ground speed radars are quite common, especially in agriculture
etc., where they can be used to calculate wheel slip etc. using true (radar)
ground speed readings and wheel (odometer) speed readings. Check out the
Dickey John units for example.
At the risk of exposing my misconceptions about Dopple Shift, here goes:
Doppler Shift occurs because the transmitter adds it's relative velocity
to the speed of the 'energy' that it's
transmitting, and that increased velocity is perceived by the receiver as
a frequency shift in the 'energy'.
So...if a car is traveling between 0 and 100mph (approx 0 - 150 fps), and
the speed of light emitted from it
is 186,000 miles/second (186,000 x 5280 feet/mile = approx 900,000,000
fps) it appears that the PIC
will be attempting to mearsure a frequency shift (or color change in the
case of light) of approximately
1 part per 6 million. The speed of light here is for a vacuum, I don't
have a figure for air, but it represents
the worst case scenario since the speed of light in air is slower.
Sound travels approximate 1000 fps in air, so the Doppler Shift for sound
would be about 1 part in 6-7.
Sounds to me like measuring the Doppler Shift in light is going to be
extremely difficult with or without
a PIC chip, however, sound doesn't look too bad! You could even time the
returning sound waves.
Mike.
P.S. If this doesn't start a thread about Doppler Shift I'll be amazed
:^)
> Think about this situation:
>
> a) A device is installed under the car.
>
> b) Forget how dirty the device will be, and the distance car x floor
> doesn't change.
>
> c) The device pulses a fast narrow light beam 450 towards the
> direction
> of movement (ahead of the car).
>
> d) An opto sensor is also pointing to 450 (same spot of the
> emitter),
> will be receiving the light pulse bounced at the floor.
>
> Question: It will have a doppler effect in the light beam frequency
> received related to the car speed?
> It will be possible to use a PIC to calculate this effect?
> (Of course sound waves could also be used)
>
> Wagner
________________________________________________________________
YOU'RE PAYING TOO MUCH FOR THE INTERNET!
Juno now offers FREE Internet Access!
Try it today - there's no risk! For your FREE software, visit:
dl.http://www.juno.com/get/tagj.
The cops measure speed with a radar. What they do is use a non linear
xtal to mix the incoming with the outgoing reference. It outputs RF from
the mixing! The rest is easy. Laser doesn't work as well in bad
conditions, but your target selection is more accurate.
--
-----------------------------------------------------------------
Time Spent With Your Dog(Kid) is Time Well Spent
-----------------------------------------------------------------
Sincerely, Dennis Gearon
Dennis Gearon wrote:
>The cops measure speed with a radar. What they do is use a non linear
>xtal to mix the incoming with the outgoing reference. It outputs RF from
>the mixing! The rest is easy. Laser doesn't work as well in bad
>conditions, but your target selection is more accurate.
Non-linear xtal .. nifty! Not that I really understand yet, but it sounds
cool anyway.
Thanks Michael, this is not a common thread really. We don't deal with
light doppler everyday. :)
The doppler can be identified using other chips, for example the use of
PLL's for audio and higher frequencies. The PLL in this case would
generate a DC voltage proportional to the "delta frequency*. I am not
sure about light frequencies, one thing that comes to mind is to use the
refraction characteristics of different frequencies, not sure if the
magnitude will be enough in this case. The PIC function here would be to
account the final data processing and control.
Laser Police Radar uses to measure the reflective time in successive
timed readings, so establishing the distance between the radar and the
vehicle in each measurements, the timed samples tells you the speed.
All the burst of readings happens in few milliseconds, so I really can't
understand what the Laser Radar Detector wants to *try* to tell you when
it detects a laser gun... perhaps "you are busted". Laser Radar uses the
same measurement principle of the RF Radar, the advantage is that a
laser is selective and "illuminate" a single spot in a single car, very
difficult to the car 1 mile behind identify what is happening up front
(what happening with RF Radar that spreads RF all along).
About the Ground SpeedOmeter: Considering the gross measurement of
298,000 km/s (light speed), it is around of 3.3557 nanoseconds per
meter, so this would be the time the light pulse would take to leave the
device below the car, hit the ground and return to the sensor. An
average car's delta speed is from zero to 180 km/h (let's hope), so it
would change the relative light speed from 298,000 to 298.000.05 km/s
when the car is at 180km/h. The delta time to light travel this 1 meter
would be 0.00001%, so impractical to be measured by "non advanced"
technology. This values would change significatively if thinking in
lower magnitude of frequencies, like ultrasound.
>
> Wagner,
>
> At the risk of exposing my misconceptions about Dopple Shift, here goes:
>
> Doppler Shift occurs because the transmitter adds it's relative velocity
> to the speed of the 'energy' that it's
> transmitting, and that increased velocity is perceived by the receiver as
> a frequency shift in the 'energy'.
