That is beautiful Beau!· I need to look a little more closely at the patent but your idea may have taken us way outside of it.· Anyway this is for me, not for sale, so the whole patent thing really doesn't apply.
Beau said... 1)·Turn on a laser while pulling the trigger and at the same time charge a capacitor to a very specific voltage. 2) Discharge the capacitor for a specific amount of time at a specific rate and note the voltage 3) Recharge the capacitor back to the very specific voltage used in step #1 and turn the laser off 4) While the receiver still "sees" the laser, discharge the capacitor at the same rate as in step #2 and note the voltage when the receiver·no longer· "sees" the laser
Your Idea makes good use of time while the laser fully excites too.· Our laser pulses would still be fast and the accuracy would be unheard of.
Could we still use an RCtime method or are we into the area of 16 bit A/D's?
We could charge the capacitor to a high accuracy 3.3v regulated source and discharge to our logic threshold and count the time.·· Then recharge, fire laser, discharge until laser is not seen, and rctime it.· Our initial RCTIME measurement would give us calibration for not only the RC circuit but any temperature affects the propeller may be experiencing.· Then the actual-factual measurement would be acurate and if we discharge at a slower rate we would expand our accuracy.
······
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
Post Edited (Sniper King) : 7/2/2008 8:09:10 AM GMT
Maybe it is not an issue because of the clock speed of the prop - but isn't the change in resistor used to "stretch' the pulse?
maybe I am reading it wrong.
Is it possible to use a single digitally variable resistor for the RC circuit?
maybe not fast enough...
quick question said...
Maybe it is not an issue because of the clock speed of the prop - but isn't the change in resistor used to "stretch' the pulse?
maybe I am reading it wrong.
· Exactly.· look at my schematic.· Super basic 5 min eagle drawing but demonstrates the idea.· Using Beau idea, the circuit still works.·
1) Charge C1· from p26·and fire LASER
2) Discharge c1 through· r4 and count time for calibration
3) recharge c1 from pin 26
·I think every minute or so I am going to do calibration cycles to calibrate both discharge times using R1 then R4 and store that data for actual firings. then I can skip 1,2,3 and start with...Turn on LASER and Charge C1 from pin 26
4) turn off laser and discharge C1 through r4
5) when detector stops seeing the LASER Stop the discharge.
6) NOW, finish discharging the laser though R1 and count time.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
At this point I would use a high precision ADC instead of measuring RCtime simply for its increased accuracy and stability.· Also, events should be gated in a way that the system clock will not throttle or have an influence on your readings.· For Example: Don't rely on software to make the decision to keep discharging the capacitor as the detector still "sees" the laser, but instead have external "glue-logic" that does this as an independent clock-free process.· ·
It is also imperative to keep paths minimal, and similar.· For Example: The only real difference between steps #2 and #4 should be how you gate the signal using external "glue-logic".· In each case, the discharge path used should be identical. · "...but isn't the change in resistor used to "stretch' the pulse?" By sending multiple pulses...· after Step #4, briefly turn on the laser and disable discharging.· Following that, repeat step numbers·#3·and #4·without recharging the capacitor.· Doing this allows you to·"fold" your distance for an increased accuracy at the cost of it taking longer to obtain a result. · · ·
For reference:
1) Turn on a laser while pulling the trigger and at the same time charge a capacitor to a very specific voltage.
2) Discharge the capacitor for a specific amount of time at a specific rate and note the voltage
3) Recharge the capacitor back to the very specific voltage used in step #1 and turn the laser off
4) While the receiver still "sees" the laser, discharge the capacitor at the same rate as in step #2 and note the voltage when the receiver no longer "sees" the laser
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 7/2/2008 5:08:34 PM GMT
In my circuit What do you think the accuracy of the system would be?· It would seem, granted, that clock cycles within the propeller would happen but I am sure that the this is fairly stable and can be subtracted from the actual time of flight reading and even if we are between cycles during an event, 4ns is a little over a meter error.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
Post Edited (Sniper King) : 7/2/2008 4:31:19 PM GMT
The Propeller's clock cycles at maximum speed (80MHz) are 12.5ns. The built-in cog counters have a resolution of one clock cycle, but the instructions take a minimum of 4 clock cycles (50ns), so you really have to rely on external circuitry for resolutions on the order of nanoseconds.
Ok, I am getting outside my circuit knowledge.· Anybody out there able to creat this circuit?· I know it is alot to ask but the learning curve is huge!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
I don't have a schematic, per say, just a fundamental approach.
In a few searches, I have found several instances where the exact principles I describe are used. The main difference being a Xenon photoflash tube rather than a laser, but the basic physics would still apply. Flash the tube, and while the detector sees the tube-flash, discharge a cap. The Xenon tube can be flashed several times to effectively "fold" the distance to increase resolution. Calibration of course would still be required, and again temperature is going to be your worst enemy. Apply temperature compensation within your circuit design where you think it may be necessary. Compensating for temperature is more of a balancing act and a way of thinking than anything else. In practice, it is applied to a circuit in such a way that any effects from temperature are cancelled out. Sometimes the technique is a differential approach, and other times it is a same-same approach where all critical signals are offset equally by some amount related to temperature. In addition to temperature, there will also be propagation delays from the circuitry that must be accounted for. Many of the same principles applied to temperature compensation will also work with propagation delays, but here the goal is often to make sure that parallel signals arrive at their destination(s) at the same time, so you purposefully "add" delay to a faster signal with less initial propagation delay so that it has time to meet up with the slower signals. Some of these issues however can also be corrected in software.
