Sensor(s) to determine if plastic pickleball has hit target

GrampsFish

Joined May 22, 2020
29
I made a 4' wide x 3' tall plywood practice wall, and drew an 18" wide x 10" target on the plywood.

I would like to make some sort of sensor array that would be positioned about 1.5" off the plywood playing surface and sound a chime each time the 2.75" diameter plastic pickle ball tripped one or more of the vertical sensors at the same time it tripped one or more of the horizontal sensors.

Depending on cost, I'd have the sensor array cover either almost the entire 4' x 3' playing surface, or just the 18" x 10" wide target area.

From the get-go, I'd like the ability to individual disable parts of the sensor so I can change the size of the target area.

My guess is something like this has already been discussed here, but I haven't been able to find the correct combination of search terms to narrow my reading to less than 155,000 articles.

I use the practice board both in the garage (where sunlight isn't much of a factor) and on the driveway (where sunlight might be a factor).

I'm most interested in suggestions for the most effective type of sensor system to use, and would appreciate being pointed towards resources for determining what a bill of materials for the system would look like. I'd especially like to hear suggestions on the types of pitfalls I should try and avoid.

Just FYI, my woodworking skills would allow me to create 4' long recessed holders for sending units that align with matching holders for receiving units, and, while I conduct Basic Hydraulic training that includes how to read hydraulic schematics, my expertise and experience with electrical systems is limited to basic household wiring.

Bottom line is that, while I'm sure I'm quite capable of making this sensing system, it is going to be a pretty steep learning curve for me. Anything you can post or link to that would shorten that curve would be much appreciated.

Attachments

• 1.1 MB Views: 27

OBW0549

Joined Mar 2, 2015
3,566
Just a random thought, related to the pictorial in your attachment. Know what's dirt-cheap right now, and would be ideal as the sending units for your light-beam curtain? Laser diode modules. Amazon.com has scads of them, such as these going for about $0.63 apiece in lots of 10. I bought a pack of 10 to play with and they work really well. They throw a nice, tight beam that would work well with a phototransistor as the receiver- maybe one of these plus a suitable pullup resistor, 10 kΩ or so. That would give you a nice pulse whenever a ball interrupted the beam. Because the laser module is putting out a very narrow beam that can be precisely aimed at its corresponding sensor, the sensor could be shrouded with a black tube that would keep out sunlight or any other ambient light so as to make it responsive only to the laser beam. As to the rest of it, you'd need electronics to process the signals from the phototransistors and ring the chime, but that should be pretty straightforward. jpanhalt Joined Jan 18, 2008 11,088 I would consider using IR sensors like this: http://www.vishay.com/docs/82463/tsop62.pdf The source can be several emitters at the appropriate,synchronized modulation frequency. The sensors in tic-tac-toe pattern. Then use a microcontroller (MCU) to handle the logic as the pickle breaks the beams. Electronically simple. You will need basic programming. I use Assembly; most people today use C. Almost any MCU will work. The modulation (whether 38kHz or other) is unimportant. That distance is probably possible but may be stretching it. Reloadron Joined Jan 15, 2015 7,062 I haven't a clue as to the ball diameter but one option to consider, keeping inexpensive, is create a matrix of LASER Emitters and LASER Sensors. I see as I sat here there are already replies. Anyway I would set the receivers (sensors) up as a matrix and scan the matrix possibly using a multiplex circuit. From there your choice as to using a uC (micro-controller) or using discrete components. Ron KeepItSimpleStupid Joined Mar 4, 2014 5,088 The key is the IR has to be modulated (carrier on/carrier off) . Then your basically looking for a missing pulse detector. low duty cycle lowers power consumption. jpanhalt Joined Jan 18, 2008 11,088 I agree that for a "beam break" device, low power laser would be most appropriate for the emitter. Thread Starter GrampsFish Joined May 22, 2020 29 Just a random thought, related to the pictorial in your attachment. Know what's dirt-cheap right now, and would be ideal as the sending units for your light-beam curtain? Laser diode modules. Amazon.com has scads of them, such as these going for about$0.63 apiece in lots of 10.

I bought a pack of 10 to play with and they work really well. They throw a nice, tight beam that would work well with a phototransistor as the receiver- maybe one of these plus a suitable pullup resistor, 10 kΩ or so. That would give you a nice pulse whenever a ball interrupted the beam. Because the laser module is putting out a very narrow beam that can be precisely aimed at its corresponding sensor, the sensor could be shrouded with a black tube that would keep out sunlight or any other ambient light so as to make it responsive only to the laser beam.

As to the rest of it, you'd need electronics to process the signals from the phototransistors and ring the chime, but that should be pretty straightforward.
Thanks!

Bernard

Joined Aug 7, 2008
5,784
Just for kicks, might look at pressure detectors. In Search Forums put in Velostat. Cricket ball
detector- Sept. 16, 2019.

