Best table tennis robot - Amicus prime or Power pong 5000

A two throw wheel robot is easier to understand.
If the upper and lower throw wheel rotate at the same speed the ball will move at a speed proportional average of the top and bottom throw wheel. However, the spin will be proportional to the difference in speed.

My Newgy 2050 has a friction pad instead of a throw wheel so the speed is proportional to the average of the throw wheel and 0. The spin is proportional to the difference speed of the throw wheel and 0. The friction pad is like a bottom throw wheel that isn't spinning. To get more top spin than a Newgy the bottom throw wheel need to rotate backwards. The spin is proportional to the difference in rotational speed but because the bottom throw wheel spins backward it actually adds to the spin. When the bottom wheel is spinning backwards the speed of the ball is still the average of the top speed and the bottom speed but obviously the top wheel must be spinning much faster than the bottom wheel to make the ball shoot out. This permits the robot to shoot slow but very spinny top spin balls or slow but very spinny back spin balls by reversing the speed of the top and bottom threads.

Without this rotating backwards feature the robot cannot throw a ball that will up out low at you after landing on the table like a good loop will.

Originally Posted by latej

Let's say the angular speed is positive, if the wheel throws the ball out, and negative otherwise. Then the speed of the ball is proportional to the sum of the speeds, and the spin of the ball is proportional to the difference of the speeds, as you say.

It doesn't look like you believe me.
Think about it.

Head which points in any direction, and can rotate as clock-hand (in plane perpendicular to throw direction), and has 2 wheels, which can counter-rotate. Maybe kids of my kids will have that.

The only reason it doesn't exist is because people don't know what can be done. The counter rotate feature would require a bipolar power supply instead of a unipolar one. The robot must already have a bipolar power supply to rotate the head back and forth to shoot the ball at different places on the table.
Why not use the bipolar power supply to also counter rotate the throw wheels.
Two other things I would like to see is to move the whole head back and forth at the end of the table so balls can be shot from different places instead of always shooting from the center. Another thing that would be good is too have a vibration sensor that detects when the ball has hit the net to trigger throwing the next ball after a programmable delay. The current robots don't care when the ball lands in the net. They simply throw another ball after a fixed delay. This is not realistic.

Throwing a ball from different positions would much more expensive to implement but counter rotating the throw wheels to get more extreme spin and using a vibration sensor to trigger throwing the next ball would not add that mush extra cost.

Originally Posted by brokenball

It doesn't look like you believe me.

I actually do. I just tried to put it more simply. It appeared a bit confusing to me, to say that the speed is proportional to the average, because I think it is intuitive to see it as proportional to the sum. But strictly speaking it is the same.

Originally Posted by brokenball

Two other things I would like to see is to move the whole head back and forth at the end of the table so balls can be shot from different places instead of always shooting from the center. Another thing that would be good is too have a vibration sensor that detects when the ball has hit the net to trigger throwing the next ball after a programmable delay.

Moveable head would be great. Sensor on the net is good idea. Counter-rotation I think we will not see in the head with 3 wheels. And with 2 wheels you need also the clock-wise rotation - more expensive (or lose side-spin, I actually don't use side-spin much).

Enough of fantasizing...

I'm the founder of the pongfox table tennis robot . @brokenball In pongfox robot we have enabled the motor 2 (below) to turn inwards for a topspin loop. We get insane amount of loop Check out our comparison video. The main advantage of the pongfox robot in my opinion is the software, Our app has features like record while playing the drill from within the app so that you can view your strokes. Let me know if you have any questions

I have implemented the counter rotation in the pongfox robot, check my answer below in the thread :-)

@palguay, You have raised the high bar for what is the best robot. Now people know the difference.
That is very good but the top head also needs to counter rotate so it can serve ghost serves. If you can rotate the head to make the top throw wheel the bottom throw wheel that would be great.
I recommend you make another video that shows the top spin balls jumping out low after the bounce and make a big deal out of it because it is something your robot can do that others can't.

