Project feasibility - acoustic target system

Hi all, Im considering a new project and would like some feedback/ideas. I am avid sport shooter and i am always looking for cheaper ways to practice. I am toying with idea of making an acoustic target system similar to this: https://www.silvermountaintargets.com/
For use with airsoft rifles.

Its basically a target with four acoustic sensors that triangulates a bullets impact on a target. What i am thinking is a projector type screen, with 4 sensors on each corner. I can project a game like trophy hunter or something simple and use the acoustic triangulation to essentially plot a mouse click where the airsoft bb hits the screen.

The range i go to uses that solo sytem so i know the theory and accuracy are possible. Im just not sure if a diy version feasable. I have decent electronics knowledge and the code aspect will not be an issue.

A couple of question:

  1. Do you think any arduino/rpi models would be able to to measure the time difference in acoustic signals at each corner accurately enough to get decent resolution.

  2. Can anyone suggest an accoustic sensor that would be reasonable to use in this scenario.

Its just an idea at this point but i am curious to hear your thoughts.

Above my paygrade - however to save some others the math - what is the time frame - i.e. sort of minimum/maximum time and resolution required

Arduino's have some fast time resolutions and there are other processors that run quicker - but some time guidance would help others help you

Good luck

I honestly do not know. I have not found much information on these types of systems, all seems to guarded proprietary information. I am really starting from scratch with this, which is why i was hoping someone had an idea of where i could start my investigation. I may have to cobble together a proof of concept and use that to point me in the right direction.

I guess the most pressing question would be what type of sensor would work best to detect a sound wave as it propagates through a canvas projector screen. Simple microphone? Or is there a better solution?

One thing to keep in mind is that sound travels through solids much faster than it travels through air. I am looking at a direction of arrival system for sound through air, ant I think I will need about 50 MHz processors to separate that gap in arrival times from two mics. Find out how fast sound travels trough the target material (might not be easy), Then consider the smallest distance resolution you will need. You'll have to be able to collect several samples from all mics in the time the sound needs to travel that distance. Probably out of the range or Arduino products.

A few thoughts on this.

  1. The "Solo" system at the link says it work with "an incoming supersonic bullet", so I'm guessing it's sensing the shock wave as the bullet passes by, just before it hits the target. The sensors appear to have two microphone elements in each assembly. An AirSoft projectile isn't supersonic, so it presumably wouldn't work with that system.

  2. One approach might be to sense the shock wave as the projectile hits a solid target, e.g. a sheet of plywood. The speed of sound is considerably faster in wood than air (about 10x) so positional accuracy for any time measurement would be that much worse. Without knowing much about it, I'd be concerned about how uniform and predictable shock propagation in wood might be and I'd expect something like a canvas backstop to be even messier.

  3. That said, it wouldn't be difficult or expensive to attach a microphone or piezo sensor to a board and instrument it to see what happens. I'd start with a pair of those cheap KY-038 clap sensor microphones and watch the output with a digital oscilloscope or a USB logic analyzer.

  4. Most microcontrollers have counter/timer(s) with input capture that can capture the time at the instant of an external pulse, generally at a clock resolution up to the processor speed. For this project, one would like a counter/timer with multiple input capture channels on a single timer. I don't believe any of the common AVR Arduinos have that capability. I've worked with STM8 and STM32 microcontrollers which do have a timer with 4 capture channels. There are probably others. Assuming a propagation velocity of 3500 m/s and a 16 MHz counter gives the equivalent of sub-millimeter resolution which suggests time capture isn't going to be the issue. The limiting factors are likely to be sensor bandwidth and peculiarities of sound propagation in the target.

Sound is a wave, not a well defined point signal. The wave is likely to be very noisy to boot, making it even harder to do time difference calculations between measurement points.

With this required resolution I have the feeling it's way out of the realm of Aduinos. It likely needs specialised microphones to correctly pick up the signal with accurate time information, that's no doubt the hardest part. The moment you have a good signal it's a matter of calculations. Still not simple, but I expect this step to be less hard.

I was curious how this might work, so tried the experiment I'd proposed above. For this test I super-glued two small piezo discs to a scrap piece of wood paneling board, soldered leads to the two discs, and hooked them up to an oscilloscope.

Photo shows setup with piezos about 21 cm apart with markings for the midpoint and 3/4 of the way between sensors giving about 10.5 cm differential path length for an impact impulse at that point.

Lacking an AirSoft gun, I struck the board with the back of a BIC pen at each of the three marked points. Striking the center point showed a pair of signals from the sensors that started at essentially the same time as expected. Striking at either of the 3/4 points showed a signal in which the closer sensor's signal started about 116 us earlier than the farther sensor. If I've done the math correctly, this suggests a propagation velocity of about 900 m/s in this setup.

This proof of concept seems encouraging. One would need a gain/limiter stage and some sort of threshold detection to generate the triggers for a microcontroller multichannel time capture, but that's left as an exercise for the student. Adjusting the oscilloscope trigger to get the signal without false triggering was a little fussy and there would be multiple triggers on an event if the single sequence mode was not used.

Striking point 1/4 of the baseline distance from "yellow" channel:

Striking point 1/4 of baseline distance from "blue" channel:

Striking midpoint between sensors:

piezoTargetTest.png

piezoTargetTest.png

Nice experiment!

I wonder though how one could reliably find the start of the signal (as that appears to be what you should be looking for, considering that you don't know how strong the signal will be (the further the strike point the weaker it will be - as your scope signals show), or even the polarity of the signal (sometimes it starts with a negative curve, sometimes a positive, sometimes a tiny bit one way and then it swings big the other way).

Thanks for the input everyone. I went ahead and ordered a few different types of sensors to experiment with (piezo vibration, piezo pickup, and the recommended clap sensors). I am not looking for pinpoint accuracy, as i do not expect this with diy setup. It just seems like a fun project, and if successful would be a fun and cheap way to practice.

I am not set on using a cloth projector screen, sounds like a wood sheet may be a better option. I will post the results of my testing for anyone that is interested.

MrMark thank you for your experiment data, that is encouraging.

bbrown, I would think dealing with amplitude of the noise level may be simpler.
the ratio of the four different amplitudes should give an idea of it's position.

Make a touch screen.

The earliest touchscreen devices used arrays of LEDs pointing to detectors in an X-Y arrangement. When your finger touched the screen it breaks some of the X and Y sensors and the computer could calculate where on the screen you touched. So build a target of a flexible membrane that the projectile would flex enough to break the LED beams. Like a finger on the ancient touch screen.

This topic was automatically closed 120 days after the last reply. New replies are no longer allowed.