This weekend I might try something with a Xenon, just to see what kind of values I can get ... The Xenon would be easier to setup as far as detector alignment and should have very similar results to using the laser assuming you have proper alignment with the laser and detector configuration.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 7/3/2008 6:28:08 AM GMT
Here is a proposed test setup using a XENON flash tube and two photo-diodes.
This test is more so to determine proper balance and response time. · When the Xenon is at one sensor or the other (shown in orange)·you should see the greatest distance in terms of time of light-flight, and since light travels at approximately 1ns per foot, then 50 feet should show a pulse width on the scope of approximately 50ns.· When the Xenon is equidistant from both sensors (shown in green), the time of light-flight should be the same and a very small blip if anything should show up on the scope.· Since all other wire routing is mirrored until you get to the XOR (literally located dead center between the two sensors) any propagation delay that exists in the wiring is·canceled out.· The next step would be to take this pulse to either charge or discharge a cap and read it with a precision ADC to·resolve distance in the sub 12.5 ns time domain·limiting·us from direct pulse reads using the Propeller.
The second image... Light Test·2.jpg ...Is how you might incorporate the test setup from the first image into a single unit.· What isn't shown is the amplifier circuitry needed at each Photo diode.· In order to optimize this,·I need to know what would be the maximum frequency you could modulate your LASER at Sniper King?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 7/4/2008 5:52:44 AM GMT
the phototransistor I have chosen for my testing is the 38kHz Infrared (IR) Receiver Module
Model: 276-640 | Catalog #: 276-640 This is due to its built in 38KHZ notch filter to really filter out background noise and make the reception definate.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
This is not a phototransistor. It requires 38KHz modulated light and takes several cycles to react. There is no way it would be fast enough for your needs. Even a standard phototransistor is not fast enough. You need a fast photodiode with a high speed amplifier / trigger to give you a clean logic level from the sharp edge of the incoming light pulse.
You are soooo right! I just did some testing of that phototransistor and it take a long time to react and the Delay between sourse of and detector reaction jumps around alot so, any suggestions here would be helpful.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
At this point what is the point of modulation? I can see us having verify for sure that we acually see light. By pulsing a few times before measuring to make sure it isn't just ambient light. I could modulate pretty fast though. What do you have in mind?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
"At this point what is the point of modulation?" - I was still considering the possibility of using phase discrimination to determine distance, by varying the modulated beam at specific fixed frequencies, you can then use the phase relationship of the signals to determine distance as well (<- without charging/discharging a cap) and you could use the same hardware setup currently proposed.
Tomorrow night I'll do some strobe tests, since tonight is the 4th of July. · Basically to re-iterate what Phil was describing in an earlier post, the method he initially describes s probably going to be the best way to go, since it will take care of a lot of potential problems.· The compensation approach here is basically same-same where each sensor goes through its own identical circuitry...· if the detection circuitry is kept reasonably close (but not too close... we don't want crosstalk)·and stepped on the circuit board... NOT mirrored! ... then ambient temperature will be tracked, and both circuits will behave the same.· Propagation delays will also be tracked, therefore canceled out, so that any measured delay will be delay in the actual signal phase that we want to measure.· This method also allows the use of sensors that only need to be fast enough to keep up with changes in the·carrier frequency. · ·
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 7/4/2008 10:26:40 PM GMT
Ill probably be using the propeller to modulate the laser.··A pulse program in its own cog and I was able to get over 120K That I could measure with the USB Ocilliscope
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
"Ill probably be using the propeller to modulate the laser"... Another subject in and of itself, but make sure that the frequency that you select is a multiple of 12.5ns or there will be a slight amount of "jitter".· 38kHz or 120kHz are not jitter free, because they don't fall within the 12.5ns multiple.· Some acceptable candidates are...
You could also change the clock crystal to something other than 5Mhz (x16 = 80MHz).· A 6MHz (x16 = 96MHz) crystal for example would
allow these jitter free frequencies to be used.
You will want to look at an avalanche photodiode for this. Go to hamamatsu.com , and look for adp diodes, they have modules at 1 gig and 100 meg. Pretty good info to check out there.
