High accuracy timer

oracacle

Senior Member
I while back I built a camera controller to allow the shutter to be controlled by various inputs (http://www.picaxeforum.co.uk/showthread.php?26768-High-speed-photography). The system was fast and I wasn't missing any shots due to its response time, however I saw that the system could be capable of a huge amount more, even with no hardware changes. I was able to add interval timer and shutter timer to allow for camera to settle before shutter activation - it is able to do this with mirror up setting too.

While the timing side of the system seemed accurate by comparison to wrist watch I was unable to confirm it accuracy. I also started to think about taking into account shutter lag for times where flash exposure was not possible.
After seeing and thinking about hippys high accuracy timer (http://www.picaxeforum.co.uk/showthread.php?14517-High-Accuracy-Timer) I thought that I could develop something that could do the job of timing just about anything I needed.
after a little testing I found that latching each input, and being able to invert them if needed was a must, this input then feed into a XOR gate before final conditioning.

next I had find a was to display the information, OLED was the simplest way, parallel being the fastest way. so that the route I went.
The final object is able to be activated through either push buttons, or through an exterior input. it can drain it power from the other item, or provide power to sensors (I feature that revealed itself during assembly).

The timer revealed that the fast photography unit as a response time of around 0.00029 seconds. it also revealed that its response time varies depending on battery power, with fresh batteries give something more like 0.00023s and flat batteries giving 0.00032s response time

here is a video testing the camera response time using a switch input to activate the shutter and a photo transistor to detect the flash which is activated when the shutter is open

 

Attachments

hippy

Technical Support
Staff member
It appears to be using the trick of counting the number of pulses presented during a period so all one needs is a 1GHz pulse stream. The PICAXE hardware doesn't support such high speeds, but it can (AFAIR) run up to 16MHz giving sub-microsecond timing, multiples of N nanoseconds.
 
Top