Framed Laser Harp: Need help picking diodes

[Stoney3K]

Using a low-cut filter (ideally a DC block filter) will give you some success, however, the filter characteristic will need to be very specific because the frequency you will need is quite close to the frequency you don't need.

Suppose you're scanning at a decent rate, for a steady beam pattern the repetition frequency is between 20 and 100Hz. A 20Hz low-cut filter is quite challenging to implement, especially if you need a very steep curve.

The problem with laser harp detection is two-fold. The whole point of the excercise is to detect the scattered light coming from your hand when it's touching a beam, and only that. There are a few tricks you can pull to accomplish that.

First of all, make sure the amount of light scattered from your hand is sufficiently bright when compared to the ambient (stage) lighting. This means, get plenty of laser power in the system, and use reflective (white) gloves to make sure most of it scatters back down and doesn't diffuse or heat up your hands in the process. Not that it's gonna hurt, but it means a lot of useful light is wasted.

Second of all, make sure your detectors are matched and pick up adequate light. Most phototransistors are more sensitive towards the red or IR range of the spectrum, and don't like picking up visible green or blue all that much. You may have more success with a red (660nm) diode laser, but don't try to mix IR or rob your 532nm of an IR filter, because of possible laser safety issues.

Step 3: Make the photodiode array more specific to your target. Part of this means to build a cowling around the detector array which blocks out any ambient light coming from directions you really don't need (such as the front of house, behind you or the sides), moving the detector closer to your hands if you need to (a microphone or drum tripod works very well for that purpose) and filtering out any colours of light that are useless if you have a monochrome system. Filters will do more harm than good if you're using an RGB or multiline laser.

When all that's said and done you can start looking at electronic means to clean up your signal. (Comparatively, try to find your favorite TV channel when the sattelite dish isn't properly aligned is a pain.)

Things you can use to clean up the signal electronically include: Low-cut filtering (aka AC coupling) and measuring the ambient light amount separately through a dedicated photodiode. If you have a microcontroller-based system, you can also bias the detection through a separate DAC connected to the microcontroller and calibrate it once the harp is running. Basically, if you're not touching ny beams and the controller knows it, it can bias the amplifiers to output zero at that point. Similarly, if you are touching a beam, the controller can set gain appropriately up to the point where the signal is strong enough to be useful.

Some other things to consider are to consult the lighting crew beforehand if you have a gig, and tell them to block the front lighting from hitting the harp's sensors. Also, make sure your beams terminate somewhere black or bounce off-stage -- when I'm testing my harp at home, the amount of scatter from the white ceiling is too much and it almost illuminates the entire room. Bouncing it forward through a set of mirrors will make sure the detector doesn't get swamped by scatter from the harp itself. Aside from that, it looks more awesome when the beams are travelling over the audience as well.

Last but not least: Don't be scared to scan slowly, since you're not going into the crowd with this thing, slow scanning is not a crime. It will help because the dwell points become much longer and beams become more visible.

[Things]

This is how I created my sensor. Just a cheap enclosure, and a MDF frame. Works brilliantly at only letting directional light in.



Bare hands work OK, although sometimes you can get repeated note triggers when you don't want them, so using something to actually play it (Strips of paper, white gloves etc) work well.