I am pretty lazy  guy. This is a main reason why my drawing has pendulum length and lever ratio control. These controls give me opportunity to play with it and find better values for them if they exists.

Visualization of data
If I want to find out more about behavior of two stage mechanical oscillator (TSMO) I would measure as much as possible values of all parameters and put them in some kind of diagram capable of showing me all relations at once giving me a chance to see and understand. This way of diagramming exists a while and is called the parallel coordinates. Much about it you could find by Google. It is very powerful tool helping us to find out things which is not possible by other methods without more effort and imagination. To save some processing time some values are measured redundantly. We will have raw data to calculate it but this need time. If we need accurate measurement we have to do it in smallest possible time slice to catch as much events as possible. Time slice is a secret of parallel coordinates. All measurements MUST be accomplished in ONE time slice, then another time slice and so on till the end of experiment.

Measurements:
1. Angle betveen pendulum and lever
2. Angle of lever to horisontal plane
3. Force in pendulum rod
4. Force added to maintain pendulum swinging
5. Acceleration of pendulum pivot (redundant measuring)
6. Movement of pendulum pivot (redundant measuring)

Parameters adjusting:
1. Control of a force added to maintain pendulum swinging
2. Changing of lever ratio
3. Changing of pendulum rod length
4. Pendulum bob weight
5. Weight at opposite sight of lever

Gauges actuators and controllers:
1. Angle measurement (relative or absolute)
2. Tenzometric cells, force gauge
3. Step motors powerful enough to do the task
4. Measurement acquisition module with analog, digital and counter inputs as like digital outputs to control step motors (USB connectable with Android phone or tablet)
5. Control and visualization device (Android phone or tablet used in the same time for measuring movement and acceleration)

Position of Android is on top of pendulum pivot (well, the best place is under Android acceleration and movement sensor. Some adjusting needed )

As you see, I choose Android. Maybe because I got it for Christmas (TAC-70051) but mainly because it is small and powerful enough for the job.

I will have to say that your stuff is very tight and well thought out. I built my earlier models off this blue print and though I have different ideas now, I feel to do real measurements, your schematic is pretty damn good.

Nice of you to notice it

For me, a little to complex to build with all the arduino sensors, but not impossible.

Android! You have several good examples on Internet how to use internal movement and acceleration modules.

If you look at this video;
 
 www.youtube.com/watch?v=tb9dgb.bdBM
 
 ; you will see that the "traditional approach" to this device may not be the best. In this variation, it uses the milkovic principle but, much easier to build, as most of the weight is kept at a low center of gravity.

Your video link is not good anymore. It seems... Deleted

check out feltenberger stuff too.. he seem's to use the pivot point differently then most,  (it seems),with  good results.

This approach is good for high pressure pumps but have the same problem as all pendulum "machines". You have to put some amount of energy in it (to place pendulum in starting position) and to maintain swinging with further addition of energy. If you ask yourself: "What is real work of such machine?", after a while you will realize that useful work is equal extra added energy minus losses, because, with every swing you repeat the first step - put a pendulum in starting position. If working machine (pump) use more energy, the pendulum swing less and you have to add more energy to maintain movement. This  is reason why Feltenberger need so big bob (9 tons as  I could recall).
My proposal with rising level of outlet reservoir show this...
http://www.overunity.com/1763/12-times-more-output-than-input-dual-mechanical-oscillation-system/msg384597/#msg384597

As lazy as you say you are,
I'll give you credit for some nice work.
 Just don't let it happen again.(I mean the laziness)
 you are forgiven

Well, I will do all in my power to behave myself

Thank you for posting in this topic!

Paolo

Ciao Andrea

I already have replay to you by PM

Paolo

One of interesting feature of proposed construction of TSMO (see the jpg in one of previous reply) is capability to control the length of pendulum and to pump a swing.

Main idea is here http://www.besslerwheel.com/wwwboard/messages/435.html
Further mathematical explanation (mathematical model) and visualization could be found here http://www.maplesoft.com/applications/view.aspx?SID=4888&view=html
And Master's Degree Thesis https://www.bth.se/fou/cuppsats.nsf/all/6bd3163040fa186ac12576e90049ad18/$file/BTH_AMT_EX_2010D01_SE.pdf with something different approach but with similar conclusions.

By pumping a swing it is possible to maintain the movement of a pendulum and introduce extra force needed to overcome losses and work on other side of a lever.

Paolo

Paolo interessanti:
Nonlinear Oscillations, undamped and undriven pendulum
http://www.elmer.unibas.ch/pendulum/nonosc.htm
http://www.elmer.unibas.ch/pendulum/lab.htm

For collecting data during experiment we need sensors, transducers, actuators and some kind of data acquisition unit.

Here is proposal for data acquisition unit, transducers and actuators:

Data collecting http://www.mikroe.com/pic/clicker/

Force transducer (dual) http://www.mikroe.com/click/adc/
Angle collecting http://www.mikroe.com/click/expand/
Acceleration measurement http://www.mikroe.com/click/accel/

Relay board http://www.mikroe.com/click/relay/
Step motor adapter board http://www.mikroe.com/click/adapter/
Step motor board http://www.mikroe.com/add-on-boards/various/bipolar-stepper-motor-driver/

Specification of boards connection http://www.mikroe.com/downloads/get/1737/mikrobus_specification.pdf

Edit: I am not a member of MikroE staff or connected in any way with MikroE except as customer (satisfied one  ). 

Browsing MikroE boards leads us to knowledge that mikroBUS is not designed as stackable bus. Let us read http://www.mikroe.com/downloads/get/1737/mikrobus_specification.pdf with more attention. Pick up the attachment, it will be easier to follow me. Stacking boards mean that we want to put our mikroBUS boards in deck of a cards fashion meaning that all of them share the same lines (this is meaning of bus in electrical circuits).

mikroBUS has incorporated three serial communications standards for connection with parts on mikroBUS board:
Rx/Tx - RS232 or RS232C standard http://en.wikipedia.org/wiki/RS-232 (3-wire standard),
SPI - http://en.wikipedia.org/wiki/Serial_Peripheral_Interface_Bus and
I²C - http://en.wikipedia.org/wiki/I%C2%B2C

Using any of this standard mikroBUS board could receive or transmit data to master board (in our case PIC Clicker). We will not use Rx/Tx (for now...)
Other two serial standards has advantages and disadvantages but for us are important lines in black frame. SPI line CS - cheap select makes us a problem because it allow only one board to be selected. We have to isolate this pins and make separate connection for every mikroBUS board which use SPI bus. With this, PIC Clicker could choose SPI connected boards one at a time and communicate with them.
I²C bus is slightly easier. If we have two different boards we will put the address of a chip in a message. For more of the same boards, chip has 3 hardware address definition lines which is incorporated in global address of a chip. This means that we could have to 8 same boards (same kind of chip on the board).

Other two lines in black frame also should be separated for each other in order to allow PIC Clicker to RST - reset particular board or receive INT - interrupt from some of the boards. We need reset in initialization phase or latter if we have to reset particular board for any reason. Interrupt line is used for alert that some board finished a job. This is particularly important if we have to deal with fast processes.

I will try to represent stacking of mikroBUS drawing in next post.

Paolo