Reflector to enhance PING))) performance available

Ed T

Fellow PING))) users,

While using the PING))) in a project I discovered that I needed performance better than the PING))) could provide. I searched the forums hoping to find a reflector that would focus the PING))) but found that others were in search of the same thing and there was nothing available.

I also discovered problems people had in using a standard circular parabolic reflector typically seen in use with microphones at sporting events. The problems included; the direct reflection from the reflector to the PING))) caused it to trigger, and the PING))) blocked the central, highest signal part of the sound beam.

I tried a bunch of different reflector designs and found that a cylindrical parabolic reflector with the PING))) pointed off axis does provide greatly improved performance of the PING))). This shape focuses the sound in one dimension (e.g. vertical) while leaving it unchanged in the other (horizontal). This concentrates more of the sound on targets at the same level as the PING))) as well as gathering more of the return signal onto the sensor. It should
also reduce the sensitivity to ground reflections. With no focusing in the horizontal dimension sensitivity to objects to the left and right of the sensor is maintained. The cylindrical parabolic shape also has a focus that is a line, instead of a point. This allows both the transmitter and receiver of the PING))) to be at the focus of the reflector.

By pointing the PING))) off axis (i.e. not at the vertex (center) of the parabola) it does not block itself from transmitting or receiving sound in its "sweet" spot and also does not trigger off of direct reflections from the reflector.

To make this available to others I had a tool machined and have found a plastics manufacturer that will vacuum form the reflectors.

You can order them at www.Perbolia.com. They ship from Amazon. Hardware to mount the PING))) to the reflector is included.

(Also now available at Gadgetgangster)

The current design size (approx 3"x3"x2") is one that significantly improves performance while not being too bulky. It is possible to make them wider on the same tool or taller on a new tool. So if there is enough interest in other sizes I could make those available also.

I hope other PING))) users will find this a useful addition to the already great PING))) sensor.

- Ed

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
___________________________________________________
www.Perbolia.com
Home of the Cyperbolic Reflector for the PING)))

Post Edited (Ed T) : 7/8/2010 2:47:32 PM GMT

Phil Pilgrim (PhiPi)

'Nice out-of-the-box thinking there, Ed! As you stated, the on-axis reflectors don't work. I know: I've tried. I ran into the same problems with an acoustic lens based on a Fresnel zone plate design.

As a service to the forum, do you have a side-view drawing of your reflector you can post that shows the wave trajectories? This will help people to understand what it is you've accomplished.

-Phil

Post Edited (Phil Pilgrim (PhiPi)) : 7/8/2010 3:29:01 PM GMT

Ed T

Phil,

Thanks for the kind words.

Per your suggestion, attached is a side photo with some sample paths drawn in. As shown in this view the reflector's parabolic shape in the vertical direction causes sound that comes from the PING))) (placed at the parabolic focus) to have much reduced divergence in the vertical dimension. The opposite occurs for the sound returning to the PING))). This sound that hits the reflector traveling parallel to the base is all reflected onto the receiver (it also has to horizontally reflect into the sensor).

In the horizontal direction the reflector is flat so that the divergence does not change in this dimension.

Note that the focus of the flat/parabolic shape is a line which the PING))) can be centered on but some of the transducer is slightly off the focus line which causes the focusing to be less than ideal.

I have also attached a graph of the measured performance of the PING))) with reflector versus the standard without.

Edit 7-10-10: added graph showing performance in other dimension.

- Ed

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.Perbolia.com
Home of the Cyperbolic Reflector for the PING)))

Post Edited (Ed T) : 7/10/2010 5:55:18 PM GMT

Phil Pilgrim (PhiPi)

Very nice! Now, could you repeat the cylinder experiment with the ping/reflector assembly on its side (i.e. rotated 90 degrees)? It would be nice to see how well the enhanced directivity works on that axis.

-Phil

Ed T

Phil,

That is a good question. I have not done that test specifically. Give me a few days.

- Ed

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.Perbolia.com
Home of the Cyperbolic Reflector for the PING)))

Martin_H

That is really neat.