>
> So...if a car is traveling between 0 and 100mph (approx 0 - 150 fps), and
> the speed of light emitted from it
> is 186,000 miles/second (186,000 x 5280 feet/mile = approx 900,000,000
> fps) it appears that the PIC
> will be attempting to mearsure a frequency shift (or color change in the
> case of light) of approximately
> 1 part per 6 million. The speed of light here is for a vacuum, I don't
> have a figure for air, but it represents
> the worst case scenario since the speed of light in air is slower.
>
> Sound travels approximate 1000 fps in air, so the Doppler Shift for sound
> would be about 1 part in 6-7.
>
> Sounds to me like measuring the Doppler Shift in light is going to be
> extremely difficult with or without
> a PIC chip, however, sound doesn't look too bad! You could even time the
> returning sound waves.
>
> Mike.
>
Actually you can make it even simpler than that. You begin with
a known frequency as your transmit frequency. Nothing non-linear
here. You send little bursts of this frequency out. While the
transmitter is quiet, the bounced and doppler-shifted echo is
amplified and mixed with the original output frequency. The
resulting low frequency "beat" frequency represents the
doppler-shifted component only. (Ignore the higher frequency beat).
Once you know the original frequency and the low beat
frequency, it just takes a little bit of math to calculate the
relative speed of the object producing the echo.
If you can isolate the receiver from the transmitter, then you
can forget about the bursts and just send out a continuous
signal from the transmitter. In general the receiver needs to be
high gain and tuned to respond to a very narrow frequency band.
Some laser type doppler-shift device works along the same principles.
It is very easy to maintain isolation between transmitter and
receiver since the laser beam is very narrow and does not diverge much.
So these are almost always run in continuous mode. A form of
interferometer mixes a small sample of the outgoing reference
signal, and 100% of the received signal (which is usually
quite small). The "beat" frequency shows up as light and dark
bands. By scanning these fluctuations optically an electrical
signal can be produced that can then be measured with a simple
counter to reveal the optical "beat" frequency. A little
math and the velocity is determined.
> Non-linear xtal .. nifty! Not that I really understand yet, but it sounds
> cool anyway.
It's called a DIODE.
The receiver is a homodyne receiver. The "magic crystal"(aka diode) sits in
the output of the transmitter, and also gets signal that is bounced off the
target. These two frequencies are mixed by the non-linear "magic crystal"
and the output is simply the difference in frequency.
Wrong.............it's NOT a diode......! It's a specially doped piece of glass,
optical quartz, or some other clear-to-the-frequency-of-interest OPTICAL
substance that has the property of non linear response to LIGHT. Most new laser
speed guns use it.It gets your speed in ONE burst by doppler.
Dave VanHorn wrote:
> > Non-linear xtal .. nifty! Not that I really understand yet, but it sounds
> > cool anyway.
>
> It's called a DIODE.
>
> The receiver is a homodyne receiver. The "magic crystal"(aka diode) sits in
> the output of the transmitter, and also gets signal that is bounced off the
> target. These two frequencies are mixed by the non-linear "magic crystal"
> and the output is simply the difference in frequency.
--
-----------------------------------------------------------------
Time Spent With Your Dog(Kid) is Time Well Spent
-----------------------------------------------------------------
Sincerely, Dennis Gearon
> Wrong.............it's NOT a diode......! It's a specially doped piece of
glass,
> optical quartz, or some other clear-to-the-frequency-of-interest OPTICAL
> substance that has the property of non linear response to LIGHT. Most new
laser
> speed guns use it.It gets your speed in ONE burst by doppler.
Dennis Gearon wrote:
>The cops measure speed with a radar. What they do is use a non linear
>xtal to mix the incoming with the outgoing reference. It outputs RF from
>the mixing! The rest is easy. Laser doesn't work as well in bad
>conditions, but your target selection is more accurate.
In a RADAR system, at 10.525 or 24 GHz, it's a DIODE.
In an optical system I imagine what you describe could be used, although it
would have to be illuminated while the return was present, so there would be
a minimum lower limit on the optical pulse width.
Doppler radars work not by measuring the time of
the reflected pulse, but by the phase changes (beat)
between the continuous, (at least during the measuring cycle)
transmitted signal and the reflected signal.
The transmitter signal needs to be present during the
whole of the measuring cycle
Even though the signal is of a high freq. the measure cycle
is very long (relatively)
eg: for 24Gig microwave doppler, resolution 1Kph
24Gig = wave length approx 10mm,
1Km per hour = 280mm per second
Times two for round trip (tx distance + reflected)
= 480mm per sec
So if you're measuring the beat freq and not the beat phase shift
(more difficult) the beats will come at 48 hz at the lowest speed
(1kph)
Ultrasonic at 40K wave length approx 20,000mm per second
Thats .024Hz beat frequency or 42 seconds per cycle
I think that rules out ultrasonic doppler for land
based vehicles at least with the current speed
limit laws.
