I'm planning to build a roll a ball horse derby game for a local carnival we are establishing in our community to get people together for some fun. The horse derby game struck me as a nice addition to the carnival allowing the participants to race each other instead of compeeting alone (against the carnival booth).
I'm a software developer by trade, but I have never touched an Arduino and have no experience with this cool technology. I've been studying videos etc. about the subject, but overall there are two questions, detecting the ball going through the hole and the track system, which I think some of you may well have some good advice about. Please see the attached picture about my initial thoughts.
Question 1:
The game will be a 4 lane setup with holes in 3 different colors enabling the horse to move 1, 2 or 3 steps corresponding to the color of the hole the ball went through. Each lane would effectively transmit a stream of data like [1,1,1,2,1,1,3,3,1,1,2...].
How would you detect which hole the ball went through and send the value to the Arduino board?
Question 2:
Each horse will travel a distance corresponding to the value recieved from its matching lane. I picture a micro switch and the end of the lanes to detect the winner and one micro switch at the start to detect a full reset of the horses once the game has ended. Traveling along the track could be done with timing belts, but it seems as if this solution would be both costly and a bit over the target in terms of the quality required.
How would you built a track for a horse to travel towards the finish line?
I hope this spawned some ideas in you which could get this project started.
Thanks for reading my post.
The image is now available.
The stream of data would be caused by the ball going through the holes several times during a game. On the first throw the ball might drop through a yellow hole causing a "1" being sent from the lane to the Arduino board. On consecutive throws, each time the ball drops through a hole a "1", "2" or "3" would be sent to indicate the color of the hole the ball went through and how much the horse should move forward.
I been looking at the optical sensors (hole detection) as indicated on the image. Is is possible to have one sensor detect multiple holes?
Optical sensors as the start and finish detector are a really good advice.
I remember those from a seaside town called Rhyl in North Wales, from when I was a kid. I have often thought it would make a good project to do with an Arduino.
I would use a micro switch to detect the ball dropping through the hole.
For the horses I would use a toothed belt like you get on the Reprap Prusa 3D printer, attached to a stepping motor.
Each lane would effectively transmit a stream of data like [1,1,1,2,1,1,3,3,1,1,2...].
No you are not thinking embedded controller here. When each microswitch detected a ball it would send the stepping motor the appropriate number of pulses to advance the horses.
Regarding question 1:
So the optical detetectors should be able to "cover" a row of holes. And by connecting the two yellow sensors and the two blue sensors I will get three signals coming from each lane.
Regarding question 2:
Each track is planned to be 2,5 meters long, making it somewhat costly if the quality is high. The timing belt i fairly cheap, but I think the gliding system shown on the mentioned 3D printer exceeds the requirements in terms of freeplay and accuracy. This seem to be a costly solution. I would prefer a cheaper and more guerilla solution.
I was thinking more about channeling the return balls on a track and detecting the balls as they roll past.
not sure what you are thinking of but it is not what I was. A stepping motor attached to a toothed belt, with a roller baring at the far end. If this is too long and produces unacceptable sag, then a few roller bearings along the way to keep it up. I don't see this as being expensive at all.
I'm not sure what you have in mind regarding the rail system. I'm following the use of the timing belt, but cant picture your vision of the rail and glider.
The timing belt runs the full length of the track. The horse is attached to the belt. You only need a switch at one end to position the system in power up and an optical sensor would do. Either reflective from a white spot on the back of the cut out horse. Or a beam it breaks. There is no rail or guide but you can use simple bearings part of the way across if the belt sags.
Or you could turn the belt through 90 degrees and have a plank support it, or even have the horse on wheels if you want.
Hi,
How did you get on with trying to copy my product?
It's interesting to see other peoples ideas on how the game can be made to work. After working on the same games for 25+ years it is easy to get too reliant on the same old methods, but what you guys are suggesting here isn't too far off the mark
I've never used Arduino personally, the games in your photos I designed around the Microchip PIC.
If you want to buy the game ready built, it's £18K
Carlsson:
I been looking at the optical sensors (hole detection) as indicated on the image. Is is possible to have one sensor detect multiple holes?
If they're all in line, yes. Restrict the view of the detector (5 cent phototransistor) and you have a 'beam' even with a wide angle source able to cover many detectors.
You could probably use fishing line to pull the horses between pulleys if the driving pulley has a pinch roller against it to ensure the monofilament turns with the pulley wheel.
-or- if you have a long wood dowel (or PVC tube, etc) and can glue a thick wire or cord spiral wound around it you would have a cheap lead screw. It could have a pitch (one complete turn) > 1" if you turn the rod slow enough (easy with a stepper motor). Just be really good about bonding the "thread" to the rod, fillet where it meets the rod. With PVC you might get away with cutting a groove.
I see you showing premade sensor modules... you could build those yourself really cheap out of led, detector, scrap and glue.
Grumpy_Mike:
Well, seeing as how I saw this game back in the 1950s it is hardly your original idea.
I see why you call yourself Grumpy Mike! :-p
Of course the original idea isn't my own, I never said it was. I can give you the full history of the game dating back to the '20s if you wish?
The 2 games pictured however, were both made by my own company to my design and that is the only reason why I signed up and wrote the post.
The original games and the larger ones that we produce, work on an electrified track system (think large scale Scalextric) with 50VAC bidirectional motors.
The smaller ones as in the photos use a German made pre-assembled toothed belt system with a standard DC motor drive. Limit sensors are reed switches.
Ball sensing is done via 3 basic pairs of photo transistors/IR LEDs
I'm also making a diy Roll a ball game and successfully made two 'playfields'. I used three v-shaped channels for the hole detection. At the end of each of the channels is a ir breakbeam-sensor from adafruit. Carlss, the professional roll-a-ball-builder, sells a ' score tray' on his website Score Tray (optical sensor V4) – Roll-a-Ball Ltd This should give you a good idea how to build the channels and where to put the sensors. I used aluminium profiles on a wooden base.
At first I used a scoreboard with a servo. It was like a clock and the arm was attached to the servo to indicate the score/progress, but this isn't as exciting as moving horses.
So now i'm in the progress of making a track with the use of nema 17-stepper motors and a timing belt.
I'm not experienced in writing code, so i'm struggling to get the new set up working. Will be testing in the upcoming weeks. How is your project doing?
Carlsson:
So the optical detetectors should be able to "cover" a row of holes. And by connecting the two yellow sensors and the two blue sensors I will get three signals coming from each lane.
Shown are two yellow sensors, two blue sensors, and one red sensor. That's five sensors (signals). If it gets critical you could wire-or the yellows and blues together to get three signals.
A bicycle chain might work for moving the horses. It's cheap, widely available, and easily strong enough. And, I think with some careful design, you could count links to know how far a horse has been moved. This would negate the need for a stepper motor.
You may be able to get away with only one motor if you could mount a shaft the width of all the lanes and attach a clutch for each lane.
