So is the plan for the 4kHz technique to output a series of short bursts of 4kHz and the measure the return time of the echo?

 

Hmmm... I didn't realize the ball would be illuminated, which would reveal the monofilament unless the laser is at the opposite end of the tube. I would also worry about the longevity of 64 units although I think you could deal with that. Large swings in position would cause problems if the filament is not kept taut. Just like fishing, you can't let your line out too fast.

What precision are you looking for? Position ± 1", 6" ?

 

Are you familiar with a soliton?

If you could figure out how to generate one at the base of the tube, defraction/dispersion may not be an issue.

 

So is the plan for the 4kHz technique to output a series of short bursts of 4kHz and the measure the return time of the echo?

Yes- this is my general idea. I can place the ranging components in a longer 'dead' section of tube to avoid the minimum detection range imposed by the sonar technique,
this seems do-able as long as I can reliably detect pings with a small number of cycles. The acoustic impedance of the ball is also a question, can I get a good return echo signal from a ping-pong ball?

This concept gives me the best configuration, I can easily shoot the laser down the bore without compromise, and the system has minimal mechanical components for high system reliability.

I am fascinated by the concept of the "Soliton" never heard of this before- can you elaborate on the idea?

 

A video is worth 10,000 words:

Can this be done with air pressure? No clue.

 

I expected a standing wave? (Which can certainly be done with sound.)

 

PID?

 

Some thoughts if string option flies: Winch drum = 1 ft / rev. then 15 deg, step would advance ball 1/2 in. Adding threaded rod to shaft could supply top & bottom limit switches or maybe an absolute code strip allowing any desired resolution & allowing a variety of different types of motors? Lift on top, light on bottom, all tubes connected to neg. pressure manifold requiring less pressure then pos. pressure.

 

....all tubes connected to neg. pressure manifold requiring less pressure then pos. pressure.

Suck 'em instead of blow 'em? Seems fine, but what's the advantage?

Oh, you mean you don't need equipment rated to 30psi for example? Natural cap at 15psi.

 

Keep enough pneumatic 'preload' pressure on the ball to keep the string taught under worst case acceleration conditions- I like it.
Only trouble I see is recovering from unexpected power fails:- Tangled Mess.

 

If pressure rises, shut it down, no tangles.

 

Good idea, Like the old tape drives with the vacuum columns.

Edit:

Stealing another idea from tape drives. You could then put a tach (quaduture) on the drive motor and go for Absolute position in the tube.

 

More ramblings: If one inch resolution is acceptable then location can be expressed in 8 bits. An 8 track for linear code track or rotary code wheel & revolution counter coupled with digital comparator would allow MC to issue tube #, ball location command & speed command, each tube control would then handle the mechanical operation. Less book-keeping for MC??

 

A spool driven via a closed-loop servo or stepping motor could drive the monofilament and ball with great precision and high speed slewing performance.
Some type of home sensor would be required to reset the system on power up, to calibrate the absolute ball position. As long as there was always some kind of preload keeping the string taught, it would work well.

 

Progress update:

Received the tubes, balls and some transducers, set up a simple rough test of the 'sonic ping' rangefinder idea, the results look promising.
I taped a long stiff wire to a ping pong ball so I could slide it up and down inside the tube, placed a piezo speaker and an electret condenser microphone at one end of the tube.
Using my signal generator and scope, I can experiment with different burst lengths and frequencies. It's interesting to see how strong the reflection is even just pinging the open tube end, the open end reflects a pulse almost as strong as when I cap it with a hard surface, but with phase shifted by about 90 degrees, which makes sense. It looks like I can get this working with a PIC micro-controller using the capture-compare module to measure the echo return time, just need to put the mic signal into a comparitor to get a clean TTL pulse output. Joining the tube sections is a pain, if the joint is not really clean, I see reflections from the joints, which might eat some of the return energy and pollute the signal. I also need to play with the pneumatics a bit to make sure the sounds generated by the fans / ball / airflow do not kill the SNR.


The first picture shows the ball close to the transducers, the second is with the balls farther away. Top trace is the generator output.

 

Cool!

 

A rough measurement of ball in free fall in tube came out as 1.5 ft/ sec. Maybe 30% difference if bottom of tube was open or cloced. Is this fast enough?
A photo reflective detector at bottom of tube might give ref. position-- if needed.

 

...................
Using my signal generator and scope, I can experiment with different burst lengths and frequencies. It's interesting to see how strong the reflection is even just pinging the open tube end, the open end reflects a pulse almost as strong as when I cap it with a hard surface, but with phase shifted by about 90 degrees, which makes sense. ....................

Rather like the reflection from an unterminated electrical transmission line.

 

Next steps are to build an air box with a DC fan and figure out how to introduce the air without disturbing the sonic properties of the tube setup.
I think I will try drilling a number of holes along the 'dead' section of tube, just above the mic and speaker, so the air does not flow past the mic, which could introduce noise.
I will build a box manifold that covers the holes on the outside of the tube, with fan attached.

Instead of using fan speed control as most attempting this, I will instead have the fan going full tilt and utilize a simple air valve controlled by an RC servo.
The valve will be proportional, going from full fan, to full vent, so the air flow in the tube can reverse - for faster ball drop, might try two fans, one sucking and the other blowing.
The servo valve should give me snappy control over the airflow, with no fan inertia to contend with.
I think I can measure the ball position every 50 ms, the servo takes 100 ms to move 90 degrees.

I ordered a 4 channel SPI serial DAC that I will use as an output to monitor the Position, Velocity and Command points with my scope.

 

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?