Whit

@Ed T

Great idea! Does the LED indicator reflect off the Perbolia enough to be seen as well?

The design keeps the wiring out of the way as well and the Perbolia provides protection for the Ping))) too.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Whit+


"We keep moving forward, opening new doors, and doing new things, because we're curious and curiosity keeps leading us down new paths." - Walt Disney

Post Edited (Whit) : 7/9/2010 2:41:01 AM GMT

Ed T

@Whit - Yes the LED is visible in the reflector, but it is only partially reflected, so it is not as bright as viewing directly.
If more reflectivity was needed some reflective material could be applied to the back surface, or possibly the front surface, if it did not degrade the acoustic reflectivity.
- Ed

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.Perbolia.com
Home of the Cyperbolic Reflector for the PING)))

Post Edited (Ed T) : 7/9/2010 12:59:22 PM GMT

agfa

Nice work Ed. I‘ve been dabbling with sonar directivity, myself. My goal is to use transducers at different locations, one at a fixed location with some type of omni directional reflector, and another on a moving platform with a reflector for the purpose of direction finding. I was thinking of using an ADC to measure the received signal of a sweeping transducer, then determine the direction based on the strength of a received signal.

I’m curious about how you collected the data for the graph, and what environment you performed the test in. I’ve been testing by using an oscilloscope to measure the amplitude of a receiver, but I’ve been getting inconsistent results. I’m having a hard time just identifying non-reflective surfaces/materials.

If anyone has any suggestions·I would appreciate it.

Thanks

agfa

Ed T

@Phil - I have added a graph to the others above that shows the results of testing with the reflector on its side. It is much narrower as expected.

@agfa - I am testing using the Parallax PING))) sensor not bare transducers. I drive the PING))) with a Propeller demo board that reads the distance info from the PING))) and displays it in a VGA window on my PC. So the PING))) handles the raw transducer signal. I have a cleared area where nothing is sensed by the PING))) when the cylinder is absent. At a given distance I move the cylinder in from the side until the PING starts seeing it. This makes a point on the graph. I believe the PING))) increases the gain on the signal the longer it waits for the return pulse. The received signal from the transducer can be very small so it needs to be amplified and filtered. You may also have a challenge with the high variability of the reflectivity of various objects and the angles they are at. For example an object with low reflectivity on center may have the exact same signal as a highly reflective object that is off at an angle. Maybe there is a way to use the data from all the angles you scan at to determine what is where with what reflectivity.

- Ed

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.Perbolia.com
Home of the Cyperbolic Reflector for the PING)))

Phil Pilgrim (PhiPi)

Ed,

That's a nice improvement in directivity! I suspect the "sideways" orientation will be how most people will use your reflector, since a narrower horizontal field of view is what the majority of related forum posts have been about.

-Phil

Ed T

Phil,

I see your point. Using the reflector on its side would be a big help for the type of thing AGFA is trying to do.

- Ed

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.Perbolia.com
Home of the Cyperbolic Reflector for the PING)))

hover1

And I like having the option!

Nice effort, Ed

Ed T said...
Phil,

I see your point. Using the reflector on its side would be a big help for the type of thing AGFA is trying to do.

- Ed

Martin_H

The second graph is really interesting. The compression of the beam results in greater range and finer resolution. A win-win if you mount the Ping sideways on a servo. But would floor reflections become more of a problem given the greater sensitivity in the alternate red axis graph?

agfa

Thanks Ed.

proximityinfotech3

As shown in this view the reflector's parabolic shape in the vertical direction causes sound that comes from the PING))) (placed at the parabolic focus) to have much reduced divergence in the vertical dimension. The opposite occurs for the sound returning to the PING))). This sound that hits the reflector traveling parallel to the base is all reflected onto the receiver (it also has to horizontally reflect into the sensor).

_______________________________________________________________________

Post Edited By Moderator (Dave Andreae (Parallax)) : 7/19/2010 10:01:38 PM GMT

Ed T

Hi All,

The reflector is now also available at GadgetGangster:

http://www.gadgetgangster.com/find-a-project/56?projectnum=361

- Ed