Building a self looping "SMOT"

[elecar]

For the purpose of this thread the term SMOT will be used. To date no one has looped a SMOT, largely because when a ball leaves the SMOT ramp it ends up at a height equal to or lower than the point at which it started. To over come that problem it is necessary to have the ball leave the SMOT ramp at a height that enables feeding the ball back to the start point.
Most previous designs utilize the ramp design in US4215330 A  which consists of an incline with two rows of magnets, one fixed either side of the incline with a track positioned in between the two. The ball is drawn up the ramp by the magnets and leaves the top of the ramp and under gravity falls off the end. Always ending up at a point lower than or equal to the start point which then leaves the problem of getting the ball back to the start point.

My design (Patent Pending) approaches the problem in a different and unique way, which I hope some of you replicate independently. 
I will start by saying I will not be showing my finished design for a while and if that is what you want to see please come back to this thread in the future when I will post a video of my unit.

What I will do however is tell you how to go about building a self looping "SMOT" of your own using the principles employed in my design. 
I will include rough drawings (sorry its not one of my strong points) I shall also include experiments for you try out which I used myself to create my own design. The experiments are worth doing, especially for the naysayers as they will show the working principle and give you chance to make adjustments before you go into a full fledged build.
Most of you who have played at trying to loop a SMOT and those who work with magnets will probably have all the parts to hand. For those who do not have the parts, there are very few and my prototype cost less than £15 (approx $23)
I will list the parts I used myself, but other parts may be used.

19 mm steel ball
40 x 1" x 1.5" x  .25" C8  magnets ( I have also used neo blocks but they were harder to set correctly)
3 meters of  2.5 mm  rod/wire for track, my prototype used aluminum but copper should be fine, you may even get away with the wire from a few wire coat hangers. Copper is probably best as you can easily solder the track, and it is reasonably flexible whilst being able to maintain its shape.


Next the working principle.

[elecar]

My design which is not called a "SMOT" but will be for the purpose of this thread works in the following manner.

A track is formed in a "teardrop shape" one long side of the "teardrop" is straight and at approximately 2/3rds of the way up is joined to the return via a junction in the track.
This junction is critical and must be designed correctly. The best way to describe the junction is that of a set of railway points where the ball can pass in one direction when travelling forward but exits in the other direction when travelling in reverse.

The straight side of the track is inclined and forms the ramp, magnets are placed on ONE side of the ramp in a manner which allows the ball to be drawn up the ramp but NOT to the end of the ramp.
The ramp does not work in the same manner as a normal "SMOT" instead the ramp pulls the ball up most of the ramp and as the ball loses its forward momentum is no longer pulled by the magnets, gravity takes over and the ball starts to roll in reverse back down the incline.
When the ball reaches the junction of the track it takes the alternate route (built into the design) and makes its way down the track and back to the entrance of the magnetic ramp.

Approximate track shape:

[elecar]

Experiment 1:  Getting the ramp to function correctly.

Your set up should be like that represented in the drawing below, raise one end of the track to form a ramp and then play with the position of the magnets until the ball is drawn up the ramp but loses momentum and starts to reverse. When you get the magnets in the correct position The ball will readily be drawn toward the top of the ramp but never quite makes the top, as it loses its forward momentum gravity will make it roll most of the way back down. the drawing shows ariel and side view



I will set up my experiments again and post videos in the next few days if anyone is having a problem getting the effect.

[elecar]

The following drawing shows an approximation of what you should aim for. The red dots represent the points on the track and their approximate set heights (you may be able to go higher or lower but those heights worked best for me with the materials I used)
The rails of the track are set at 12 mm outer edge to outer edge for the 19 mm ball.

[elecar]

And when completed should represent the schematic below.