Well ... how to start?

 

The basic idea is known, I want to make my own laser harp. I have some additional ideas about what is basically a laser harp. I want nine notes and a control footswitch. I also want an LCD screen that will tell me quickly important things in the context of live use. The footswitch will also have a matrix display for the current preset number.

 

Other choices relate to the equipment and technology to use. Microcontrollers seems obvious, relevant and easy development platform attracts my interest almost naturally : Arduino. So I order a starter kit and some extra accessories. Laser side, my choice is an analog RGB module. This will allow me to precisely control the dosage of each color. As this is a complete module, I carefully avoids the problems of coupling and control of each color.

 

The choice of a good power supply was not complicated. The Arduino 5V operates, laser requires 12V. A PC power supply seems perfectly fit the bill. It delivers good power in all the necessary voltage and is inexpensive. There are many things to be defined as : the size of the matrix display of the footswitch and the footswitch itself... ! A good footswitch also inexpensive (49 €) is available at Marshall with four switches. Perfect !

 

First developments, first tests ...

 

It took me some time to tame the Arduino. Very convenient, very easy! Regarding the deviation of RGB laser, I first tried to satisfy me with a stepper motor. First disappointment (predictable way) : too slow, too flickering... Well, okay. I order XY scanners closed loop with its management and supply. Ah! Much better! Perfect in fact... For its implementation, it took me design a DA converter to provide an analog signal from the PWM-generated aduino. Notwithstanding this desire to create the best product possible, it appears that this converter is not up to par. So I developed another that is much simpler and that suits me perfectly. It is time to think about the sensor, what I fear most about this project...

 

Sensor ? There is a sensor ?

 

Yes ! It has a sensor ! We must detect the hand through the beam ! Of course, this sensor is not at the top of the beams. It would be too simple and it would lose its magic ... It is on the floor in front of the harpist. How does it work ? In principle, it is quite simple. In practice, this is not a foregone...

The principle is based on the reflection of light by the hand of the harpist. The glove reflects light to the sensor placed in front of him. Then the problem is to determine which beam was interrupted. Again, the principle is simple. The microcontroller knows which beam is currently generated. At this time, if it receives light, it means that the beam has been interrupted. It then remains only to play the note corresponding to that beam. Otherwise, if the sensor does not detect light when generating a beam, it means the hand is not there : the note corresponding to the beam is not played.

 

Some constraints ...

 

First problem with the sensor : the small amount of light reflected by the glove. After many tests, I finally returned to the simple white textile glove. But it is important to note that the laser power must be at least 500 mW to permit a stable and comfortable operation.

Second problem: the ambient lights. The harp must not detects anything other than the laser. Forget the color filters since we work with RGB laser. The best solution seemed to scan the laser at a different frequency of most lighting. This is done almost automatically as we seek to avoid flicker. The scanning frequency must be high. Perfect ! filtering detections at lower frequencies.

 

Note: I still noticed that the video projectors are detected by the sensor and interpreted as reflections of the hand. This happened during the unfortunate projection on me! On this side, improvement is conceivable ...

 

Third problem: how to detect a wider angle ? After trying several reflectors of various flashlights (and even bicycle lamps), the results often remained in a very just limit. The more moved away from the center, the more had to go down there in the beam. My only solution has been to use a reflector intended for LED spotlights. This type of reflector available in various forms of transmission (used here to capture rather than to diffuse). The combined result with a good sensor setting is sufficient. The additional idea would be to use a more powerful laser, but this presupposes a higher cost.

 

Fourth problem: too short duration of the detection pulse. A very simple solution to compensate for this: the use of a monostable based on ... 555 ! The pulse width remains adjustable so as to enable modification, in case of ...

 

Finally, the sensor has requested three prototypes before properly fulfill its mission. The first based on LDR was operational but too sketchy, influancable in short, basic. The second has been the grind of the third. Finally, the solution is retained in the third version of the detector. An idea of improvement is still possible : detect the hand height to produce a specific effect ...

 

Play notes ...

 

The harp produces no sound. It controls a synthesizer or a computer using the MIDI protocol well known in the world of electronic musical instruments. When a beam is interrupted, it transmits NoteOn and when it is no longer the case, it transmits NoteOff. For now, the channel used is 1. When the configuration will be possible via a convenient interface, each note of each preset could be assigned to a particular channel.

 

And now?

 

The harp is functional and responsive. Certainly, many things could be improved. Side of the harp configuration, nothing is really created. Presets are included in progressively in the firmware before recompiling. From a purely physical standpoint, the harp is still lacking housing.

 

And after?

 

It is planned to develop an Android application to set the harp from a tablet application. The harp is also already now equipped with its Bluetooth module. A housing is under study to protect the internal electronics and mechanics of the harp. In addition, smoke distribution system is also under study to spread smoke in the beams.