I have some experience with airborne Doppler navigation RADARs from a
`mumble' USAF bomber project. It was a standard JANUS (ie: the Greek God)
system with dual lobes looking fore and aft. Two beam pairs to the
front-right, two to the front-left, same with the rear. It simply looked at
the difference in frequency between the front and rear to derive ground
speed. The difference in left to right gave us cross-track info and the
antenna assembly was slewed to null-out the difference. The dual pairs in
each quadrant was used to correct altitude ambiguity. I think the Klystron
(not a `Maggie' in this case) was tuned to around 24 GHz. I do remember that
it was small, red, and very difficult to reach for tuning and testing which
meant that one often got a 30-40KV high frequency `wake-up call'... That was
a long time ago and I'll be 50 years old tomorrow and you know what they
say; "The first thing to go..." ;-)
- Tom
At 11:18 AM 1/22/00 -0500, Wagner Lipnharski wrote: {Quote hidden}
>Thanks Michael, this is not a common thread really. We don't deal with
>light doppler everyday. :)
>
>The doppler can be identified using other chips, for example the use of
>PLL's for audio and higher frequencies. The PLL in this case would
>generate a DC voltage proportional to the "delta frequency*. I am not
>sure about light frequencies, one thing that comes to mind is to use the
>refraction characteristics of different frequencies, not sure if the
>magnitude will be enough in this case. The PIC function here would be to
>account the final data processing and control.
>
>Laser Police Radar uses to measure the reflective time in successive
>timed readings, so establishing the distance between the radar and the
>vehicle in each measurements, the timed samples tells you the speed.
>All the burst of readings happens in few milliseconds, so I really can't
>understand what the Laser Radar Detector wants to *try* to tell you when
>it detects a laser gun... perhaps "you are busted". Laser Radar uses the
>same measurement principle of the RF Radar, the advantage is that a
>laser is selective and "illuminate" a single spot in a single car, very
>difficult to the car 1 mile behind identify what is happening up front
>(what happening with RF Radar that spreads RF all along).
>
>About the Ground SpeedOmeter: Considering the gross measurement of
>298,000 km/s (light speed), it is around of 3.3557 nanoseconds per
>meter, so this would be the time the light pulse would take to leave the
>device below the car, hit the ground and return to the sensor. An
>average car's delta speed is from zero to 180 km/h (let's hope), so it
>would change the relative light speed from 298,000 to 298.000.05 km/s
>when the car is at 180km/h. The delta time to light travel this 1 meter
>would be 0.00001%, so impractical to be measured by "non advanced"
>technology. This values would change significatively if thinking in
>lower magnitude of frequencies, like ultrasound.
>
>Any actual experiences? weblinks?
>
>Wagner
------------------------------------------------------------------------
Tom Handley
New Age Communications
Since '75 before "New Age" and no one around here is waiting for UFOs ;-)
Tom, that reminds me of a RADAR MTI (Moving Target Indicator) mode. I was
once on a project with TI as the prime contractor, where we were doing
pioneering work in the area. It use to be highly classified... Basically, we
took advantage of the fact that a moving target (a truck on the Ho Chi Minh
trail in our case...) would leave a "Doppler Phase History" on the received
pulse. The RADAR was multi-mode with ground mapping, terrain following,
terrain avoidance, and other functions so we had to live within those
constraints as far as frequency, PRF, dish size and shape, etc. It required
a great deal of signal processing which was mostly done in analog with some
hard-coded math done in, what was then 5400 series logic from TI. It amazes
me to think what I can do now for under $100 not counting the microwave
components and mil-spec power supplies...
- Tom
At 11:35 AM 1/22/00 -0500, Thomas McGahee wrote: {Quote hidden}
>Actually you can make it even simpler than that. You begin with
>a known frequency as your transmit frequency. Nothing non-linear
>here. You send little bursts of this frequency out. While the
>transmitter is quiet, the bounced and doppler-shifted echo is
>amplified and mixed with the original output frequency. The
>resulting low frequency "beat" frequency represents the
>doppler-shifted component only. (Ignore the higher frequency beat).
>Once you know the original frequency and the low beat
>frequency, it just takes a little bit of math to calculate the
>relative speed of the object producing the echo.
>
>If you can isolate the receiver from the transmitter, then you
>can forget about the bursts and just send out a continuous
>signal from the transmitter. In general the receiver needs to be
>high gain and tuned to respond to a very narrow frequency band.
>
>Some laser type doppler-shift device works along the same principles.
>It is very easy to maintain isolation between transmitter and
>receiver since the laser beam is very narrow and does not diverge much.
>So these are almost always run in continuous mode. A form of
>interferometer mixes a small sample of the outgoing reference
>signal, and 100% of the received signal (which is usually
>quite small). The "beat" frequency shows up as light and dark
>bands. By scanning these fluctuations optically an electrical
>signal can be produced that can then be measured with a simple
>counter to reveal the optical "beat" frequency. A little
>math and the velocity is determined.
>
>Fr. Tom McGahee
------------------------------------------------------------------------
Tom Handley
New Age Communications
Since '75 before "New Age" and no one around here is waiting for UFOs ;-)
USTR.NET>
