The night was always a mystery. Then mankind found fire and could light up the night. After a few more years mankind developed light bulps, photography and finally digital photography. And now I come along and present you a cheap device that allows you to take pictures in the night without intervening the scene.
The idea came up, when I wanted to develop a device to take pictures at a party without blinding everybody with a strong flash.
So what do you need:
Some IR-LEDs, the stronger the better (20-60 pieces)
a power transistor (i.e. BD649)
some capacitors (1000uF or more)
one resistor (50Ohm)
an old mouse button
and some breadboard
To use it, you need of course an IR-sensitive camera. You can build these by yourself or get someone to convert it for you. Look at these guys here: www.lifepixel.com
My other instructable is about how to make a Nikon 4300 IR sensitive. Look here!
Step 1: The Principle
The principle is to let the LEDs only flash up for a very short time but with much more current than normal.
Some LEDs are suitable for this, others not. You should try to find out before ordering a large quantity. In the data sheets this value is called peak forward current or surge current. It should at least be something around 100mA or more, while the normal forward current is 20mA for most LEDs.
In the picture you see the data for some diodes called SFH4650 with a normal current of 100mA and a surge current of 1A!!! The higher the peak current is, the longer has to be the time between two events. Say you flash 1A for 10ms then you should wait at least another 90ms before flashing again. But normally this is not a problem when triggering by hand.
Also check the wavelength of the IR-LEDs. Anything above 780nm is enough to be out of the visible range but in the center of the sensitivity of a IR-Camera. 850nm is nearly perfect, those with 920nm are too dim for the camera to see.
Step 2: The Circuit Diagram - and Why It Works
The things are pretty easy here, as we don't need any micro controller or complex ICs. Its just a capacitor (C), a resistor (R) and the transistor (T).
The capacitor and the resistor build something that is called a differentiate device. When at the input side there is a change in voltage level, the output side will transmit a small peak. Look at the picture for a sketch.
We use this characteristic to switch on the transistor for a short time when the button is pressed.
When the button is released the peak occurs with negative voltage level, but this won't have any effect on the transistor.
By the values of both devices you could coarsely time the duration of the flash. The time constant of the e-function describing the decrease of voltage at Uo is calculated by t = ( R * C). After one t interval the voltage level is Umax / e.
I used a 56 Ohm Resistor with a 2000 uF capacitor, which gives a flash time of some hundreds of a second.
As you can see in the formula, it is possible to use different combination of Rs and Cs to get the same time base. What changes is the current that drives the transistor. The bigger the C is, the lower the R must be, but the higher the current is that drives the transistor. So it is advised to use a rather big value for your capacitor, as then the transistor gets enough current, will be totally switched on and the flash is bright enough.
The transistor I used is a Darlington Transistor, which is in principle two transistors together in one housing. The advantage is the high amplification. A small current at the base switches a few hundred times higher current between the collector and the emitter. Just what we need. :-)
Another limitation is the power source. If the BD649 is switched on, it's similar to a short circuit for the battery. This means that the voltage breaks down and with it the current. So we add some back-up capacitors, where the power is taken from during the flash. Contrary to classical flash lights they don't have to keep high voltages, but large amounts of 'current' or capacity. This is why I took two super-caps and three normal high capacities.
Step 3: The Miracle of the Mouse Button
The mouse button is a very special pushbutton. First of all it is quite cheap, because you can always find and old mouse somewhere. The second point is: It switches between two connectors and this we need for to discharge the capacitor before the next firing. There might be other pushbutton switches with the same functionality, but I like the mouse buttons best.
Step 4: Solder on Breadboard or Etch Some PCB
As the circuit is this simple you don't really need to make your own PCB, but it doesn't harm though.
I did it, because the LEDs are a bit too big to fit in two adjacent rows of a 2.54mm bread board. Although it was quite a hustle to drill all the holes...
And of course I wanted something really good looking! ;-)
Step 5: Try It Out With Your Camera
last step is find a suitable housing and try it out with your camera.
Have fun and don't break the law!
Step 6: Post Scriptum
- The flash is not synchronized to the camera. At the moment I take a picture with an exposure time of 1s and then flash manually while the shutter is open. This won't work at partys when there is ambient light.
- When there is ambient light then either the picture gets blurred, when taking long exposures or the triggering of the flash is too difficult. You can't trigger the flash to meet a 1/60s exposure.
- When you use an IR-filter on you camera to cut off the visible light, then you also reduce the flash to some amount, because no filter is perfect, resulting in a rather dark picture again...