Is a micro processor appropriate for this project?
DavidSmith
Posts: 36
I am an amateur with no experience w micro processors.
I have a project that might use one, or someone may tell me to just get a 4 track tape recorder.
I am a member of a caving club, and we are trying to "invent" an easier way of mapping caves than the string and compass system we are currently using. (yes, cavers in Italy are doing a beautiful job w a commercial laser system , , , , ,).
The idea is sonic, and will require 4 microphones and an electromagnetic receiver. The user will touch the wall of the cave w a wand and trigger a simultaneous electromagnetic pulse and sonic pulse. The electromagnetic is essentially instantaneous and serves as the time mark. The time of arrival of the sonic signals will give the location of the wand.
We do not intend to do real time, but record and manipulate the data later.
We would like to use sound frequencies up to 100,000 hertz, but can live w lower. We will probably take 1,000 measurements a day (X 5 for all receivers) w about 1 second of recording time for each.
Is a microprocessor equipped w a data logger a good selection? Or would it be easier to use more prosaic methods of recording the data? If a microprocessor is a good choice, a propeller, basic, or several basics?
Any of you experienced people have an opinion?
Thanx,
DavidSmith
I have a project that might use one, or someone may tell me to just get a 4 track tape recorder.
I am a member of a caving club, and we are trying to "invent" an easier way of mapping caves than the string and compass system we are currently using. (yes, cavers in Italy are doing a beautiful job w a commercial laser system , , , , ,).
The idea is sonic, and will require 4 microphones and an electromagnetic receiver. The user will touch the wall of the cave w a wand and trigger a simultaneous electromagnetic pulse and sonic pulse. The electromagnetic is essentially instantaneous and serves as the time mark. The time of arrival of the sonic signals will give the location of the wand.
We do not intend to do real time, but record and manipulate the data later.
We would like to use sound frequencies up to 100,000 hertz, but can live w lower. We will probably take 1,000 measurements a day (X 5 for all receivers) w about 1 second of recording time for each.
Is a microprocessor equipped w a data logger a good selection? Or would it be easier to use more prosaic methods of recording the data? If a microprocessor is a good choice, a propeller, basic, or several basics?
Any of you experienced people have an opinion?
Thanx,
DavidSmith
Comments
Well, I don't know much about what is needed to map a cave. However, the Propeller is perfect for datalogging at high speed and could probably do what you want with minimal external components. It's multicore, which seems like it would come in handy for what you want to do.
It may even be able to compute your data real-time, so you don't have to do any post-processing. I can't be sure about that, though. As I said, I don't know what is needed in mapping a cave.
Do you intend to rely on the acoustic signal differences for compass direction and inclination? Having mapped some caves myself, I would be concerned that the acoustic method might not give the kinds of angular (compass and inclination) accuracies you need, and these inaccuracies would become compounded as you move through the cave. However, I think you might be able to integrate a standard acoustic distance measurement with an electronic compass module and maybe a tilt measurement to create a kind of survey/theodolite device that could be easier to use in cramped spaces. Something like the Propeller chip could probably handle that, and record the data onto an SD card. But I don't know enough about the accuracies of the compass modules and tilt measurement chips to judge whether or not it would be good for your needs. If I remember correctly, I think our old hand-held devices were only good to about 0.5 degree but I'm sure the cold and wet and fatigue made our measurements much worse than that. What sort of accuracy were you hoping to achieve?
Also, you need to think about what kinds of effects the acoustic reflections will have on your device and maybe whether or not the 100kHz would drive the bats... batty.
In my previous post, I presumed by "sonic pulse" you were talking about a signal traveling through the air of the cave. But now I'm wondering: were you talking about a sonic pulse traveling through the rock to microphones (seismic sensors) mounted on the earth's surface, etc.? I don't think 100kHz would travel very far through rock, so I don't think it would work for seismic measurements.
The idea isn't great accuracy, but just faster than the current system.
Okay, but do you intend to have the sound energy travel through rock or air?
Are the microphones on the earth's surface above the cave? or inside the cave somewhere?
How far apart are the microphones?
Everything will be air, inside the cave. The microphones will probably be mounted on surplus camera tripods and be as far apart as the size of the room allows. Have not decided on a cable system or a bluetooth. Bluetooth would be more convenient, but the cable system is simple.
The reason we think this may work is we are not trying to do what bats, submarines, radar, etc. do. We will not send out a signal to be bounced back and interpreted at the receiver. The "target" will send the signal and that should be clean enough to get us the necessary four distances to plot it pretty close. Any "echoes" will arrive at the microphones after the initial wave front that measures distance. We have plan for a simple statistical plot to build up details where we can't reach (like up about 40 feet). It should work, but will be a fun application of math first, program the math later.
Well, to answer your original question, I'm sure a microprocessor would be great for this, especially the Propeller. But my main concern is whether or not you can get the kinds of resolution you want with sound. I'm not very knowledgeable about ultrasonics, but it seems to me that this might work for very large rooms but maybe not for passageways that are narrow because it would be too difficult to get a good triangulation. Have you done any calculations on this to see if your microphone separation at the desired sonic frequency can give you the triangulations you need?
My main job is topographic surveying (usualloy with DGPS systems), and I now the seller I buy from (in Ancona) has this kind of laser based commercial stuff.
If you plan to develop a new system for the leaning challenge a Propeller is a great option. I'm not a good programmer, but if you are close I can help you to get stared.
On the other hand for a quality job an used laser scanner could be a better option... check the guys in Ancona or other sellers for refurbished units...