kelvin james, · Thanks I will keep that in mind. · Sniper King, · Fortunately my wife works at a church, and I was able to use (with permission from the pastor) the wide open consistent space of the sanctuary to setup this experiment.· Some tests that I·performed with the strobe tonight worked, but I really needed to tweak the sensitivity of the sensors to get them to properly behave.· This is mostly because the energy level of my strobe at 50ft and even at 25ft is very weak and not modulated.· The strobe basically has a characteristic of a fixed 100uS pulse upon which the repetition rate can be adjusted.· Because the pulse is not modulated (no carrier), I ran the risk of measuring light intensity rather than speed, but I was able to visualize a 25ns pulse at each end 1/4 the way towards the center from each sensor, and a 50ns pulse directly at one sensor or the other.··...and as predicted there was no pulse in the middle. (no perceivable pulse).· Also, because I could not find a decent XOR that had equally balanced rise and fall times based on various dependencies on the combination·of the inputs I·used two 330 Ohm resistors to create·a summation stage·before going to the Scope omitting the XOR all together. · · All this proves, is that the sensors and arrangement can work, but the more I think about it, the more I'm absolutely convinced that you should go completely digital and not rely on the charging or discharging of a capacitor.· · · · Thinking about it further... the 4th of July was actually very helpful... The lag of the sound of the explosion verses the visual of the explosion....Anyway, at the distances that you want to accomplish, a totally closed system might work where all you need to do was to look at a frequency counter to determine distance. · (See attached image) · At 1 mile the·propagation delay of light is·about 5.368·micro seconds... double that for a round trip time of 10.736 micro seconds.· If you look at this in terms of a standard schmitt trigger style oscillator, if you had·a propagation delay of·10.736 microseconds, you'd be oscillating right at about·46.57 KHz.... piece of cake for a 74HC14 schmitt trigger.· Even if you include the propagation delay of the inverter plus the laser, plus the detector, you’re only going to be contributing nano seconds, and because it would be a relative constant it could easily be corrected for. · Suppose you were measuring 1 mile plus 1·meter (5283 feet)·instead of just the 1 mile above....·· At 1 mile + 1 meter, the propagation delay would be about 5.371 micro seconds... double that and you get 10.742 micro seconds.· That doesn't sound like much, and it's not but your oscillator is now running at·46.545 KHz instead of 46.57 KHz. · If you sample or count pulses for·1 second you have a difference or resolution of about·25 pulses per meter (46570 -·46545 = 25) ... The longer you take a sample the better discrepancy you will have...·5 seconds = a resolution of 125 pulses per meter · In situations where you would want a·closer measurement, an optical delay line could be used where the laser beam is folded several times before it's allowed to leave.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 7/7/2008 6:30:47 AM GMT
Could I get away with a two telescope desighn for this?· looking at your schematic, we could manually fire the laser until the detector "sees" the return then let the resulting occillator kick in until we can get a sample of the frequency.· Is this what you are thinking here?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
Sniper King, "Could I get away with a two telescope design for this?" - I think you have a better handle on the optics than I do.... I see one need to focus the beam smaller to compensate for atmospheric lensing that normally causes the beam to grow larger.· Another need is to re-focus (from the target) all the available energy you can from the reflection of the laser. "looking at your schematic, we could manually fire the laser until the detector "sees" the return then let the resulting oscillator kick in until we can get a sample of the frequency.· Is this what you are thinking here?" - Absolutely correct, provided you can focus enough of the beam energy from the target into your sensor.
I am going to try a variant of the Closed Loop test tonight (see attached image) that would require minimal beam focusing and alignment.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Beau Schwabe
IC Layout Engineer
Parallax, Inc.
I suspect if your receiver is sensitive enough, you might get false measurements due to differences in humidity!
(either the heat from something in the distance or exhaust from a chimney)
At the same time, those things could totally scatter your signal.
I work on radars and would make the assumption that light is just the same energy at different wavelengths (minus the Rf modulation/etc). To be brief, as a radar signal passes through a weather target, it actually changes phase. I don't know the proper formula to quote, but I'm sure there is one!
Is the laser beamwidth static? I assume it would widen as it transits.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ <FONT>Steve
What's the best thing to do in a lightning storm? "take a one iron out the bag and hold it straight up above your head, even God cant hit a one iron!"
Lee Travino after the second time being hit by lightning!
steve_b, · More than likely you are exactly right, which is probably why you don't see laser range finders going that far.· 1 mile is beyond my scope to actually test with, before·spending a few dollars more than I want to. · Sniper King · Tonight I did two tests at different lengths, but because my mirrors weren’t that great, I had·some unwanted beam splitting at the glass mirror surface and at the mirror backing.· I did not have any flat polished surface mirrors to work with. · The largest distance I was able to setup was about 880 feet (see image 880Feet.jpg)· With a similar setup I also did the test at 440 feet.· Both tests were able to oscillate using the feedback method at about 38 kHz (440feet) and 37 kHz (880feet)· · The·propagation delay within the system (Includes Laser, Oscillator, and Detector) was about 12.6us this was measured by completely closing the loop and measuring a frequency of about 39.7kHz · A distance of 440 feet adds a 447ns propagation delay resulting in a calculated frequency of 38.4405 A distance of 880 feet adds a 894ns propagation delay resulting in a calculated frequency of 37.1615 · The calculated·numbers coincide with what I was able to measure, but to be honest, there was enough fluctuation beyond the decimal that I could not get a solid lock on the frequency. · · · While this was a fun educational experience, I did have an ulterior motive with the possibility of contributing·something to our existing robotic sensor portfolio.· An affordable "Optical Echo Location"·would be a nice addition, and while the Strobe method has some merit, it still needs much work. ·
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Beau Schwabe
IC Layout Engineer
Parallax, Inc.
I do find this interesting and wish I wasn't so busy when this thread started....and still wish I wasn't busy at the moment to play too!
I hope you continue looking at these things...
At one point I had been looking at using fiber optic hardware to try and clock/gate light signals. It was only an idea that I scratched on some paper....but with the proper lens one should be able to use a fiber optic detector as well.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ <FONT>Steve
What's the best thing to do in a lightning storm? "take a one iron out the bag and hold it straight up above your head, even God cant hit a one iron!"
Lee Travino after the second time being hit by lightning!
"With your tests, what do you think the accuracy of a propeller chip measuring the frequency?" - The scope that I used was a Tektronix TDS 220 (100MHz 1GS/s).· I don't think that the Propeller would be a fair comparison.
