Go Down

Topic: Spark chamber camera trigger (Read 3848 times) previous topic - next topic


Hello all,

We are building a spark chamber. The basic idea is that we have three scintillators which we'll hook up to the analog input to perform coincidence/anti-coincidence. We'd like to trigger a camera when we get particular signals from these scintillators.

We have two ideas for this; we can either have a webcam hooked up to the computer constantly recording video, and record timestamps when an event occurs, and then we'll look at the frame at the timestamp time. This isn't an elegant solution though. What would be better is if we could have the arduino trigger the camera to take a picture, and either push it to the computer or just store it on a memory card on the camera until a later time.

Does anyone know of good camera hardware to use for this, or have suggestions for how we should go about doing this? I know very little about triggering cameras, or what sort of camera we might want for this. We don't need anything fancy, we just want to keep this as simple as possible.



Won't you miss the leading edge of the event when you trigger?

Maybe constant recording is the best solution.


Speed is of the essence, therefore software is to be avoided.

Much more true at the LHC, where they use several layers starting with hardwired detection, sifting down thru programmable gate arrays and finally at the lowest speeds and final stages the very best data gets all the way to computers running software.

Mottos aside, the ion lifetime in your chamber will be the most severe restriction on your instrument, and pressure is a big component of lifetime.  If your chamber can preserve an ion trail for milliseconds then Arduini could be in the innermost loop.  Thats probably the way to go, but chamber lifetime is probably more appropriate discussion for physics forums.  If you can get the timing requirements for your chamber figgured out in number of mS or uS, then many folks here can help.  Arduino can reach to microsecond regime for simple operations.

Using hardware AND for coincidence detect takes a few tens of nanosec, and feeding its output to Arduino is the first step to speed things up if naked Arduino cannot keep pace with your chamber.

If not enough, then use continuous recording, which requires fast camera and enough memory to hold all the video until your Arduino can figure out what parts to keep.  And a way for Arduino to control all of the faster parts.

Its a difficult problem, but most things worth doing are not easy.  Good luck on your project.


Out of interest, what is the maximum latency you can suffer without missing these events?
"Pete, it's a fool looks for logic in the chambers of the human heart." Ulysses Everett McGill.
Do not send technical questions via personal messaging - they will be ignored.


Thank you all for your responses.

I have looked into this and the ion recombination time is on the order of 5 microseconds. Is this long enough for the arduino to perform coincidence/anticoincidence of 3 scintillators? If so, then we have the two options; continuous recording or otherwise. What would be the best choice of camera to interface to the arduino for this?

WillR: I am not sure that I understand your question. The ion trail sticks around for a little while after a muon shoots through the chamber - we trigger the pulser and the camera very shortly after in order to see the spark.



Mar 30, 2011, 03:30 am Last Edit: Mar 30, 2011, 03:38 am by WillR Reason: 1
That gives you time to execute 80 instructions. My answer is no.

I should be more clear -- 80 single byte instructions...
access 40 integers or whatever..

The time is a bit short. 

The Mega is a 16MHz Clock. So the period and the available time give you the maximum number of instructions you could execute -- or the memory fetches.

The newer ATMEL 8 bit processors with a 32MHz clock -- maybe.

The ARM (32 Bit) versions of ATMEL with 32MHZ clock and up -- maybe. It depends on the code and you did not provide any or an algorithm -- so it is just a best guess.

I would love to see the Arduino style of software on the 300-800MHz embedded systems -- for projects like yours and mine I would happily pay the $200 to $500 these systems cost. It would be a deal for really fast turnaround on the project in terms of bread boarding and testing. I would buy one today...

Go Up