looking for the right IR emitter, receiver, and help with OR(ing)

Hi Cory,

If I understand what you want correctly, you need 40 light sensors with just one output - if any sensor loses light (interrupted beam) the output triggers.

If you could explain the geometry of how they are positioned hand how large is the interrupting object it might help us to optimize. If they are close together, you may get light from multiple sources hitting each detector. If this is outside, then sunlight may saturate all detectors -you might have to modulate the light to differentiate it from ambient light.

1) If you could have retro reflectors on the object and use light sources beside each detector - then the trigger would occur when a light was received instead if interrupted. Then you could simply use photo transistors or photo darlingtons (more sensitive) for detection and they can all be "ORed" simply by connecting all of their collectors to the same pull-up resistor. (parallel open collector configuration) If any photo darlington gets some reflected light the output will go low.

2) This could also be done with light interruption as well by connecting the output of each sensor to an open-collector transistor, but this requires wiring up an extra 40 transistors.

3) If the speed of the object is not too great, you could use cadmium sulfide detectors or cadmium selenium which is better for IR. Just connect them all in series (a single daisy chain = minimum pins and connections). Since they can be less than 1K ohm when lit up and 20K when dark, you just need a single pull-up resistor and a comparator to detect an "ORed" interruption trigger. You might even get away with using a transistor instead of a comparator - you could adjust the threshold using diodes or a zener.

Doug

The object is a ping pong ball. I am building a ping pong table that will be inside. Once I did all the math I realized I need 40 sensors on either side of the table and each would be about 1.5 inches apart so no mater where the ball lands on either side it will always trip a beam. I am looking for the ball to just be recognized that it hit the table on whichever side it does land on which is why i just need the l/0 output for either side.

The diode OR sounds promising.

It just so happens I just finished building a sensitive shock sensor. In addition to an output it illuminates an LED when it detects a vibration.

It can detect a pin drop from a couple of mm.

To provide some indication of whether this could work for your application, I jammed the sensor on the underside of my dinning room table (I don't have a ping pong table) and bounced a ping pong ball on the table.

In the video you can see the circuit board powered by a USB battery at the far side of the table where the sensor is. And you can see the LED flash when the  ball hits the table. The sensor could be made more sensitive since there is an amplifier in my circuit - I just don't need more sensitivity for my application. Attachment to the table could also be improved, I just stuck it in a crack.

Note that this table is split in the center to allow extra leaves to be inserted. When I bounce the ball on the other side of the split the impact is not detected. This would be ideal for ping pong tables that have two separate halves.

Doug

Edit media

Dimensions x Small Medium Large Custom

Subject (required) Brief Description Tags (separated by comma) Video visibility in search results Visible Hidden

Parent content

RE: looking for the right IR emitter, receiver, and help with OR(ing)

Poster

Upload Preview

Ok

Cancel

Ping pong sensor

not sure but I would think there is a need for separation of sensors across the net, if the idea is to detect if a player indeed cleared the net and hit the table, bouncing on your own side may not count after it has just arrived ?

just a thought, so is it possible to have each side separated without a gap in the table ?

LOL, sorry Doug, for some reason I did not see the last line of your post, ignore me lol

Well this gets into the system design choices that depend on accuracy requirements, cost, complexity and some unknowns that may need to be tested. You could probably use either amplitude or propagation speed to determine which court was impacted.

With a single sensor in each court, you could determine which signal was larger to determine which court was impacted. There might be ambiguity near the net, so. this may need careful calibration for impacts near the net.

If propagation delay is used, the first sensor to trigger would "win" and indicate which court was impacted. Basically the first trigger would "lockout" the other one for some period.

yes i was planning on having a separation as my goal is to make this look as much like a ping pong table as possible. What is the schematic for the circuit you are using or where did you buy it? I will give that a shot as I said I have successfully managed to hook up 3 IR sensors in series to one pin using a diode for each sensor and so far it is working perfectly, but your idea of the vibration circuit is very interesting I could theoretically stick one to the bottom of each side, and that would be much cheaper than buying and wiring 40 IR emitters and 40 IR receivers since i only have to buy a few parts to make the boards.Here I show the 3 sensor test that I have set up now. It's not the cleanest I just wanted to see if it would work first.