"Do you think your fluctuations may have been due to the mirrors having·glass between the reflective surface and the front of the glass?" - It could be that, but more likely at that distance (880 feet), normal ground vibrations can be an issue... something I had not considered before.
One series of tests that I did not do that could have an improved signal determination is to modulate the laser at a fixed carrier frequency.· On each detector you would need identical circuitry... a tuned resonant LC at the carrier·frequency and a schmitt input of some kind to convert the received carrier into a digital signal.· From there you look at the phase difference as it has been suggested earlier in this thread.· By XORing the signals and looking at it as a PWM signal 0% would mean that both sensors are balanced ... 100% would mean that the sensors are 180 Deg out of Phase with one another.· The rule would be that whatever carrier frequency that you are modulating at, you could only measure at half the distance of the entire wavelength.· Take for example 50kHz... it has a wavelength of 19,671 feet ... The maximum distance that you could measure before your numbers started counting backwards would be 9,835.5 feet (1.86 miles).
Detecting the carrier frequency becomes another issue... higher frequencies are much easier to "tune" than a lower frequencies.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 7/9/2008 9:24:27 PM GMT
One of the problems you may face with a dual scope range-finder, is beam alignment (not just beam-collimation) - with two scopes, you will run into a complication where the transmitted beam may have a fractionally different angle that the receiver's POV - this is "Parallax Error" (how great is it that it bears this forum's company name?·).· This may make it hard for the receiver to "see" the lased point on the target.· A way to think of this is if you have a point-and-shoot camera with a separate view-finder from the capture lens.· What you see in the viewfinder is not exactly what you will see on the captured content (it will be slightly shifted).
Two eyed humans compensate for this problem by crossing our eyes the closer the "target" gets, and paralleling our eyes the further out the target is.· With two scopes, you will need to build in some function to correct this (this could be a geared-down, high-precision stepper that "crosses" the two scopes together unit it detects a good signal at range - this could also help with quick estimation based on the stepper position).
However - that method is complicated and cumbersome (and let's not forget it too will need to be calibrated, which could be hard if you "bang around" your gear).· I would recommend an optical cube-beamsplitter, at the eye-piece of the scope.· This will ensure that anything that is "viewable" in the scope's reticle will also be viewable on the return path.· The big issue with the setup is probably going to be your laser's transmitting alignment - since with optics, the angle at entry determine's the angle of departure.· You need to get the pitch and yaw (tilt/pan respectively) of the laser aligned with the optical chain, as well as the horizontal and vertical centering of the laster just right for it to "hit" your target.
If you manage to go with a cube beamsplitter, I'd recommend two - one for the range finding (closest to the scope's optical chain), and one for the operator's target aquisition.· If you are using a high-powered IR laser with this, do your eyes a favor and put in an IR filter to block the seering light from the return path...· a small camera aligned to the optical path is a safer bet (it removes the risk of high-intesity laser light from passing to the operator's eye).
One other item I though of while reading this thread - if you use more than one of these on a single range, you may have an issue with cross-talk.· You can address this by using a similar tactic as automotive radar is using, where not just a modulated carrier goes out, but a code pulse is modulated on that carrier, to ensure that the received pattern came from THIS transmitter.· This could be as easy as "firing" the laser at a coded interval for better noise rejection.
For instance if you used a 6-bit unit address (this could be whatever you want I presume), you would consider a "bit" to be a sample period of carrier activity (lets say 60uS of transmitted carrier "burst" per bit). If your 6-bit unit address was 011010, you should get a reply like:
Bit 5 Range (MSB) = No Carrier, distance out of range
Bit 4 Range = 1,435 feet
Bit 3 Range = 1,442 feet
Bit 2 Range = No Carrier, distance out of range
Bit 1 Range = 1,437 feet
Bit 0 Range (LSB) = No Carrier, distance out of range
You can tell this was a received carrier that was transmitted·by this transmitter.· Also, you can use "box-car-averaging" to narrow down the actual range (three values added then divided by three gives you a rough range of 1,438 feet - probably as close as you'll be able to manage...·O-Scopes like the Tektronix line do this to get better acccuracy because they only have 8-bit digitizers).
If you got a return signal like:
Bit 5 Range (MSB) = No Carrier, distance out of range
Bit 4 Range = 1,435 feet
Bit 3 Range = 1,449 feet
Bit 2 Range = 1,501 feet (Should not be getting carrier here!!!)
Bit 1 Range = 1,417 feet
Bit 0 Range (LSB) = No Carrier, distance out of range
You would know that there is probably a signal to noise problem, or another device operating on your carrier in your theater.· You could then make some decisions about the data you've received:
If you had very few "bit errors", where range data was received when it shouldn't have been, you can chose to omit that unexpected data
If you had a lot of bit errors, you can notify the operator that the range data is unreliable, and some other means should be used (if you are doing the two scope method, the parallax alignment method would be a good guestimate of the range based on the angle of deviation).
With the ammount of "errors" you receive you should be able to indicate the "confidence" of the range finder that it has valid data from its transmitter.
By doing a coded-pulse-train for your laser burst, you should give you the side effect of being able to do longer ranges without concern for out of phase alignment, since it gives you a much slower received data stream to align to the known transmitted data stream (not just a singe-cycle phase range - but a longer data phase range over many cycles).
my 2 cents...
