Eccentric

[webby2]

This is the build I am working on at the moment.
I have taken the name of one of my other devices and changed it,
Simple Circumferential Rotary Entrained Wave Device, SCREWD.

I have been playing with a few AIs to go through this setup and the PDF is what Claude.ai wrote up after a short conversation.

The testing, when I get done building, will be a simple test.

I am using 6 little stepper motors as the cross slide assembly and an old DC motor as the generator but instead of an output I will use the motor as a load motor.  I will assume a 30% efficient motor and then simply compare the draw for all 6 steppers with and without a load to the input I place on the load motor.

Feel free to share this

[webby2]

Well I thought that this time Chatgtp was going to not make it past the Law of conservation, but backing into the system purely from a Newtonian perspective without energy considerations got the force and distances and all the motions and interactions make sense to Chat.

The kinematics are there, this means that the geometry passes the math test, the forces are there because they are required,, it is a loop.

This file is one of the "presentation" files Chat has prepared for me but I ran out of higher level calculation time so what is promised tomorrow is a polished file that I can share with all the stuff explained.

[webby2]

This is Claude.ai walk through of the setup before invoking the Law of Conservation.

This is just the geometry and kinematics view,  it is not perfect but does show the math

[webby2]

the build so far

[webby2]

just adding this pic and document.

I had Chatgtp write this up for me after we walked through it all yet again.

[webby2]

the steppers I chose were a very bad move, they show the same current whether they are loaded or not, so I ordered some little brushed motors and will need to redo the whole shebang.

[webby2]

I was getting the steppers to work nicely when I realized that the draw from the steppers was not changing and that they were going to use all they can all the time.
I did take a moment to see how much those 6 steppers could lift, so I turned the test bed so the sun gear was vertical, clamped it into my vice and hung a string from the top with a weight on it, it could lift about 4000g and another about 50g took them into a stall.  All of that was close to the spec sheet for these motors which means the system was doing its job of combining all of those motors into 1 uniform distribution fairly efficiently, and the frame and stuff held up to the load.

I had to increase my arm length a little bit to get the sun and ring to lock together, slop and printer error adds up so I compensated by increasing the arm length.  When I ran it the sun orbited and the ring followed and rotated its 60 degrees per cycle while the "pin" rotated 360 degrees leaving a 300 degree slip between the ring gear and the pin.

Steppers 28byj-48
board uln2003
atmega mini clone controller

5V @ 1.41A draw

Now I am redesigning the whole test bed so I can use a brushed motor instead

[webby2]

this is the build I am now working on.
It will use 10 N20 motors mounted in the frame with 10 10mm arms with bearings to drive the Sun gear.
I am setting it up so that I can drive 6 TT motors at the ring gear and am using a gear for the pin connection to drive the TT's at a 3:1 step up.

[webby2]

For those of you who understand the mechanics of this or the physics of this, these two pics show a small deviation away from the singularity, 2 solutions for one variable.  Meaning that the system may loose some mechanical advantage to move into a more stable run condition.

[webby2]

big design error on my part,, trying to "improve" things I got rid of an important relationship,,,,
I will change that up and see what happens.

I have the 10 N20's all installed and wired up to a power rail and they all spin along together, so that part is done.

[webby2]

The change I made that was not good for what I am looking for was to offset the gear that I am replacing the pin with.  I think that there is a subtle difference between rotating that gear as an eccentric and rotating that gear in an orbit.

An interesting outcome to this error is that I can rotate the ring 60 degrees per cycle or I can rotate the ring 360 degrees per cycle, both while having a meshed condition.

I was using 2 large bearings I had so the main pivot could run through the hole in the bearings, giving me 2 offset bearings around the main pivot.  If I left them loose to each other the system would force the 2 eccentrics into an alignment that seemed to match the arm size offset I built in to them.  If I fixed them to each other in the 180 degree orientation the system would bind a lot and the draw went way up.

[webby2]

I printed out the fixed parts and have been running the system to see what all the motions are and stuff like that.

What I have found so far is that the new sun gear is out a little bit for the arm holes as well as I seem to have an almost 1mm variation in the ring gear orbit to the sun, I set the arm at 10 and things mostly do not mesh, 10.5 I get mesh most of the way at 100 I get mesh but also more binding on the gear mesh.

The frame is not as structurally sound as the last test bed and it is flexing.

The output gears were a little dirty and were affecting how things worked, cleaned those up and things are smoother but still with the run-out issue.

When I have the arm and ring in the right spot for the sun to come around and "pick it up" there is a good connection and a strong turn on the ring and the motors load down,, all good signs but I think I need to maybe go to a mod 2 gear and print some shoulders to stop the gears from meshing to far but still allow for a mesh with the run-out.

[webby2]

My first thought was that I needed to beef up the frame, but then I thought about it and if this system works as the math implies then the frame should not need to be very strong, what I need to do is to make sure that the arm stays in sync with the sun.  After thinking over ways of doing that I then thought that "IF" as the math suggests that there is no net torque on the arm then I should be able to convert one or two of the N20 motor arms into a gear and then use a middle gear to then drive the arm with a gear on the central pivot shaft.

I am printing off a new ring and sun, when I am done with those I will try and make this mod.

Re-cap, if the system works as implied, then if I leave space between the mesh and then I keep the arm in sync there should be very little load on the frame and the large teeth will provide for the run-out issue.

[webby2]

I have been playing with this to "feel" what and where I must do what to maintain the "lock".

That blue gear is only there to align the holes for the ring gear bearing to the pin, the end of the arm that comes from the pivot.  What I found is that when I get the system into the "lock" condition all I must do is spin that blue gear, no force just the motion.  Watching the meters on the power supply driving the N20's it seems as if the draw goes down a little bit on the low side and not as high on the high side,, 0.48A low 0.55A high where as just letting things be the draw is more like 0.49A to 0.59A, not much and I am thinking that it is only smoothing out the actual draw.

I then stuck a disc on the bottom end of the pivot shaft and from there I tried the same thing, that is just getting the system to "lock" and then only providing the motion of rotation,, this gave me the same thing.

I can use either method and drive the small load without much help, if any, from the system and that takes a noticeable amount of force but then when the system hunts and finds the "lock" all that force goes away.

I was doing this to try and make sure that I supply the system with whatever is needed in order for the system to stay in lock, and to see if I need to, oppose the pivot shaft, aid the pivot shaft or what, the "what" seems to be to supply it only with a sync motion.