GrampsFish

Joined May 22, 2020
29
I would consider using IR sensors like this:
http://www.vishay.com/docs/82463/tsop62.pdf

The source can be several emitters at the appropriate,synchronized modulation frequency. The sensors in tic-tac-toe pattern. Then use a microcontroller (MCU) to handle the logic as the pickle breaks the beams.

Electronically simple. You will need basic programming. I use Assembly; most people today use C. Almost any MCU will work. The modulation (whether 38kHz or other) is unimportant. That distance is probably possible but may be stretching it.

If the sensors can be configured in series, would an easier/less expensive/more simple alternative to a MCU be to instead use some sort of relays with one relay tripped whenever any one of the horizontal sensor beams was broken, and a second relay tripped whenever one of the vertical sensor beams was broken... and then having the chime sound any time both relays were tripped?

Joined Jan 15, 2015
7,062
Whatever you choose for sensors I would be thinking a matrix array. Just for example a 4X4 matrix would use 16 sensors but only require monitoring 8 points. Reducing the number of monitoring channels is why a matrix design is used. A 10 X 10 Matrix covers 100 points and requires only 20 channels. Matters not what the sensors are be they LASER, pressure or other optical types.

Switches in series all need to be closed before anything happens, switches in parallel only need a single switch closure but how do we know which switch closed? See what is happening here?

Ron

jpanhalt

Joined Jan 18, 2008
11,088

If the sensors can be configured in series, would an easier/less expensive/more simple alternative to a MCU be to instead use some sort of relays with one relay tripped whenever any one of the horizontal sensor beams was broken, and a second relay tripped whenever one of the vertical sensor beams was broken... and then having the chime sound any time both relays were tripped?
Relay costs v. MCU? I think MCU's win. However, it can be done with discrete logic, which is sort of like relays.

1) You need to create the correct signal. Two 555's or a 556 can do that. It can also be done with Schmitt inverter (555's in that configuration are actually actin as Schmitt inverters). A simple, single mosfet amplifier could drive all IR sources in parallel.
2) You will need "one-shots." A 555 and several other chips do that. That will open a window so the x and y sensors do not need to be simultaneous. I understood your original post to want that. One could get around it with an RC coupling to give a small window. I suspect that would work.
3) You will need some descrete logic. They come with up to 6 or 8 gates per IC. Probably "AND" or "NAND" type (i.e., bot x and y need to be detected at about the same time) .
4) Finally the detectors and, presumably,weak laser emitters; although, if you use light guides on the detectors (piece of small aluminum tubing) you might get by with just one or two ordinary IR LED sources. It all depends on how narrow you want your beams/areas of detection to be.

Bernard

Joined Aug 7, 2008
5,784
If a hit anywhere on target is OK then if Rx output is low with light then all outputs could be ORd.
Also only one set of sensors ( 4 Rx on 18" ) would be needed.
Modulated Tx should be narrow beam so that only one Rx would be activated by one Tx.
One possibility is LTE-2871, 940 nm, 16 deg. with a short focus lens like Anchor Optics PN: 27435
9.5mm X 14.5 mm, 34mm focus. With this combination ( not modulated ) @ 18" detection was about +- 1". Could use Photo Logic receiver OPL561 OC with a modulation frequency of your choice, output lo with light, & open collector.

Bernard

Joined Aug 7, 2008
5,784
The Velostat pressure sensor seems quite simple. a sandwich of Al foil 10" X 18", Velostat & Al foil
12" X 20" with maybe a sheet of plastic film for Al protection. Just finished a 10" X 12" example.
Assembly pressure gives about 5k, 2.5" rubber ball strike gives about 500 ohms- easy trigger for a 555 chime driver.

Bernard

Joined Aug 7, 2008
5,784
11" X 11" available from Digi-Key, PN: 1528-2211-ND, US $4.95 or elsewhere, ( ETSY ) 36" X 12"$9.99 / ft.

GrampsFish

Joined May 22, 2020
29
Whatever you choose for sensors I would be thinking a matrix array. Just for example a 4X4 matrix would use 16 sensors but only require monitoring 8 points. Reducing the number of monitoring channels is why a matrix design is used. A 10 X 10 Matrix covers 100 points and requires only 20 channels. Matters not what the sensors are be they LASER, pressure or other optical types.

Switches in series all need to be closed before anything happens, switches in parallel only need a single switch closure but how do we know which switch closed? See what is happening here?

Ron

It seems to me that if I put the horizontal sensors in a series circuit and connect them to one relay, put the vertical sensors in another series circuit and connect them to a second relay, the logic becomes pretty simple - the chime sounds any time both relays are tripped at the same time.