-Tim
P.S. Oh - Beau, you mentioned a robot laser "ping" type aplication...· your are basically talking about "lidar", or "laser radar".· Take a look at the SICK Laser Measurment Systems, their units are essentially a standard apperatus on the DARPA Grand-Challenge bots...· Once you have a range finder built, it just takes a few aligned/motorized·mirrors to do X and Y grid measurements and mapping... -T
P.P.S. Sniper King, if you go through the motions of having the coded signal - guess what, you now have a line of sight communication link too...· Think of it, take a keyboard entered string of data, blast it out through the laser at an agreed upon target surface, and any other range-finder on your carrier can read out the transmitted burst. -T
Post Edited (GreyBox Tim) : 7/11/2008 1:53:48 AM GMT
I think it's possible that Beau's measurement fluctuations may have been due (in part) to variations in the refractive index of air with temperature. That is, air convection currents in the path causing "shimmer". I could not quickly find from Googling how big these variations may be.
Taking up Beaus idea of modulating the laser at a fixed carrier frequency, instead of introducing a resonant tank at the Rx end why not use a quadrature demodulator. This would:
1. Provide a signal averaging function to help with the detection of small return signals.
2. Provide a means to measure the phase of the returned signal.
OK, the time has come to test all this stuff with a laser, telescope and avalanche photo diode that I am getting ·from Perkin-Elmer (c30737e).· I am getting a small telescope and since we are working in the near infrared, we can use a refracting telecope.· i have worked with thermal so long I almost forgot that it is quite a difference in wavelength!·
Here is the test.· Tell me if I am dumb...
250mw infrared laser at 940nm pointed at nearby cliff face of the majestic Huachuca Mountains.· I am not going to use a telescope yet with the laser because the range is only about 3 miles.· The Photodiode will be on the telescope at the point of focus (not the focal point).· Using a camcorder i will be able to see the spot on the mountain.· Then pointing the telescope at the same spot I should see the spot with the photodiode.
Now the fun part and this is where I may be dumb.· I am going to use my trusty Parallax O-scope and attach channel 1 to the + power· on the laser.· Channel 2 on the photodiode.· I will use the propeller to modulate 50khz on the laser.· I should see the 50khz on the direct connect to the laser and I should see the returned laser pulses on the photodiode. I should also see a wave shift on the returned signal.· If I am wrong somebody fix me.
Michael King
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
Comments
Your Idea makes good use of time while the laser fully excites too.· Our laser pulses would still be fast and the accuracy would be unheard of.
Could we still use an RCtime method or are we into the area of 16 bit A/D's?
We could charge the capacitor to a high accuracy 3.3v regulated source and discharge to our logic threshold and count the time.·· Then recharge, fire laser, discharge until laser is not seen, and rctime it.· Our initial RCTIME measurement would give us calibration for not only the RC circuit but any temperature affects the propeller may be experiencing.· Then the actual-factual measurement would be acurate and if we discharge at a slower rate we would expand our accuracy.
······
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
Post Edited (Sniper King) : 7/2/2008 8:09:10 AM GMT
Maybe it is not an issue because of the clock speed of the prop - but isn't the change in resistor used to "stretch' the pulse?
maybe I am reading it wrong.
Is it possible to use a single digitally variable resistor for the RC circuit?
maybe not fast enough...
1) Charge C1· from p26·and fire LASER
2) Discharge c1 through· r4 and count time for calibration
3) recharge c1 from pin 26
·I think every minute or so I am going to do calibration cycles to calibrate both discharge times using R1 then R4 and store that data for actual firings. then I can skip 1,2,3 and start with...Turn on LASER and Charge C1 from pin 26
4) turn off laser and discharge C1 through r4
5) when detector stops seeing the LASER Stop the discharge.
6) NOW, finish discharging the laser though R1 and count time.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
·
It is also imperative to keep paths minimal, and similar.· For Example: The only real difference between steps #2 and #4 should be how you gate the signal using external "glue-logic".· In each case, the discharge path used should be identical.
·
"...but isn't the change in resistor used to "stretch' the pulse?"
By sending multiple pulses...· after Step #4, briefly turn on the laser and disable discharging.· Following that, repeat step numbers·#3·and #4·without recharging the capacitor.· Doing this allows you to·"fold" your distance for an increased accuracy at the cost of it taking longer to obtain a result.
·
·
·
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 7/2/2008 5:08:34 PM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
Post Edited (Sniper King) : 7/2/2008 4:31:19 PM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
I don't have a schematic, per say, just a fundamental approach.
In a few searches, I have found several instances where the exact principles I describe are used. The main difference being a Xenon photoflash tube rather than a laser, but the basic physics would still apply. Flash the tube, and while the detector sees the tube-flash, discharge a cap. The Xenon tube can be flashed several times to effectively "fold" the distance to increase resolution. Calibration of course would still be required, and again temperature is going to be your worst enemy. Apply temperature compensation within your circuit design where you think it may be necessary. Compensating for temperature is more of a balancing act and a way of thinking than anything else. In practice, it is applied to a circuit in such a way that any effects from temperature are cancelled out. Sometimes the technique is a differential approach, and other times it is a same-same approach where all critical signals are offset equally by some amount related to temperature. In addition to temperature, there will also be propagation delays from the circuitry that must be accounted for. Many of the same principles applied to temperature compensation will also work with propagation delays, but here the goal is often to make sure that parallel signals arrive at their destination(s) at the same time, so you purposefully "add" delay to a faster signal with less initial propagation delay so that it has time to meet up with the slower signals. Some of these issues however can also be corrected in software.
