OOAK Table Tennis Forum

I'll be very interested to see how this turns out.
Hope it'll work for you!

Okay here is the video of it in action, I must warn you it is not yet complete, this is still testing. The code will still need some tweaks and the board is not mounted to the rig yet cause I couldn't find screws there were long enough to hold the motors, but you can see the idea behind everything.

http://s223.photobucket.com/albums/dd17 ... uff036.flv

ps. if you were wondering why it is test 2, its cause test 1 failed horribly.

_________________
-Galaxy 896 cpen with 2.2mm Cream MRS FH and 2.0mm FXEL Super Soft BH

I think that looks very good actually... getting the balls to come out continuously without getting stuck, is definitely one of the big challanges! Were you getting much spin on the ball?

_________________
OOAK Table Tennis Shop | Re-Impact Blades | Butterfly Table Tennis bats
Setup1: Re-Impact Smart, Viper OX, Victas VS 401 Setup2: Re-Impact Barath, Dtecs OX, TSP Triple Spin Chop 1.0mm Setup3: Re-Impact Dark Knight, Hellfire OX, 999 Turbo
Recent Articles: Butterfly Tenergy Alternatives | Tenergy Rubbers Compared | Re-Impact User Guide

getting the balls to come out one by one was the biggest issue, test 1 of the machine, balls were getting stuck in the mechanism so it failed. So a little redesign of the hopper mechanism solved the issues.

right now it only has only 2 degrees of spin on the ball, side spin left or side spin right. To get the rig to give me a underspin or topspin ball I'll need to add 2 more motors to the mix, that will be at a later date.

Materials:
Arduino microcontroller $30
Motor controller $9
2x motors $19
Servo $15
Acrylic glue $3
Acrylic boards $6

$82 Total, I think that beats paying $300 for a low end table tennis robot. Besides this thing has a 300 ball bucket capacity, bigger than any high end robot you can buy. No one said anything about picking all those balls up though.

_________________
-Galaxy 896 cpen with 2.2mm Cream MRS FH and 2.0mm FXEL Super Soft BH

Hi, how is it going with the project. Have you made any further progress with it.

It would be interesting to see how far you can take this robot-project.

That looks so impressive, Ray!
May I ask some questions please! Sorry if they are stupid ones;)
What force makes balls to fly off the tray? Are they pushed off with 2 rotating wheels in opposite direction that have rubber tyres on them? If yes, then how do you set the right gap between the tires- I guess it will be very sensitive to this gap and the pressure on the tires?
Are the motors constantly rotating? I noticed you set state variable once and it is not set back to 0. Why the control sequence of the motors is so complex?
Side spin: is it just setting slightly different speed of rotation of the motors?
How does the device serves the balls one by one? I guess there is a photo sensor at the bottom is that right?
How accurate is the system? Does it land the balls on the table in the same spot?
Or am I wrong about how it all works? Would be nice to see the construction scheme if it is not a secret;)
Cheers!

_________________
729 FX 1.5 | glue | 729 Black Whirlwind | glue | 729 FX Lighting 2.0

I have the parts together to begin construction and have a donor Newgy recycling net, have you continued to work on yours. I am weak on the computer end of things so seeing how you have developed yours would certainly be helpful.
Thanks,
Thomas

Its been a while, really busy with work but I have partially finished the machine. Couple of flaws still exist though. I have the 2 motors running side by side to launch the ball, the problem here is that it does launch approx a foot, but thats not enough. To launch further I will need more powerful motors as well as concave wheels that would give the wheels the most surface area to grab the ball. I have used the robot a couple times before. There is nothing wrong with the program i wrote for the arduino, this was more of a hardware problem.

To power more powerful motors I will need a dedicated power adapter with the proper amps along with more torque/rpm speed. Don't forget that the wheels will need to have a concave shape along the treads so there will be a lot less torque loss. Torque loss coming off a standard wheel I didn't take account for when I first built it (I'm a computer engineer, its why i'm not a mechanical engineer ). For me to use my current motors, the equations say I will need wheels approx 4 inches in diameter for each motor. That would eliminate the portability of the robot so its not a viable method of approach.

If I had to redo the robot again I would suggest going the spring compression method, like a plunger on a pinball machine but electronically. Rubber bands might work but they snap easily. Maybe one day I will come back to this.

_________________
-Galaxy 896 cpen with 2.2mm Cream MRS FH and 2.0mm FXEL Super Soft BH

I used hard rubber wheels that are more like two inches, but the ones on the butterfly are closer to four and are very lightweight foam, they seem to wear concave with time. I have had a variety of robots, over the years with no wheels and both one wheel and two. I have one motor to rotate the head and one for each wheel.

My power supply is an old 220/110 60/50 herz one from a defunct laptop. Plenty of zing for the three motors.

As to your portability issue, I consider the Newgy to be portable as opposed to the butterfly 3000. I guess anything can be portable if you are big enough, I remember when a portable TV required muscle and some real determination to move it around:).

On a version years ago I tried to use air to propel the balls but that didn't work well other than making a pretty decent dead ball, so the wheel thing seems the way to go. I guess I am more mechanically inclined when it comes to the computer/mechanical battle.

What I need now is a way to control what I have with the computer, and then to write programs for drills, etc. I can reverse the wheels, slow them down or speed them up or reverse the direction independently but only on a case by case basis using switches and rheostats Movement of the head from side to side isn't a problem, but it would need to also be integrated into the computer program. Slowing the wheels and reversing them is not an instant process either, so ..on it goes.

I know that the new Newgy can interface with a comp[uter and even the older ones had an extra port (serial as I recall) on the power supply that never really came to any use.

If you have any more thoughts on the topic I would be happy to share what I have learned.

Thomas

we seem to have a few engineers on this site...


i was looking at a pretty funky ucontroller board a couple of days ago sporting an AD 7024 uprocessor. Tempted to get one, but couldn't figure out what I was going to do with it. I think they run the same bas%#rdised version of C++ that those ATMELs run

_________________
Chasse Patate

what motors are you using? RC Car? The new brushless motors have crazy tork & speed.
Electric 4wd that does 70+mph.

_________________
Gambler Black Diamond Light: *primary setup*
Black-Dawei 388C-1 Red-Quattro UL 2.2
YinHe M6: Outlaw & Six Shooter Pretuned
Hard Bat: LKT Instinct
Dr.Evil
Power is not revealed by striking hard or often, but by striking true.
-Honore de Balzac

Did you complete the build of TT robot ?
Any photos or videos ?
Any recommendation for someone new trying to make one ?