An Infrared Webcam for the bird's cage
The selected camera is the D-Link DCS-930L. This camera was selected because it is one of the cheaper wireless cameras. I preferred a camera without infrared LEDs because I would like to minimise the light intensity myself as much as possible. After testing the robustness of the wireless connection for some time (something that made an earlier attempt with a Wanscam camera fail), I made the infrared LED bar. Only then I discovered this D-Link camera has a very strong IR filter. So, it had to be removed first.
Removal of the IR filter
As described on this blog, it is quite easy to remove the IR filter of the DCS-930L.
First, open the camera carefully. I did this by placing a knife between the silver rim and the top cover. See the first picture for how the housing is clamped, when it is partially open.
The camera consists of three parts. There are no screws that fix the electronics board.
Cut the square rim of the lens mounting, and pull the lens mounting out. Now, kill the glass IR filter. In the left-hand picture, the filter can be see, in the right-hand picture it is gone. The lens-mount can be pushed back easily, and it sticks immediately.
The first picture is taken with the camera in the original state, The light is artificial light from an energy-saving lightbulb (CFL). In the dark, with infrared illumination, there was no response at all (completely black picture). The second one after removal of the filter with the same CFL light, no infrared The colors did not change. The picture is out of focus: an advantage for me because I wanted a shorter focal distance anyway for the small interior of the bird's cage.
In the ultimate infrared test (6 IR LEDs, 20 mA each), a black and white picture can be seen. In the final picture, there is both IR light and CFL light. In that case, the color has changed. It is a little bit bluish.
When building the camera in the bird's cage for the first time, I discovered that the focal distance was too small. This is no surprise because I have removed a glass plate from the lightpath. Objects at a distance of 5 cm from the lens looked in focus, anything further away was blurred. I had to do something, because the avarage distance of the birds in the cage will be 40 cm. For a lens we know that 1/f = 1/b + 1/v. When v has to go from 5 cm to 40 cm, while f stays the same, then b has te become smaller: the lens must be closer to the CCD. To do this we have to modify the lens mount (the black object). So, I remove the four plastic pins on the lensmount and I used sandpaper type 1200 to remove plastic from the lens mount until the objects at a distance of 40 cm were in focus. This is the result. The picture is taken at daylight in another room in another setting as before, but it is clear that we can see the eyes of the penguins again.
Build into the cage and optimise light
After the birdhouse was assmebled, I determined that using three infrared LEDs, the minimum light intensity was achieved at 14 mA. A circuit was made that switches on the infrared LEDs when the outside light is low. However, I did not realize that the old black/white cameras are much more light sensitive. my old analog BW camera could record images with only the external daylight coming through the 33 mm bird hole. With the color camera this is not the case, so I had to cover the light sensor permanently in order to have infrared light all the time.
The birdhouse itself is shown below. It was made from old floor panels.
The D-Link DCS-930L camera has the advantage it can send pictures automatically to a server by means of the FTP protocol. After that, it is only a matter of embedding the picture on a webpage that runs on the same server.
Images are on the bird's pages.