This weekend I might try something with a Xenon, just to see what kind of values I can get ... The Xenon would be easier to setup as far as detector alignment and should have very similar results to using the laser assuming you have proper alignment with the laser and detector configuration.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 7/3/2008 6:28:08 AM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
Here is a proposed test setup using a XENON flash tube and two photo-diodes.
This test is more so to determine proper balance and response time.
·
When the Xenon is at one sensor or the other (shown in orange)·you should see the greatest distance in terms of time of light-flight, and since light travels at approximately 1ns per foot, then 50 feet should show a pulse width on the scope of approximately 50ns.· When the Xenon is equidistant from both sensors (shown in green), the time of light-flight should be the same and a very small blip if anything should show up on the scope.· Since all other wire routing is mirrored until you get to the XOR (literally located dead center between the two sensors) any propagation delay that exists in the wiring is·canceled out.· The next step would be to take this pulse to either charge or discharge a cap and read it with a precision ADC to·resolve distance in the sub 12.5 ns time domain·limiting·us from direct pulse reads using the Propeller.
The second image... Light Test·2.jpg ...Is how you might incorporate the test setup from the first image into a single unit.· What isn't shown is the amplifier circuitry needed at each Photo diode.· In order to optimize this,·I need to know what would be the maximum frequency you could modulate your LASER at Sniper King?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 7/4/2008 5:52:44 AM GMT
Model: 276-640 | Catalog #: 276-640 This is due to its built in 38KHZ notch filter to really filter out background noise and make the reception definate.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
"At this point what is the point of modulation?" - I was still considering the possibility of using phase discrimination to determine distance, by varying the modulated beam at specific fixed frequencies, you can then use the phase relationship of the signals to determine distance as well (<- without charging/discharging a cap) and you could use the same hardware setup currently proposed.
Tomorrow night I'll do some strobe tests, since tonight is the 4th of July.
·
Basically to re-iterate what Phil was describing in an earlier post, the method he initially describes s probably going to be the best way to go, since it will take care of a lot of potential problems.· The compensation approach here is basically same-same where each sensor goes through its own identical circuitry...· if the detection circuitry is kept reasonably close (but not too close... we don't want crosstalk)·and stepped on the circuit board... NOT mirrored! ... then ambient temperature will be tracked, and both circuits will behave the same.· Propagation delays will also be tracked, therefore canceled out, so that any measured delay will be delay in the actual signal phase that we want to measure.· This method also allows the use of sensors that only need to be fast enough to keep up with changes in the·carrier frequency.
·
·
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 7/4/2008 10:26:40 PM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
"Ill probably be using the propeller to modulate the laser"... Another subject in and of itself, but make sure that the frequency that you select is a multiple of 12.5ns or there will be a slight amount of "jitter".· 38kHz or 120kHz are not jitter free, because they don't fall within the 12.5ns multiple.· Some acceptable candidates are...
31.25kHz
32kHz
40kHz
50kHz
62.5kHz
64kHz
78.125kHz
80kHz
100kHz
125kHz
128kHz
You could also change the clock crystal to something other than 5Mhz (x16 = 80MHz).· A 6MHz (x16 = 96MHz) crystal for example would
allow these jitter free frequencies to be used.
37.5kHz
38.4kHz
40kHz
46.875kHz
48kHz
50kHz
51.2kHz
60kHz
75kHz
76.8kHz
80kHz
93.75kHz
96kHz
100kHz
120kHz
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
·
Thanks I will keep that in mind.
·
Sniper King,
·
Fortunately my wife works at a church, and I was able to use (with permission from the pastor) the wide open consistent space of the sanctuary to setup this experiment.· Some tests that I·performed with the strobe tonight worked, but I really needed to tweak the sensitivity of the sensors to get them to properly behave.· This is mostly because the energy level of my strobe at 50ft and even at 25ft is very weak and not modulated.· The strobe basically has a characteristic of a fixed 100uS pulse upon which the repetition rate can be adjusted.· Because the pulse is not modulated (no carrier), I ran the risk of measuring light intensity rather than speed, but I was able to visualize a 25ns pulse at each end 1/4 the way towards the center from each sensor, and a 50ns pulse directly at one sensor or the other.··...and as predicted there was no pulse in the middle. (no perceivable pulse).· Also, because I could not find a decent XOR that had equally balanced rise and fall times based on various dependencies on the combination·of the inputs I·used two 330 Ohm resistors to create·a summation stage·before going to the Scope omitting the XOR all together.
·
·
All this proves, is that the sensors and arrangement can work, but the more I think about it, the more I'm absolutely convinced that you should go completely digital and not rely on the charging or discharging of a capacitor.·
·
·
·
Thinking about it further... the 4th of July was actually very helpful... The lag of the sound of the explosion verses the visual of the explosion....Anyway, at the distances that you want to accomplish, a totally closed system might work where all you need to do was to look at a frequency counter to determine distance.
·
(See attached image)
·
At 1 mile the·propagation delay of light is·about 5.368·micro seconds... double that for a round trip time of 10.736 micro seconds.· If you look at this in terms of a standard schmitt trigger style oscillator, if you had·a propagation delay of·10.736 microseconds, you'd be oscillating right at about·46.57 KHz.... piece of cake for a 74HC14 schmitt trigger.· Even if you include the propagation delay of the inverter plus the laser, plus the detector, you’re only going to be contributing nano seconds, and because it would be a relative constant it could easily be corrected for.
