Direction location - blind kayaker tactile guidance

Thread Starter

JRiley

Joined Mar 21, 2016
5
Ah, so the built-in AGC is what produces 0v when IR is detected or high (5v) when it is not - either on/off (digital)? Desirable would be an analog reference to determine relative strength - hence direct control of AGC. Sound right?
 

Bernard

Joined Aug 7, 2008
5,784
With IR receiver chips having AGC, it might be difficult to determine which receiver signal was the strongest.
With home brew receiver it might be possible to OR the individual AGC control Vs & use output as - input to
a comparator, + input is the individual AGC control Vs. The activated comparator is the receiver with the strongest signal.
 

Lestraveled

Joined May 19, 2014
1,946
How the scanner works
- On a spinning scanner, there will be a sensor that will produce a pulse when the scanner passes zero. This pulse would reset a continuously running counter to zero. As the scanner rotates, the number in the counter represents the angle.

- The scanner is rotating an IR sensor that measures the IR intensity and outputs a voltage. When the sensor scans across an IR emitter, the voltage momentary increases. (Lets call this signal, video.)

- If the zero pulse was set to occur looking directly forward of the Kayak, and the IR emitter was also directly forward, the video pulse would occur exactly at the zero pulse. If the IR emitter was to the left, the video would occur earlier than the zero pulse. If the IR emitter was to the right, the video would occur after the zero pulse. (Assuming clockwise rotation.) (It is actually better to have the zero sensor in the aft position, but for the example I set it forward. If you build this you will find out why this is true.)

How the AGC works
- The video of the IR sensor is fed to a peak detector. If the output of the peak detector in greater than a preset voltage then the gain is slowly reduced. If the output of the peak detector in less than a preset voltage then the gain is slowly increased. This is the AGC function.

How detection works
- Once the AGC has normalized the the video of the highest IR source to a preset level, a simple comparator can be used to sense the position of the video. When the video pulse occurs, simply look at the counter. That is the angle, relative to the zero pulse, that the IR emitter is at.

The only part that has to be processed in analog is the IR sensor video. The rest could easily be done in a PIC chip.
 
Last edited:

Bernard

Joined Aug 7, 2008
5,784
Could we use an oscillating IR receiver, say 180 deg., & avoid slip rings? Also how about a code disc with a limited no. of positions, say 16 or less ?
 

Lestraveled

Joined May 19, 2014
1,946
Could we use an oscillating IR receiver, say 180 deg., & avoid slip rings? Also how about a code disc with a limited no. of positions, say 16 or less ?
Rotating a mirror would be easier. The aperture for the sensor is pretty small so it could look up through the center of a bearing. Perhaps a 3/8" ID (or smaller) bearing, or gear, mounted on a hollow shaft, which is where the sense diode would be mounted. The mirror would be mounted at 45 deg above the sensor and rotate with the top housing. The bearing or gear are off the shelf items. Just have to build a clear top housing for the mirror.

An incremental shaft encoder would work nicely. A little pricey though.
 

jpanhalt

Joined Jan 18, 2008
11,087
I am sorry to come late to this very interesting discussion and have only browsed the previous posts, so my apologies if I mention something that has already been suggested.

It seems that guide 1 and guide 2 form a line that the disabled kayaker needs to stay within a margin of error from. In ancient days of aviation, there were AN beacons. One transmitted and "A" (dit-dah) and the other an "N" (dah-dit). The pilot navigated to stay within the beams so he got pretty much a steady tone when on a direct line (https://en.wikipedia.org/wiki/Low-frequency_radio_range ).

I wonder whether a variation on that approach might be adapter here. Each lead kayak would transmit a dit-dit-dit to the right and a dah-dah-dah to the left. If the disabled received only a dit's , he would know to turn left and visa versa for only dah's. Of course, the dit's and dah's could be tactile. For example, encode a dit as a 200 ms pulse and a dah as an 800 ms pulse. The decoding should be simple, and it should be easy to tell the difference between the two. The transmitters would need to be directional, but would not need to move. The receiver would also be fixed -- not sure whether one would need one or two (i.e., one on each side of the disabled's kayak).

Regards, John
 

jpanhalt

Joined Jan 18, 2008
11,087
After a little sleep on it, my suggestion in post #28 would not alleviate the need for the guide transmitters to stay oriented at each other (i.e., swivel) while the guide kayaks wiggle in the water. Conceivably,one might keep a constant orientation by having a separate, non-directional transmitter on each kayak with the other kayak tracking that for signal strength (similar to these old projects: 1) http://www.dprg.org/projects/1998-08a/ ; and 2) http://users.frii.com/dlc/robotics/projects/botproj.htm). That would add a lot of complexity, and I concerned that discrimination between two IR signals may not be sharp enough.

John
 
Last edited:

Bernard

Joined Aug 7, 2008
5,784
I discovered a 1 W IR LED in my junk box so hooked it up to a 555 & driver, 4 kHz, 20 % duty & listened to it with IR transistor, bunch of clip leads & audio amplifier. Works great up to 2 in. Re tried with silicon sensor ( Electronic Goldmine G15405 ) & audio amplifier; now detectable at over 30 ft. May next try a AGC controlled amp. detector.
 
Last edited:

Bernard

Joined Aug 7, 2008
5,784
No amplifier yet but some thoughts on setup.
IR detector oscillates 158 times per minute,( I have a 79 RPM motor ) over a range of 120 degrees; position is read from a Gray code disc. Gray is converted to binary , binary decoded to 1 of 8, 0-6 used so that a center segment can be used
to indicate on course , maybe alternate L-R buzzing on vest. As signal shifts to another segment buzzed stops on one side. As signal drifts further off course, buzzer period shortens.KAYAK #1 00000.jpg
 
Top