Water drop setup

[rick55]

I got my setup reconfigured, and there are a couple of things others might like to know.

I rebuilt the setup to make it easier to adjust the various pieces. As you'll see, it's overkill for the one-trick pony of water droplets, but I think it will be great for a lightbox. I apologize in advance for most of the measurements being non-metric, but in the case of attaching a flash bracket, it has to be.

The key piece is this clamp that I got from Amazon. They're unavailable on Amazon at the moment, but here's a link to the manufacturer's page, with a reference to another source.



It clamps onto two rods up to 5/8" in diameter, which is just under 16mm, so I'd guess 15mm is the limit. I don't know what metric sizes of threaded rod are easily available. I used 3/8" rod: I imagine 10mm would be equivalent. As you can see in the picture of the clamp, I have a coupling nut and wing nut (to lock the coupling nut) that the clamp grabs, to avoid messing up the threads. There are probably 3 choices: 1) get plain rod, and thread the last 2 inches; 2) use coupling nuts like this; or 3) just clamp to the threads, and don't worry about the damage. If you'll always use the clamps, and never thread anything onto the rods, (3) should be fine.

The overall setup looks like this:



The sensor boom is out of view at the top. The flashes are set up on pieces of 1/4"-20 threaded rod with coupling nuts. This is where you have to use inch size.

The base is just 3/4" plywood with a pair of 36" rods in front, and a pair of 24" rods in the rear.

So with being able to adjust the height, and tilt and rotate the booms that the flashes are on, it's very adjustable.

The sensor is on another boom, with a homemade adjustment at the end:



You can see a couple of extra holes in the aluminum bracket: that's because I originally had the blocks that hold the hose at the top, which was stupid. It meant that I had to fiddle with it to make sure the drops fell through the sensor. This way, it has to be pretty far off level to be a problem.

The sensor (datasheet here) is now working quite well: it consistently triggers when a drop leaves the tube, which means my shutter timing is much more reliable. You wouldn't have to hack out aluminum the way I did: just put the sensor up on a couple of #4 spacers so that the wires have room to get out, and put the tube on a block that's a little farther from the aluminum. On the electronics side, I'm reading the output with the Arduino, but you could use a comparator IC to go directly to a relay. There'd be no delay, but doing some continuous firing might work.

The other thing I wanted to pass along was that I wasn't getting the "ball on a spout" images I wanted. Looking back at suggestions, I saw two possibilities: a deeper basin, or a rinse aid. I tried the deeper basin, and it made no difference. Here's a typical spout from the deeper basin:



It looks about the same as I was getting with the paint tray.

So I added rinse aid, and sure enough, that improved things. However, watch out for how much you use. I found a page for a commercial rinse aid (here), and they specified 0.3ml/l. When I added that much of an off-the-shelf product (Rinse-Aid), it made things a little difficult with bubbles in the water, as seen here:



My recommendation would be to start with a smaller amount, and see if you can get the effect you want, but fewer bubbles. I probably won't have any more time to fool with it this week, but I will work some more with the rinse-aid stuff.

I posted a couple of results here: they certainly don't look any better than what Jeroen has been posting, or what Jim has done. The advantage of this setup is the convenience. I just checked back, and as an example, when I was working with the rinse-aided water, in a series of 63 exposures, 41 of them had drop-to-drop collisions. It takes 10-20 shots to get zeroed in, so you figure that's a good 50% success rate.

Cheers,
Rick

[carregwen]

Rick

You seem to have this well sorted, even better than Ash over on Talk Photography forum. I must confess I'm fazed by the electronics. I've ordered some of the aquarium pumps that Ash uses to get the drops at a better hit rate. One question, and this is probably a dumb one, how critical is the height of your rig that the drop has to fall? If there's an error margin (you said you don't get 100% hit rate) surely a higher distance will mean less error margin, or have I got that wrong?

We await some stunning shots...........

[rick55]

Thanks, Rob. Some shots (not stunning, but cute, I think) are here.

Not dumb at all: there would probably be less margin for error with the outlet of the tube lower, although for me, it seems to have to do with the drop spacing. Since I'm just pumping two drops, I think if one is kind of "hanging there" from the tube, it drops off quickly, then there's a longer pause until the second drop, even with the same pump speed. My current rig has the outlet about 20"/50cm above the water when using the deep basin, 24"/60cm above when using the shallower tray. I've had it at various heights during this stuff, but I see little difference in behavior, which surprised me: I expected it to be pretty important, since I assume the energy the drop contains when it hits the water controls the splashback. I guess what's happening is that the distance has to change more. The velocity goes with the square root of change in height, then the energy is proprotional to square of velocity. So you need a 50% increase in height to hit with 50% more energy.

On the sensor, I'd like to see if I can do it with a comparator and timer IC. If that can be done, it would be pretty simple.

Cheers,
Rick

[rick55]

Jim just asked a question, and I realized I had a mistake here. For the pics in the other thread, I have the flashes rotated 90 degrees from what you see here. That let me pull the two colors closer together, and spread the light better front to back.

Cheers,
Rick