The pickleball is about 2.75" in diameter, so I expect I'll put the sensors about 2.5" apart, which means the ball will actually break two horizontal beams and two vertical beams any time the ball strikes the target area, but I don't think that affect the logic at all.

Attachments

• 111.9 KB Views: 3

GrampsFish

Joined May 22, 2020
29
11" X 11" available from Digi-Key, PN: 1528-2211-ND, US $4.95 or elsewhere, ( ETSY ) 36" X 12"$9.99 / ft.
I certainly sounds like an easy way to go. I wonder:

- if the ball will bounce differently off the material than it bounces off plywood. The pickleball is much lighter than a rubber ball, and since my goal is to make the chime ring 10 to 20 times in a row, if the material absorbs much of the force of the ball, it will defeat the whole purpose;

- when you say the 2.5" rubber ball strike gives about 500 ohms and it is an easy trigger for a 555 chime driver, does that mean "all" I would need to do is connect a power source to the material and run a wire to a chime driver?

Bernard

Joined Aug 7, 2008
5,784
You would run 2 wires to the sensor, one for ground , one to pin 2 of 555.
Maybe if the whole board were covered with cloth, covering sensor, the ball would react the same no matter where it hit ?

Deleted member 115935

Joined Dec 31, 1969
0
A thought, its easier to make a receiver narrow beam than a transmitter.
so a light along two sides, cna be detected by detectors on the other side.
just pit a "tube" on the receiver. Does not evne have to be a tube, just two holes separated by a few mm works.

Its much easier to detect AC signals than DC ones, so amplitude modulate the emitters, means you can use an AC coupled filter as the receiver, with filtering if needed . If you MCU is fairly fast, such as a tensey Arduino, then you can apply filtering of the signal sin the MCU, to avoid false detection,

Also using the receivers like this, allows you to easily have a none linear array, you just have to move the holes / receivers.

Bernard

Joined Aug 7, 2008
5,784
I dropped a ball from 5' on to hard floor & sensor on hard floor, rebound was at same height
to both.

GrampsFish

Joined May 22, 2020
29
I dropped a ball from 5' on to hard floor & sensor on hard floor, rebound was at same height
to both.
Thanks again!

If I can impose on you again, would you take a look at the following and let me know if I've missed anything?

I'm going to find a local electronics supply store that is kind of like what Radio Shack was like in the 1970's, hand the nice guy behind the counter the Mores-code practice oscillator circuit shown in Figure 5 at https://www.nutsvolts.com/magazine/article/using-the-555-timer-ic-in-special-or-unusual-circuits , and ask him to sell me all the parts I need to make the circuit (except I won't be using a Morse key and I'll use a small speaker instead of the headphones). (I picked that circuit because both the pitch and volume of the tone are adjustable)

After I get the circuit to work, I'll buy a 19" wide x 11" high sheet of Velostat.

And then I'll make a "Velostat sandwich," with:
- a 20" x 12" sheet of double sided tape attached to my 4' x 3' plywood playing surface;
- an 18" x 10" aluminum foil target;
- the 19" x 11" sheet of Velostat, which will completely cover the aluminum foil target and be held in place by the 20" x 12" double-sided tape;
- a 19" x 11" sheet of aluminum foil which will just sit on top of the Velostat;
- a 20" x 12" piece of plastic sheeting (I make customized magnetic-backed dry erase calendars, and have a self-adhesive dry-erase plastic sheeting that I use when turning a customer's photo into a dry-erase fridge calendar)

As I make the "Velostat sandwich," I'll attach a wire or piece of conductive foil tape to each of the pieces of aluminum foil, being sure to isolate them from each other, and connect those wires to the electrical circuit in place of the Morse key.

I'm guessing I should use a heavy-duty aluminum foil to get the longest possible life from the target.

Does that sound like it will turn your idea into a functional target?

And, if you don't mind, two other questions:

When I was looking at the code-practice oscillator circuit, I guessed that the "R" stood for "resistor," but had no clue what the "C" stood for (the Internet quickly told me that, in this case, "C" is for "capacitor" and NOT for "cookie").

What the Internet hasn't suggested is what kind of power source to use for the circuit. We use Eneloop AA batteries for most thing around here, and I'm wondering if I'd be better off using somewhere between 4 and 8 Eneloop AA batteries, or a single 9v battery like we use in our smoke detectors.

I try and have three or four 20-minute sessions smacking the ball against the board each week. My guess is that I'd be hitting the target and sounding the chime as much as 400 times each session, which might add up to 1,600 times a week.

So, the first of the questions is, do you have any suggestions for the power source?

And the second question is, could the code-practice oscillator circuit be easily adapted to include an L.E.D. counter and an adjustable timer? Right now, I'm not looking for any guidance on how to add a timer and a counter to the circuit - just interested in knowing if it would be very difficult or expensive to do.

Thanks again for all your help!

Attachments

• 418.4 KB Views: 10