Originally Posted by palguay

I'm the founder of the pongfox table tennis robot . @brokenball In pongfox robot we have enabled the motor 2 (below) to turn inwards for a topspin loop. We get insane amount of loop Check out our comparison video. The main advantage of the pongfox robot in my opinion is the software, Our app has features like record while playing the drill from within the app so that you can view your strokes. Let me know if you have any questions

Are these available in the UK?

Originally Posted by brokenball

@palguay, You have raised the high bar for what is the best robot. Now people know the difference.
That is very good but the top head also needs to counter rotate so it can serve ghost serves. If you can rotate the head to make the top throw wheel the bottom throw wheel that would be great.
I recommend you make another video that shows the top spin balls jumping out low after the bounce and make a big deal out of it because it is something your robot can do that others can't.

Thanks for the suggestion, as far as making the other wheels go inside we can do that easily without rotating the head

Originally Posted by Basmundo

Are these available in the UK?

Yes, Someone from Edinburgh is interested and we are looking into the shipping details and should be able to do it within a couple of weeks

Hi Brokenball,

I have the 2055 robo pong, have you had any issues with the ‘L’ shaped spring breaking? It’s labelled as part 50 in the manual, part# 2050-159A BF spring, medium (L shaped)
2 of these springs have snapped at the attachment point.
wondering if this is a problem other robo pong owners have?

cheers.

Originally Posted by brokenball

Throwing a ball from different positions would much more expensive to implement but counter rotating the throw wheels to get more extreme spin and using a vibration sensor to trigger throwing the next ball would not add that mush extra cost.

Sorry but I didn't quite understand why they would have to counter rotating to get more extreme spin? I have a Butterfly Amicus Prime, and yes, it is not able to counter rotating, but when set to maximum spin (top or back) it is so extreme that it is practically impossible to return the ball to the opponent's side. Max spin is inhuman.

Watch the first 30 sec of the video where max topspin and backspin are shown.

Originally Posted by IB66

Hi Brokenball,

I have the 2055 robo pong, have you had any issues with the ‘L’ shaped spring breaking? It’s labelled as part 50 in the manual, part# 2050-159A BF spring, medium (L shaped)
2 of these springs have snapped at the attachment point.
wondering if this is a problem other robo pong owners have?

cheers.

l used to have a 2050 robot and all the springs right angle and long and short one broke .. l got spares but months later they broke again... the flaw is at the attachment point ... l made all the springs stronger by using a vice and two pliers stretch the springs to give you room to make a circle so the screw can fit thro the hoop and give you space to get the screw thro ,the most difficult bit ... ,they never broke again

Originally Posted by Basmundo

The robots that you are discussing are high end and too expensive for many, however good they are. Are there any cheaper robots available that can do more than simply oscillate backwards and forwards but can handle excercises like the falkenberg etc.?

Yes, Our robot can handle those as well in our opinion what makes our product better is that it is software driven, we have a feature where you can select a drill and start with a record option that records a video of you playing

Someone wrote that the Amicus robot can't counter-rotate, so I assumed that was the case, but now I went to check with my robot.
For topspin all three wheels rotate in the same direction (forward), but for backspin the upper two rotate in the same way as for topspin, but the lower one rotates counter.

No, it's not caused by a ball touching them. The wheels is rotating before the ball is thrown.

Originally Posted by brokenball

@ycjason, that doesn't answer the question about whether the Amicus can reverse the motors for extra top spin and back spin. To simulate slow but very spinny loops the bottom throw wheel must be able to spin backwards.

I guess I don't know the exact answer, the control used to manipulate spin is not based on motors. Basically you can control (each can be controlled independent from others)
* top/under spin
* left/right side spin
* speed of the ball
* projection of the ball

btw, speed + project and to some degree, the spin, controls the landing position of the ball

So you can definitely program slow and spinning ball (both underspin and topspin). How they actually achieved it, I don't really know. Hopefully this helps