Now looking at a laser engine to do the ball illumination-

Found this beast in China for $1,300.00, its 300 mw of coaxial RGB laser beams that can be externally modulated, so PWM dimming will be attempted.
The opto-mechanical aspects are probably not great, but this application is not too critical as far as beam geometry is concerned.

http://www.cnilaser.com/

Does anyone have experience with this type of laser engine?

You said there will be 64 of these tubes? $1300 for illumination hardware each? This project is getting into some serious money. What is the total budget for the project? Seems like it could be $250 to $500k range (or more)?

 

300mW laser seems overkill and dangerous . Methinks you'll be drilling holes in the ping-pong balls .

 

At this stage, I am developing a proof-of concept prototype for approval by the client.
Yes, it's potentially a very expensive project, but some of these things do get built.
The laser is indeed overkill, but the unit accepts a modulation input so we can determine what power level is really necessary for the final application, very hard to guess at these parameters.
We start with overkill, and work our way back to plain old "kill"

Attached is a photo of my mock-up built inside a piece of 40 mm PVC pipe, along with the analog front end I am designing. The mic and speaker are both glued inside the tube.

The analog portion is doing the nasty hard stuff to provide the MCU with a simple pulse to measure.
The tricky part is the fact that the echo amplitude drops as the ball moves to the far end, I need to track the amplitude of the pulse to set the the detector threshold correctly, so that it always triggers on the correct portion of the returned waveform, regardless of amplitude.

I am using two peak-hold circuits to capture the amplitude of the return echo, one is used to store the peak value from the previous pulse, while the second captures the amplitude of the current pulse. The peak-holds are alternately reset every other cycle, the trigger amplitude is always derived from the last pulse, so the trigger level tracks the signal level.

 

We start with overkill, and work our way back to plain old "kill"

Pretty much how I do my research for most projects... otherwise things take a year to complete

 

Some thoughs on air valve:
Thought that it might be possible to make a valve with readily available parts. like 1 1/2 " & 1 1/4" PVC pipe. Turned down 1 1/4 pipe to make a snug easy fit into larger pipe. Idea is for servo to rotate part B inside of pt. A. Darn-- time ran out-- must pick up my date.
Note, ID pt A is 41 mm

 

What is the spacing between tubes? Unmarked dimentions are mm.
Pt H, not shown is a mounting ring, 48.5 mm ID X 76 mm OD X 4 mm thick. It slips over bottom of A & glued in place.

 

Some thoughs on air valve:
Thought that it might be possible to make a valve with readily available parts. like 1 1/2 " & 1 1/4" PVC pipe. Turned down 1 1/4 pipe to make a snug easy fit into larger pipe. Idea is for servo to rotate part B inside of pt. A. Darn-- time ran out-- must pick up my date.View attachment 80459
Note, ID pt A is 41 mm

Bernard,

I am working on an implementation of a valve, based on your idea - with minor differences.

The valve will be mounted on a "T" in the lower section, so the laser can shoot axially up the bore of the main tube.
The valve will have 50% overlap between the two ports, so it will have a proportional output with no 'dead' travel in the middle - avoid servo 'hunting'
The valve will have small diameter central shaft to reduce the bearing surface, there will be clearance between the inner rotor and the housing to minimize friction.
The rotor part will just be a "C" shaped section of cut tubing, just enough to block the ports with minimal extra mass.
This whole valve contraption will get mounted inside a box lined and baffled with foam, to damp sound. The sound damping helps keep the pinger sound from escaping and also reduces the fan noise coupled into the microphone- the fan is quite loud, and this introduces timing jitter into the position measurement.

The tubes will be spaced far apart, so no need to worry about clearance around the tube.

Thanks for the clever idea!

 

How long are the tube sections? I just assumed about 5 ft.
Just some thoughts on joining tubes: A sleve would be distracting, there are some glass glues- how strong?, then there is heat fusion. I'm no expert, last time that I joined two soft glass tubes was back in 1948, glass blowing Christmas present. If tubes were rotated with precision alligned glass lathe it might be possible to preserve ID with a plug cast from lost wax investment material. Plug normally is a tight sliding fit but if it becomes stuck, it can be disolved with water.

 

The client says that he can get pre-joined tube sections made, presumably joined on a glass lathe and annealed - somehow, thank god this is not my problem.

Attached is the schematic of the analog portion of the system, what I have built so far.
The analog ping pulses come from a 5 bit DAC in my MCU, I am using a PIC18F46K22.
The digital interface consists of two BLANKING outputs and one RANGE_PULSE that comes back.
The MCU also generates the pulse train to control the RC Servo, and three PWM outputs for the laser.

Only hardware details I have to add now are the DMX receiver and laser PWM outputs. The laser seems like it takes an analog input, so I will put the PWM into a pot, so I can vary the power level while still retaining 8 bits of PWM resolution.

The PID code is from an old motor control project, I just need to fiddle with it a little to format the input and output, then maybe add an SPI DAC to monitor what is going on in there.

 

Here are a few pics of the valve that I built for controlling the ball.
The valve slides into the airbox like a cartridge, so I can get at it.

Got the thing working, tricky part is feeding it with the correct pressure and flow, the input air has a large impact on the system gain and therefore ball stability and performance.

 

The project is finally complete, here are a few pictures of the contraption.
The laser engine is total overkill, but it makes the effect spectacular. At 300 mW per color, I have to sand the printing off the ball so the back ink doesn't absorb too much power and melt the plastic.

 

That's awesome! Any chance of video, whether now or after the actual installation is complete? Also, this seems like a great project for the "Completed Projects" forum if you're willing to share the relevant details. Either way, congrats!

 

Thanks for taking the time to let us know how it turned out. We always like to hear the end of the story. This thing looks awesome and I can't think of anything close we've ever seen around here.

 

We want video!!!!

 

Video is Here:

 

Impressive how smooth it works. Congratulations!

Laser color:
depends of the height?
The same for all tubes?

What about the noise of 64 of these in action?

Have you tried a different colour for the balls?

 

Cool, I want one! An array of these would be as mesmerizing as the Belaggio fountains in Las Vegas.

I guess a mirrored ball would blind everyone?

 

I guess a mirrored ball would blind everyone?

I'm guessing that even the white ping pong ball is a little too rough on the eyes with a 300mW laser lighting it up!

 

Impressive, great control. Thanks for report.

 

Thanks all for helping to make it possible.
It was a very interesting project with just the right amount of challenge to be really fun.

The videos show just 2 sections of glass tube, (3 meters) it will have far more impact when implemented at the final 6 meter length.
I wrote a simple test program using LCedit (DMX lighting controller) to move the ball around and change the colors; DMX Channel 1 controls ball position, 2 = RED, 3 = GREEN, 4 = BLUE.