Synchronization of a camera with a rotating shaft

[ishaan]

I wish to synchronize a thermal camera FLIR A300/600 with a test section on a rotating shaft such that whenever the section becomes orthogonal to the line of sight of the camera, an image is clicked by it automatically. The camera uses a TTL pulse input. Being a mechanical engineer, I have little idea as to the electrical components involved. However, I have thought of placing a proximity sensor on the mutually orthogonal axis in front of the section, such that whenever it detects an obstacle(the test section), a pulse is sent to the camera. I wish to confirm whether this will work before actually implementing it on the expensive camera. Also, if someone has used a similar setup can guide me as to what all settings are needed in camera software and whether the pulse output by proximity sensor(PNP) is in compliance with the camera.

 

[berkeman]

I wish to synchronize a thermal camera FLIR A300/600 with a test section on a rotating shaft such that whenever the section becomes orthogonal to the line of sight of the camera, an image is clicked by it automatically. The camera uses a TTL pulse input. Being a mechanical engineer, I have little idea as to the electrical components involved. However, I have thought of placing a proximity sensor on the mutually orthogonal axis in front of the section, such that whenever it detects an obstacle(the test section), a pulse is sent to the camera. I wish to confirm whether this will work before actually implementing it on the expensive camera. Also, if someone has used a similar setup can guide me as to what all settings are needed in camera software and whether the pulse output by proximity sensor(PNP) is in compliance with the camera.

Welcome to the PF.

You should be able to find a proximity sensor that outputs a TTL pulse. You can use some of those words in a Google search to see what pops up. I like to use Google Images for an initial search for things like sensors, since you can scan the results visually much faster than clicking into a bunch of links from the search.

Can you attach a picture or two (or link to them if you have them on a hosted website)? It would help to see what you are working with...

 

[ishaan]

Image

Here's a rough sketch of the circuit I am hoping to use

 

[mjhilger]

I have not used proximity sensors in a long time. But since you did not mention it, you need to verify that the speed of rotation of the shaft will work for the sensor chosen. Most shaft encoders use optics with a disk on the shaft which has window openings, in your case only 1 would be necessary. Then you might have to advance your sensor location (whatever type is used) because of any delay in the image capture from the pulse. This could be dependent on the speed of the shaft depending on the speeds involved. Big difference between 300 rpm and 10,000 rpm in sensing and capture. So if you are looking at covering a wide range of rotational speeds, you might need something to adjust the timing.

 

[berkeman]

Here's a rough sketch of the circuit I am hoping to use

I don't see any need for the opamp. The output of the AND gate should be sufficient to drive the TTL input of the camera.

 

[Ryoko]

One thing that needs to be considered is the speed of the camera. The models cited use microbolometer arrays which are not all that fast to respond to temperature changes. A spinning shaft may be a bit too fast for them to capture in any detail.

 

[meBigGuy]

You might want to consider hall effect sensors if you can attach a magnet to the shaft.

 

[Viggo]

One thing that needs to be considered is the speed of the camera. The models cited use microbolometer arrays which are not all that fast to respond to temperature changes. A spinning shaft may be a bit too fast for them to capture in any detail.

I agree. All cameras require exposure time to capture photons. And each pixel on the sensor may have different timing.

 

[Ryoko]

I agree. All cameras require exposure time to capture photons. And each pixel on the sensor may have different timing.

It's not so much that each pixel might have different timing, but rather that microbolometers are in essence tiny thermometers and like thermometers, they don't respond instantly. So a spinning shaft will likely be seen as just a blur even if the camera is timed correctly.