Edit media

Dimensions x Small Medium Large Custom

Subject (required) Brief Description Tags (separated by comma) Video visibility in search results Visible Hidden

Parent content

RE: looking for the right IR emitter, receiver, and help with OR(ing)

Poster

Upload Preview

Ok

Cancel

As for the rules, they would be standard so the system on which I detect the ball wouldn't matter. If the ball hits one side more than once than it is a point to the other person. if the ball hits the net and goes over then that would count. This is where the one of the only faults i see with the shock detector method. If the ball gently lands on the side( for instance it hits the net but gently falls over so it lands bounces maybe once then roles) then that may be a problem. I have thought about that and was going to add to the code to prevent a problem like that with the IR method anyways so it's not a big deal because that is still easier and cheaper.

I'm pretty sure a sensor could be set up to detect a ball falling just a couple of inches.

It is definitely an inexpensive sensor.

I would prefer to wait on publishing the circuit until the person I designed it for has had a chance to publish her research paper. Not a question of rights and it is a completely different application, its just a courtesy. I will discuss it with her.

Yes, I was thinking it could be detected, if I could replace 80 sensors with 2 chips that would save me a lot of money. So, since it would be less expensive I could at least build the circuit and test it to see if it works. If it doesn't then I wont lose much money , and I could probably find a way to use it later with another project. I don't mean to sound pushy if I do I'm sorry, but do you happen to know when she has to turn the paper in? I do understand and it is fine for me to wait on her to turn it in though. This table is actually tuned down because my original design was to make a grid of IR sensors for each side to make a grid so the machine knew exactly where the ball was then where it landed make an animation at that location, but I couldn't afford it so I settled with just being able to sense the ball altogether. That is why I started out thinking the IR sensors were best. I will continue to work on that IR method while I wait because in the future when I do have the money( I'm talking 10-15 years) to build the table the original way I will be able to just build it and not worry if it will work or not. Thank you for telling me about your circuit though it will be nice to see if it will be able to see if it will work which I'm pretty sure it will.

If you had one of these shock sensors under each corner of the table, you could use time-of-arrival of the shock wave to determine X-Y ratios and pinpoint where the ball hit.

You might be able to do everything with 4 microphones instead. They could detect both the ball hitting the bat and the table, giving you position of when and where the bat hit the ball and when and where it hit the table. This also gives you velocity and direction of the ball. If you want a better idea of what height the ball hit the bat you would need at least one more microphone. Of course, you would also need to  figure out how to prevent or ignore other sounds.

That is  a good idea. Thank you. Not only would that save me time but money as well. I think my best bet would would be the sock sensor. I don't think I would be able to do the microphone idea because it can get pretty loud when playing unless I can find the frequency range that comes when a ball hits the table. I will be waiting to see about that shock sensor though so because as of right now that is the easiest and cheapest way I can think of. Thank you again for the help.

The shock sensor solution gives you some information about how fast the ball is travelling - just by the time between ball strikes on the table. It wouldn't be super accurate, because you don't know exactly where the bat is, but still pretty useful for simulating the action. And if you know where the ball hits on both sides and the rough speed, you actually know roughly where the ball gets hit from.

The shock sensor solution gives you some information about how fast the ball is travelling - just by the time between ball strikes on the table. It wouldn't be super accurate, because you don't know exactly where the bat is, but still pretty useful for simulating the action. And if you know where the ball hits on both sides and the rough speed, you actually know roughly where the ball gets hit from.

That would be for later on down the road when I can afford the parts for that advanced project. The version I am making just needs to know if the ball hits the table and which side it hit. so I could just have a shock sensor on each side and that could detect if the ball hits that side or not. As for the advanced project, my goal was to have big tvs act as a monitor for each side so 2 tvs total and I could have them show the animations to where the ball would hit. That is why I can not afford it right now, and so I am just settling with making this version.