Technique: Photos of Low-Speed Impacts
Overview: Why? What? How?
The goal is to freeze different stages of some fairly quick process (at least when compared to the time resulotion of the human eye). Since I don't have fancy toys like a high speed film camera, lack of good equipment has to be compensated with a smart setup applying just a an ordinary consumer-type digital camera and some self-made electronics.
Unfortunately, the camera's minimum exposure of about 1ms is too long for the purpose and furthermore the delay from pressing the button to actually opening the shutter is around a second - quite long. So, the basic idea is to make the complete room dark, open the shutter of the camera, drop the ball, ignite the photo flash after some time and close the shutter again. This way the real exposure time is given by the duration of the photo flash which was measured by me (using a DSO) to be less than 100us (better than 1/10,000 second) when set up accrodingly.
Dropping and sensing the ball
Flash delay trigger
After the start signal, some time passes during which the ball drops down and then finally touches the sand. This time is in the order of several 100ms for usual drop heights below 1m. I designed a delay trigger which will start running when the light barrier sends the start signal, wait some time and then fire the flash. The time can be varied in steps of half a millisecond (500us) using an impulse switch ("Drehimpulsgeber"). The intrinisc accuracy of the microcontroller (due to interrupt handling) is 4us and the base clock has a tolerance of less than 100ppm.
Below, you can see the circuit of the delay trigger including the light barrier and the required amplifier as well as the connection to the photo flash.
The 3 LEDs connected to the microcontroller are a delay change indication for visual feedback when turning the switch. Additionally, the JP3 connector was connected to my self-made frequency counter to measure the actual delay time. There is some additional comfort like a switch which will enable auto increment mode where the microcontroller will increase the delay automatically by a fixed amount after each trigger event...
The microcontroller runs a self-developed firmware (written in C, compiled with avr-gcc) and was downloaded onto the controller using my USB-based Atmel programmer.
The complete setup
So, with the above setup, what was actually done: When all the testing is done successfully and everything is functional, the light is switched off. The cycle begins when the camera's shutter is opened and the ball is dropped by switching off the coil. As soon as the flash lights up, the shutter is closed again. Now, the delay is incresed by a fixed amount, the ball is put onto the dropping coil and the sand is levelled again by shaking the container. Putting the container back into its predefined position ends the cycle.
Additionally, the delay can be checked with the frequency counter (in time measurement mode). All that is done with lights switched off and it's best to have two people operating the devices.