Massimo
http://en.wikipedia.org/wiki/Gunfire_locator
In the center of the cave I would place 2 (or3) omni directional microphone about 10feet apart.
When walk around with the speaker wand and hold it to the wall and press a button to send a beep.
Now it would be up to the software to trianglete the direction and cancel echo.
Maybe easier if wand have internal clock that is in high precision sync with receiver.
A clock that rolls around from 0-1000ms and only sends a beep at 0ms
and the receiver only start listing at 0ms
If you're talking about using a Bluetooth for sending the start pulse for timing the ultrasonics, you might also consider using a cheaper, bright LED emission, perhaps an infrared pulse to start the timer. Because your measurement pathway will need to be a straight line and clear of objects, I would think an LED or IR emitter would work for setting this timing pulse.
Yes, I can see how that would work in big rooms, but in all unexplored caves that I ever mapped, people spent most of their time with one shoulder on the floor and the other shoulder squished against the ceiling, so it's hard for me to imagine a sonic system could triangulate in those circumstances.
http://forums.parallax.com/showthread.php?135899-Four-Buck-Ultrasonic-Sensor
One is controlled by a prop, it sends a ping.
The other one is hardware programmed (no mcu or a tiny one) to send a signal back but wait 2sec to do so (to let any echo go away.)
followed with a 2second dead time so it does not answer back it's own echo.
The prop calc: distance= time-2000ms/speed-of-sound/2
You calibrate by holding the sensor exactly 10meters away from the Prop at boot-up sequence.
The prop have a e-compass and you use a sight to see the sensor,
also a up/down angle e-sensor so you can get height difference.
Distance/Height and Compass angle are recorded to eeprom.
Massimo, I did do some web research and learned about the laser systems being used in Italy (you?). You're right, I want the challenge. I don't want to plunk down my money and have it done correctly. I want to do it incorrectly my own little way. (PS I'm an amateur blacksmith and I figure my return per hour is about $.09 - does that tell you I'm more interested in the doing than the achieving?)
Electric Aye, 100,000 hertz would be accurate to about an inch (2.54 centimeter). Much higher resolution than we really need. I picked that frequency because it is above the range used by bats. Also, we wouldn't even go in to a cave w a known bat population w the current problems of white nose syndrome. Just too much chance of doing serious harm.
About small areas (volumes): In theory this could be used for any volume bigger than a few feet across. In practice, I suspect there will be places where we will get out the string, compass, inclinometer as the simpler of two awkward solutions.
My original question about using a micro processor was because we plan to process the data afterwards, the same thing could be done by replacing the microprocessor/flashdrive combo w four synchronised tape recorders. Or, even just a straight passthrough to digital recording (something I do all the time w a single channel). The microprocessor might be overkill as all it will do is cache the signals (briefly to avoid overlap) and then forward them to storage w an identifier as to which microphone caught this signal. Possibly no calculation at all and depending on the final design, maybe even no separation of signals to different channels. (?????).
Interesting variation. I'm at the pencil pushing stage right now, I'll include your idea w the other possibilities.
Thanx
As I recall, space and weight are things you want to keep to a minimum when dragging your exhausted hypothermic-corpse-of-a-self through a cave at the end of a mapping job, so I'm guessing a microprocessor system with an SD card recording the timing data, would be the best way to go. The Propeller would probably be ideal for this. I would guess such a system would be much cheaper and more ruggedizable than a tape system, too.
BTW, where are you caving?
I'm located in Sacramento California USA. I'm part of the Mother Lode Grotto.
Oh, yes, about 10 miles from the Parallax office/factory.
By the way, you're in the east and you have caves w squeezes? We joke about how the easterners think it's a tight cave if they have to stoop. California is very marginal cave country. I'm not a big man and I've failed to get through squeezes because I couldn't hold my (exhaled) breath long enough to make the wiggle. (some of the ladies made it of course - and not all of them).
I used to do caving in Kentucky and Tennessee. It was our insanity to find new caves and map them out. Big rooms glistening with speliothems were always our dream, but most of our time was spent on our stomachs, lower-lip-deep in cold mud and bat Smile. Crazy dangerous but fascinating, too.
Late joining here, did you consider Inertial navigation?
A gyroscope and accelerometer track your absolute relative postition and orientation in space.
Does not care about rock or geomagnetic or GPS signals.
There is some drift, but I think it would not be prohibitive at your speed and range.
I would link to the german quadcopter example, but I don't have it handy.
Might be simpler and more accurate than the scheme you have described (maybe I don't fully understand that).
But there's a bunch of folks interested on INS, and the prop would be nice for that, and a single prop could probably do INS and accoustic at the same time.
Check also david laser scanner
http://www.david-laserscanner.com/
There is a free version available. In general it requires a target behind the object, but if you have a predictable line sweep you can scan with no background:
http://www.david-laserscanner.com/wiki/user_manual/scanning_without_background
I don't know if it would work in your case, and you would still need a micro like the propeller to drive the laser sweep with precise motions.
Massimo
Thank you, but I kind of want to do it the hard way (more fun).
prof-braino,
Inertial would work for the way points (benchmarks) but I don't thing it will do what I want for mapping. I need thousands (millions?) of plottable points, preferably taken as close to the wall (inches) as possible.
ElectricAye,
Yeah the joys of caving: Freezing, falling, suffocation, bad air, the redolent odor of dead bodies (some even mammal), having to drag your own broken body out of the cave after fall 'cause there's not enough room for the rescue team to do it, bat Smile, and amberrat. About the last two: Thank god for the twentieth century and micro pore filter breathing systems. I don't want to die with important parts rotting out. Interesting hobby, what?