Picking things back up

[webby2]

Getting a little confused.

I decided to add a pulley on the go-between part that rotates at a 1:1 with the input and output.

I measure that to the output and I get close to a 1:1 value, that is 1000g of pull on the prime mover needs about 900 to 950g on the resistance side, either direction of rotation.

I measure that to the input side and I get a much larger difference, that is close to 1200g for the prime mover needs about 900 to 950g on the resistance side, either direction of rotation.

I was getting a difference one way when measuring between the input and output, then I advanced the output and locked the output in place and now it is close to a 1:1 either direction.

[webby2]

This is odd.

I can move the system with the same force on each side, if I start at the right non sticky spot that is.
I am not taking any notes on this but one way I can keep adding in more prime mover force while maintaining the same resistance force.  This is not the same for all ways I can measure I think but the simple fact that I can do it at all is very odd.

Maybe it is just tweaking the frames and connection belt??? not sure.

[webby2]

This is one of those things that has bothered me for as long as I can remember.

I come up with some mechanical idea, do a rough sketch and a very crude approximation of forces and distances and all that, build and test.

Now in the process of all this I get these force amplification relationships. In a lot of systems what I see, and can calculate, is a massive force being created but since it is against a non-changing part there is no work being performed and the very small reflected force that can move happens to match the input to the output for a 1:1 work value.

The part that bothers me is that I can build something to where that non-changing force goes to infinite, I can build this in the real world but of course things don't hold up to an infinite force but I can see these ridiculous force values trying to manifest.  The bothersome part is HOW does the system only reflect the exact amount for a 1:1 work transfer? what is 0.1 percent of infinite?

I can't help but think that there is another level of communication that is happening between the input and output that I am missing, maybe we all are missing it.

[webby2]

Maybe the simple thought that if I can run this setup with a loss of work performed I should be able to "reverse" that loss into a gain,, is to much.

I have been trying different things and maybe got a little oversized even, but right now one way is so close to a 1:1 that I am happy to call it that, after all that is what it should be, but when I run it the other way I am now loosing almost 1\2 of the work I am putting in.

What is frustrating about that is that I can clearly see where and why that loss is occurring but have not been able to figure out how to use that to my advantage.

[webby2]

picked the test-bed back up,, I need to remove some parts that are not going to be used and while I had it apart I thought to "test" the output section with only part of the system and it still does its thing, it seems to be more touchy about RPM and mass and stuff but it still shows the same feel.
Simplified like this I think I was still overthinking what is going on, and in this setup it would be very easy to think that the work out is from the mass falling under the influence of gravity but I think that is wrong.