·
Suppose you were measuring 1 mile plus 1·meter (5283 feet)·instead of just the 1 mile above....·· At 1 mile + 1 meter, the propagation delay would be about 5.371 micro seconds... double that and you get 10.742 micro seconds.· That doesn't sound like much, and it's not but your oscillator is now running at·46.545 KHz instead of 46.57 KHz.
·
If you sample or count pulses for·1 second you have a difference or resolution of about·25 pulses per meter (46570 -·46545 = 25) ... The longer you take a sample the better discrepancy you will have...·5 seconds = a resolution of 125 pulses per meter
·
In situations where you would want a·closer measurement, an optical delay line could be used where the laser beam is folded several times before it's allowed to leave.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 7/7/2008 6:30:47 AM GMT
Could I get away with a two telescope desighn for this?· looking at your schematic, we could manually fire the laser until the detector "sees" the return then let the resulting occillator kick in until we can get a sample of the frequency.· Is this what you are thinking here?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
"Could I get away with a two telescope design for this?" - I think you have a better handle on the optics than I do.... I see one need to focus the beam smaller to compensate for atmospheric lensing that normally causes the beam to grow larger.· Another need is to re-focus (from the target) all the available energy you can from the reflection of the laser.
"looking at your schematic, we could manually fire the laser until the detector "sees" the return then let the resulting oscillator kick in until we can get a sample of the frequency.· Is this what you are thinking here?" - Absolutely correct, provided you can focus enough of the beam energy from the target into your sensor.
I am going to try a variant of the Closed Loop test tonight (see attached image) that would require minimal beam focusing and alignment.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
(either the heat from something in the distance or exhaust from a chimney)
At the same time, those things could totally scatter your signal.
I work on radars and would make the assumption that light is just the same energy at different wavelengths (minus the Rf modulation/etc). To be brief, as a radar signal passes through a weather target, it actually changes phase. I don't know the proper formula to quote, but I'm sure there is one!
Is the laser beamwidth static? I assume it would widen as it transits.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
<FONT>Steve
What's the best thing to do in a lightning storm? "take a one iron out the bag and hold it straight up above your head, even God cant hit a one iron!"
Lee Travino after the second time being hit by lightning!
·
More than likely you are exactly right, which is probably why you don't see laser range finders going that far.· 1 mile is beyond my scope to actually test with, before·spending a few dollars more than I want to.
·
Sniper King
·
Tonight I did two tests at different lengths, but because my mirrors weren’t that great, I had·some unwanted beam splitting at the glass mirror surface and at the mirror backing.· I did not have any flat polished surface mirrors to work with.
·
The largest distance I was able to setup was about 880 feet (see image 880Feet.jpg)· With a similar setup I also did the test at 440 feet.· Both tests were able to oscillate using the feedback method at about 38 kHz (440feet) and 37 kHz (880feet)·
·
The·propagation delay within the system (Includes Laser, Oscillator, and Detector) was about 12.6us this was measured by completely closing the loop and measuring a frequency of about 39.7kHz
·
A distance of 440 feet adds a 447ns propagation delay resulting in a calculated frequency of 38.4405
A distance of 880 feet adds a 894ns propagation delay resulting in a calculated frequency of 37.1615
·
The calculated·numbers coincide with what I was able to measure, but to be honest, there was enough fluctuation beyond the decimal that I could not get a solid lock on the frequency.
·
·
·
While this was a fun educational experience, I did have an ulterior motive with the possibility of contributing·something to our existing robotic sensor portfolio.· An affordable "Optical Echo Location"·would be a nice addition, and while the Strobe method has some merit, it still needs much work.
·
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
I hope you continue looking at these things...
At one point I had been looking at using fiber optic hardware to try and clock/gate light signals. It was only an idea that I scratched on some paper....but with the proper lens one should be able to use a fiber optic detector as well.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
<FONT>Steve
What's the best thing to do in a lightning storm? "take a one iron out the bag and hold it straight up above your head, even God cant hit a one iron!"
Lee Travino after the second time being hit by lightning!
Do you think your fluctuations may have been due to the mirrors having·glass between the reflective surface and the front of the glass?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group
Post Edited (Sniper King) : 7/9/2008 8:52:20 PM GMT
"Do you think your fluctuations may have been due to the mirrors having·glass between the reflective surface and the front of the glass?" - It could be that, but more likely at that distance (880 feet), normal ground vibrations can be an issue... something I had not considered before.
One series of tests that I did not do that could have an improved signal determination is to modulate the laser at a fixed carrier frequency.· On each detector you would need identical circuitry... a tuned resonant LC at the carrier·frequency and a schmitt input of some kind to convert the received carrier into a digital signal.· From there you look at the phase difference as it has been suggested earlier in this thread.· By XORing the signals and looking at it as a PWM signal 0% would mean that both sensors are balanced ... 100% would mean that the sensors are 180 Deg out of Phase with one another.· The rule would be that whatever carrier frequency that you are modulating at, you could only measure at half the distance of the entire wavelength.· Take for example 50kHz... it has a wavelength of 19,671 feet ... The maximum distance that you could measure before your numbers started counting backwards would be 9,835.5 feet (1.86 miles).
