Experiments in High Speed Digital Photography
|
What? |
Aron & Justin over at
www.hvguy.com figured out a way to take
incredible high speed photographs using a digital camera and some homebrew
strobe circuitry. Go check out their work. Their work inspired
me. |
|
How? |
High speed
photography involves exposing a recording media such as film or a CCD to an
illuminated scene for an incredibly small amount of time. There are two general
approaches. The first involves using a mechanically fast shutter - much faster than
the shutters of normal cameras. The second method involves leaving
the shutter open in a totally dark environment and turning on a light for a
very brief amount of time. We'll focus on that approach here. |
|
Why? |
In the infamous words of my hero Larry Walters, "A man just can't sit around!" Seriously, if something deep within your being doesn't innately understand why this is important, and must be done, there is little hope that my words will make a difference... |
What sort of speeds are we talking about here?
Good question. For starters, let's try to take a look a .22 rim
fire rifle bullet. These little guys are slightly subsonic traveling at
about 1200 feet/sec. I figure that in order to get a decent picture of
one, it should not move more than about 1mm while the strobe is on. My
rough calcs show that the strobe duration needs to be on the order of 0.25
microseconds.
Time = distance / rate
Time = 1 mm / (1200 ft/sec * 254
mm/in * 12 in/ft) = 0.27 uS
So, how do we go about making a 1/4 microsecond light source?
OK, let's get started!
TRIAL 1 - Modified Xenon Strobe
Aron and Justin had pretty good results using a modified xenon strobe so I
figured that would be a good starting point. I realized that I had most
of the supplies already laying around in my High Voltage lab. You do have a HV
lab, don't you? For starters, I dug up a strobe given to me by my buddy
Kevin Ottalini. The first step was to make sure it still worked. The
strobe was easily verified by ripping the flash circuitry out of a disposable
camera, discarding the little weenie xenon tube it came with, and replacing it
with my big-daddy xenon tube. See the following picture.
Although the flash brightness was nothing to write home about, everything functioned just fine. It was reassuring to see that the strobe would fire with the little trigger transformer and the 300VDC that's stored in the photoflash capacitor. The next step involved getting a clue as to what sort of capacitance, voltage, and energy levels are required to actually get good picture of high-speed stuff. Justin clued me in that 6J creates enough light for his pictures. The energy seems to define the amount (brightness?) of light created. The time that the light stays on is dictated by the inductance, resistance, and capacitance. To keep the time constant low, you want as small of a capacitor as possible that still has enough energy to produce the required amount of light. The first order of business was to figure out how my xenon strobe behaved under high voltage. As the picture below shows, I rigged a small variable DC supply to get an idea about how everything behaves.
Self Trigger = depends on main capacitor size, somewhere around 3.5 KVDC for
100nF cap pictured
Retrigger after fire = has problems with this above 2.5KVDC, may be due to
shoddy mechanical trigger switch
Reliable operation = 100nF main cap charged to 2 KVDC, 2.32 KVpp (770 Vrms) out
of Royer, 4 VDC into Royer
This is only 0.2J so it's not a very bright flash. I was excited to find
that the flash empties the capacitor in 2 uS. This isn't terribly far from
my goal of 0.25 uS. Kevin's Royer pumps the cap back up to 2KV in 125mS.
New data! 3/18/03
| Notes | Capacitor (uF) | Voltage (VDC) | Energy (J) | Discharge Time (uS) |
| Flash assy from disposable camera | 160 | 300 | 7.2 | 600 |
| 8" xenon tube, one green cap as shown above | 0.1 | 2000 | 0.2 | 2 |
| 8" xenon tube, four green caps | 0.4 | 2000 | 0.8 | 3 |
| 8" xenon tube, 3 of Reed's caps, damn bright! | 500 | 500 | 62.5 | 1500 |
The experiments above were worth while because they gave me a feel for discharge times and forced me to learn a lot about how strobes work. It's obvious from the data that shorter discharges time are had with smaller capacitors. We need a certain amount of brightness from the strobe and this has to do with the energy that we dump through it. Needing higher energy and smaller capacitors leaves us no options but to increase the voltage. This can be done until the xenon strobe self-triggers or retriggers after a successful firing. Further reading suggested that even if I find the perfect balance, xenon simply glows for too long after being energized and will always give me multi-microsecond long pulses.
TRIAL 2 - Improvised Air Flash Lamp
Jim Lux, Scott Hanson, and several of my HV buddies pointed out resources explaining that much faster flashes can be had by dumping a HV capacitor across a spark gap placed on the surface of a quartz tube. (Apparently Edgerton figured this out 40 years ago. I guess a little research is in order before firing up a soldering iron!) Perhaps the best explanation of a homebrew short duration strobe is given in a column for The Amature Scientist in August of 1974 on Schlieren Photography. That column is shown in the link below. http://hot-streamer.com/bob/AmSci01/1974/08/1974-08-fs.html Another option is to use an EG&G 549-11 "Microflash" electronic flash unit. It's capable of producing flashes in the 0.5 uS duration. I haven't stumbled across the technical details of this unit, but I assume that it uses an air spark to get its fast times. Not having a "Microflash" unit handy, I'll get started on the device detailed in the Amature Scientist column.
...Searching for components... more soon...
Useful Links
http://www.edgerton.org/biography.html - The dude who started it all!
http://www.rit.edu/~andpph/text-agfa-1280-hs.html - a professor at RIT does high speed photography with digicams for his lab!
http://www.geocities.com/CapeCanaveral/Lab/5185/circuit.html#SOUND - sound triggered flash circuit
http://www.xtrsystems.com/strobe/strobe_light - this guy fooled around with strobes a bit
http://www.dewtronics.com/tutorials/lasers/leot/index.html - excellent LASER power supply stuff
Please let me know if you see any errors or have any ideas on how I can improve the performance of my strobe system.
[ Back to Tesla Coil Research Home | Contact Ross ]