Another Milkovic pendulum pump replica from India:

Gravity Energy Machine (The Pendulum Pump)
Diploma in Mechanical Engineering Final year (2013) project made by me and my team Jatan Desai, Amey Patel and Vinay Sawant.
http://www.youtube.com/watch?v=g4gLRODJAIU

Pendulum Pump
Gravity Energy machine (The pendulum Pump)
Its our final year diploma project. We are Mech engg. students from Mumbai.
http://www.youtube.com/watch?v=hXLe6SCH7Qg

After building and experimenting with 2SOs, you will find the device is more efficient when either the "counter weight to main lever pivot point" is longer than the "pendulum pivot point to main lever pivot point", or the counter weight is approximately 2x or more heavier than the pendulum bob (assuming both distances mentioned above are equal).  The reason is you want to counter weight oscillation to occur quickly (that is the up and down movement).  A quicker movement means more force, and thus more work can be done in less time (more power).
The standard Milkovic 2SO is a break even design (assuming no loses due to friction or air resistance).  Has anyone given much thought as to why?  Build one and experiment with it.  The reason will become clear.  It's acceleration.  I probably shouldn't have said, but I'll let you determine where and how to eliminate it.  Until you do, you won't progress beyond the standard design and it's results.
Looking at different designs is great for inspiration, but eventually you are going to have to build one.  There is no substitute for the hands on knowledge you will gain.  Forget the simulations, they won't help you "feel" what you need to know to move on.

Quote from: zoelra on November 25, 2013, 02:03:22 PM

The standard Milkovic 2SO is a break even design (assuming no loses due to friction or air resistance).  Has anyone given much thought as to why?

I would predict any 2SO will be at best a break even device. The device stores energy the pendulum. Once that energy is used up it stops. Exactly the same amount as was put in will be available as output, minus any losses to friction or air resistance etc.

 I have done much experimenting with the TSMO, and one thing I think a lot of people overlook is the fact that you can change the weight, dimensions, pivot points, leverage, pendulum swing, pendulum length, etc., etc. You can build small models to experiment with, but a large model with a 50 to 200 lb.(OR more even) pendulum may really make you think twice about this device. There is an amazing the amount of work done with relative ease.
@ zoelera
Feltonburger,s tsmo method also seems to work well, despite not using the lever aspect. Or maybe it works better.  Not sure.
Greg

Quote from: gdez on November 25, 2013, 05:02:05 PM
... You can build small models to experiment with, but a large model with a 50 to 200 lb.(OR more even) pendulum may really make you think twice about this device. There is an amazing the amount of work done with relative ease.
Greg

Rather than make such statements, that offer no attempt at quantizing the results, actually measure the inputs and outputs with some accuracy and then comment.

If you do what you will find is that energy is conserved and there is no over-unity present.