Hey guys:
I am working on a project that tracks an object on a 2D plane. IE, I only need 2 axises to work with. I saw the ultrasonic sensors that could be used for this, but the price (30$) is way too steep.
Could capacitive sensing work? I'm not sure how far (or accurate) it can sense...
Or should I use something totally different?
Maybe its just not possible. Thanks guys!
--Dylan
Thanks for the quick reply.
The object is the size of a finger (actually, it is a finger).
I am trying to make a DIY "toutchscreen" in front of a monitor. The uC takes in the data of where it has detected a finger (or really any object), and turns it into XY chords and relays that to the computer.
I am hoping for a resolution close to 1cm to an 1inch (1 cm - 2.54cm).
On board power: yes, using a wall-wart & 5v regulator (my Arduino is built on a breadboard)
On board computing: yup (atmega328?)
The calculations are the easy part, just sensing the finger is what is holding me up 
Good luck. I have yet to see anyone claim success.
Just wanted to get some ideas on how to get the position of an object
I also have yet to see someone /try/ this with IR or something of the sort.
OTOH, there are some relatively inexpensive transparent touch screen products available. Have you rejected them for some reason?
I've seen these overlays before. Quick google search brings up a price of 104$(US) for an 18.5" screen. Ebay has a 15" touchscreen overlay for 117$. Hardly inexpensive :o
Well, when I asked you for the area, you did not respond, so we had no basis for evaluation.
Sorry bout that!
The screen is about 24" diagonal.
Have you tried this?
They claim you can do probably 22" screen on their PC pen
I simply don't have that kind of money lying in my drawer for it but the principle seems sound.
Have you tried this?
They claim you can do probably 22" screen on their PC pen
Huh - looks promising. Very promising!
Maybe figure out how it works, and try to implement it on a larger scale.
Looks like it uses ultrasonic to get the position of the pen, and the pen determines wether it is touching the screen or not, and relays the info back to the hub via IR.
Maybe it could be done with something like this:
The receivers (ultrasonic) around the screen can get the position of the finger, and the IR can sync the pulses as to do away with timing ussues and overlapping pulses.
The finger has a glove or ring that emits the pulses, as well as the sync pulse on IR.
What do you think? Can an ultrasonic receiver position /That/ accuritly?
Thought something like this would work... Ah well.
Does anyone know of a ultrasonic /Sender/ only? Or would a piezoelectric speaker @ 40 KHz work fine?
I have some recievers in mind also: http://www.allspectrum.com/store/product_info.php?products_id=1300
OK, thanks. I really appreciate your honest opinion 
Is it that it's not accurate enough?
You won't get reliable "pings" off of a human finger
Ah, you wouldnt be pinging off the finger. The finger is emmiting the ping via its own "glove", with battery pack and attiny + IR emmiter for syncing issues.
EDIT:
See my 1337 art skills on one of my previous posts in this thread
But the diameter of the finger exceeds your desired resolution, and you can't control the rotational position of the finger.
True... I thought about putting it in a more "pen" form, much like:
http://www.penandfree.co.kr/english/overview/introduce.asp
They seem to get very good accuracy, and the pen seems to be emitting the ultrasonic ping (see the "cage"). Plus, they are only using 2 sensors. A sensor on each side of the screen, in my mind, would give much greater accuracy (not to mention the ability of calibration).
If put in a "pen" form, rotary position would not matter.
Not to mention that putting on a glove with active circuitry is something most people might be reluctant to do.
Meh, im not planning to sell it
I think the accuracy of a sonic ranger is near 1~2mm. The one I got has that accuracy. The pen that I mentioned (made by a Korean company) is similar to a sonic ranger, except the sender of the sonic wave doesn't receive it. I think the pen and the receiver clip use IR to synchronize when the sonic wave will be sent. Then the pen sends out the wave, and the receiver (stereo) receives the wave on two microphones. Use geometry one can find out the exact location of the pen tip, seemingly with mm accuracy.
I think the same company makes a larger version for white boards. Just search around. You may not need to build one yourself.
I also have yet to see someone /try/ this with IR or something of the sort.
This project was designed to be an overlay for a touch screen originally:-
http://www.thebox.myzen.co.uk/Hardware/The_Sound_Square.html
But the $30 limit is a bit of a killer.
I think the accuracy of a sonic ranger is near 1~2mm.
Perfect
I think the same company makes a larger version for white boards. Just search around. You may not need to build one yourself.
There pens are cool, but there is still some things that bug be about them (Well, not really bug me, but I have some ideas for improvements ;D ).
Did someone say multitoutch?
Or gestures?
Open source, anyone?
IF I get this going, ill be sure to share source and give credit were it is due
but I have some ideas for improvements
It is all very well having the idea for improvement any one can come up with that. It is how those improvements are going to be realised that is important.
I have the idea to improve air transport with an anti gravity device. However realising this is something else.
Patent 7,479,949:
A computer-implemented method for use in conjunction with a computing device with a touch screen display comprises: detecting one or more finger contacts with the touch screen display
This has nothing to do with using sonar to detect the position of multiple objects. It dosent even need to make contact with the screen to detect position: its totally different technology.
It is all very well having the idea for improvement any one can come up with that. It is how those improvements are going to be realised that is important.
I have the idea to improve air transport with an anti gravity device. However realising this is something else.
Also all very well, but you (and I) dont have the know how to implement an anti-gravity device.
However, I believe sonar has been implemented before, as we can see with the Duo pen thing.
We used sonic 2D digitisers on a PDP-11 in the late 70s.
Cost a lot more than $30!
We used sonic 2D digitisers on a PDP-11 in the late 70s.
They had computers then o.O
Haha I bet it did. Wasnt it like 5 million somehting for a 12 MHz IBM mainframe?
We used sonic 2D digitisers on a PDP-11 in the late 70s
Yes but did they detect two points simultaneously?
I believe sonar has been implemented before,
Yes but not two at the same time. I know you can use sonar imaging techniques but that is way above your pay grade.
but you (and I) dont have the know how to implement an anti-gravity device.
Please speak for yourself. I am a physicist and all physicists know how to do it. It's just we are meant to keep it a secret from you.
Please speak for yourself. I am a physicist and all physicists know how to do it. It's just we are meant to keep it a secret from you.
Err sorry :-[ I'm only in HS Chemestry, didnt mean to offend :P.
Yes but not two at the same time. I know you can use sonar imaging techniques but that is way above your pay grade.
To detect 2 points:
Send an IR Pulse to the detector on the monitor
Monitor detector starts a timer
Send a Ultrasonic pulse out of the "pen"
----a little (very small amount of) time passes
One ultrasonic reciever recieves the pulse
Monitor detector saves the time
Another reciever on the monitor gets a pulse
Saves time again
---goes on for however many recievers it needs to
All the pulses are accounted for ( or they timed out)
Calculate distances
Done, send XY data to computer over serial interface. Let Autohotkey or python worry about interpreting the choordinates.