In a recent comment to this thread (
Looking for some help setting up a Cactus laser trigger setup - PentaxForums.com), BrianR asked if I had measured the delay of the Pentax AF-540-FGZ when used as a slave flash:
Originally posted by BrianR
have you ever tested the response time of the 'dumb' slave mode? It looks like SL2 mode on your flash (page 52 of the manual) is a dumb slave that responds to other flashes.
Now I have.
So, here are some more flash results that might be of interest to those trying to do stop-action shots (with and without slaving).
I have on hand a pair of really cheap Neewer flash units that I got some years ago on Ebay for about $10 each. These can be used in slave mode. I also have an old Pentax AF200SA flash unit, which is not slavable, but has a light sensor for control of flash power. And, of course, a Pentax AF-540-FGZ that can be used in slave mode.
The Neewers feature pretty short flash duration duration. Here’s what the time profile looks like for one of them. The effective duration is about 100 microseconds.
Here’s what the profile is like for the AF200SA - first at full power, and then in sensing mode. The full power blast is about 800 microseconds long, but in sensing mode, the flash can be very short - just 20-30 microseconds or so.
What happens when I try to trigger the AF540? Here’s there result, using the AF200SA as the triggering flash. It takes somewhat more than 100 microseconds for the AF540 to start firing, and the light output does not peak until some 140 microseconds after the trigger flash starts to fire (for the AF540 at 1/32nd power).
The Neewer used in slave mode fires substantially quicker than does the AF540. The Neewer starts firing within about 50 microseconds of the beginning of the trigger flash, and the peak light output occurs just 80 microseconds after the initial firing of the trigger flash.
What does this mean for Bobbotron’s initial goal to watch his crossbow arrows slice though water balloons?
For single flash guns, light curve durations from a few tens to 100-200 microseconds should be relevant. He reports that his arrows should be moving at about 400 feet per second (~120 meters per seconds; why is somebody in Canada using that old-fashioned “feet” unit

!? ) So, in 100 microseconds, the arrow would travel about 0.04 feet (half an inch) or a bit more than a centimeter. Unless you really want to see the arrow stopped, this seems good enough to watch the balloon explode.
Since the slave delays are also around 100 microseconds (perhaps plus/minus a factor of two), similar distances (a centimeter or so) would occur for multiple images of an arrow illuminated by 2 or more flashes slaved together. Whether that is a problem, or just a nuisance, or immaterial, I leave up to the user!
(In all the above scope images, the relative peak amplitudes of the flash curves are NOT related - I was pointing the flashes as necessary to keep the intensities within the linear range of my photodetector.)