Detecting the carrier frequency becomes another issue... higher frequencies are much easier to "tune" than a lower frequencies.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 7/9/2008 9:24:27 PM GMT
One of the problems you may face with a dual scope range-finder, is beam alignment (not just beam-collimation) - with two scopes, you will run into a complication where the transmitted beam may have a fractionally different angle that the receiver's POV - this is "Parallax Error" (how great is it that it bears this forum's company name?·
Two eyed humans compensate for this problem by crossing our eyes the closer the "target" gets, and paralleling our eyes the further out the target is.· With two scopes, you will need to build in some function to correct this (this could be a geared-down, high-precision stepper that "crosses" the two scopes together unit it detects a good signal at range - this could also help with quick estimation based on the stepper position).
However - that method is complicated and cumbersome (and let's not forget it too will need to be calibrated, which could be hard if you "bang around" your gear).· I would recommend an optical cube-beamsplitter, at the eye-piece of the scope.· This will ensure that anything that is "viewable" in the scope's reticle will also be viewable on the return path.· The big issue with the setup is probably going to be your laser's transmitting alignment - since with optics, the angle at entry determine's the angle of departure.· You need to get the pitch and yaw (tilt/pan respectively) of the laser aligned with the optical chain, as well as the horizontal and vertical centering of the laster just right for it to "hit" your target.
If you manage to go with a cube beamsplitter, I'd recommend two - one for the range finding (closest to the scope's optical chain), and one for the operator's target aquisition.· If you are using a high-powered IR laser with this, do your eyes a favor and put in an IR filter to block the seering light from the return path...· a small camera aligned to the optical path is a safer bet (it removes the risk of high-intesity laser light from passing to the operator's eye).
One other item I though of while reading this thread - if you use more than one of these on a single range, you may have an issue with cross-talk.· You can address this by using a similar tactic as automotive radar is using, where not just a modulated carrier goes out, but a code pulse is modulated on that carrier, to ensure that the received pattern came from THIS transmitter.· This could be as easy as "firing" the laser at a coded interval for better noise rejection.
For instance if you used a 6-bit unit address (this could be whatever you want I presume), you would consider a "bit" to be a sample period of carrier activity (lets say 60uS of transmitted carrier "burst" per bit). If your 6-bit unit address was 011010, you should get a reply like:
Bit 5 Range (MSB) = No Carrier, distance out of range
Bit 4 Range = 1,435 feet
Bit 3 Range = 1,442 feet
Bit 2 Range = No Carrier, distance out of range
Bit 1 Range = 1,437 feet
Bit 0 Range (LSB) = No Carrier, distance out of range
You can tell this was a received carrier that was transmitted·by this transmitter.· Also, you can use "box-car-averaging" to narrow down the actual range (three values added then divided by three gives you a rough range of 1,438 feet - probably as close as you'll be able to manage...·O-Scopes like the Tektronix line do this to get better acccuracy because they only have 8-bit digitizers).
If you got a return signal like:
Bit 5 Range (MSB) = No Carrier, distance out of range
Bit 4 Range = 1,435 feet
Bit 3 Range = 1,449 feet
Bit 2 Range = 1,501 feet (Should not be getting carrier here!!!)
Bit 1 Range = 1,417 feet
Bit 0 Range (LSB) = No Carrier, distance out of range
You would know that there is probably a signal to noise problem, or another device operating on your carrier in your theater.· You could then make some decisions about the data you've received:
With the ammount of "errors" you receive you should be able to indicate the "confidence" of the range finder that it has valid data from its transmitter.
By doing a coded-pulse-train for your laser burst, you should give you the side effect of being able to do longer ranges without concern for out of phase alignment, since it gives you a much slower received data stream to align to the known transmitted data stream (not just a singe-cycle phase range - but a longer data phase range over many cycles).
my 2 cents...
-Tim
P.S. Oh - Beau, you mentioned a robot laser "ping" type aplication...· your are basically talking about "lidar", or "laser radar".· Take a look at the SICK Laser Measurment Systems, their units are essentially a standard apperatus on the DARPA Grand-Challenge bots...· Once you have a range finder built, it just takes a few aligned/motorized·mirrors to do X and Y grid measurements and mapping... -T
P.P.S. Sniper King, if you go through the motions of having the coded signal - guess what, you now have a line of sight communication link too...
Post Edited (GreyBox Tim) : 7/11/2008 1:53:48 AM GMT
Taking up Beaus idea of modulating the laser at a fixed carrier frequency, instead of introducing a resonant tank at the Rx end why not use a quadrature demodulator. This would:
1. Provide a signal averaging function to help with the detection of small return signals.
2. Provide a means to measure the phase of the returned signal.
In fact could it not be a version of Beau's spectum analyser for the Propeller http://forums.parallax.com/showthread.php?p=663985
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
Here is the test.· Tell me if I am dumb...
250mw infrared laser at 940nm pointed at nearby cliff face of the majestic Huachuca Mountains.· I am not going to use a telescope yet with the laser because the range is only about 3 miles.· The Photodiode will be on the telescope at the point of focus (not the focal point).· Using a camcorder i will be able to see the spot on the mountain.· Then pointing the telescope at the same spot I should see the spot with the photodiode.
Now the fun part and this is where I may be dumb.· I am going to use my trusty Parallax O-scope and attach channel 1 to the + power· on the laser.· Channel 2 on the photodiode.· I will use the propeller to modulate 50khz on the laser.· I should see the 50khz on the direct connect to the laser and I should see the returned laser pulses on the photodiode. I should also see a wave shift on the returned signal.· If I am wrong somebody fix me.
Michael King
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.
Michael King
Application Engineer
R&D
Digital Technology Group