mechanical gain device

[webby2]

Doing some preliminary testing of the main input control system.

It is not as efficient as I would like but it is within what I was expecting,, I can always hope for ideal but I except what things are,,

I am still waiting on a few parts to show, I ordered some slippery tape to use between a few parts that slide on each other, I thought that using the tape would be cleaner than using lubricant and easier to change things out if needed.  I am also waiting on a flange that will solidly lock the main shaft to the main input control system.

I will start up on the ancillary parts next week and when those are done I will be able to test the whole system.

So far all the testing I have done is what is expected and they are in line with what my simulations predicted.

[webby2]

I found out a few interesting things today,,

A razor blade knife\ utility knife,, blade, has a center tang that is 3mm wide,, a very handy thing if you are making grooves to fit a 3mm wide bearing!

Too much exposure to the vapors from Super Glue will make you feel like crap,, flu like symptoms,, not happy about that.

I got carried away and rebuilt, a few times even,, an internal part of the main system, it is now smaller and lighter and is all held together by bearings running in grooves, I am still using 7 bearings for the whole system but now they are running in grooves,, once all cleaned up and washered it will be very nice.  I made a small change to it to reduce some sliding friction and the tests showed a good change so I thought that if I removed all such things I would only be left with the gear mesh as a large frictional loss.

I have to grind down washers so that they become bearing spacers,, not the best way but it works.

[webby2]

Still playing with the main system,, I broke a few parts a few times and so thought that it would be prudent to make sure the system can run for a few hours at a time before moving forward.

So far things are going fairly close to what the sim was predicting, and if the main system follows those predictions then I see no reason that the full system should not also follow what the sims are showing.

I change how I was using some bearings, well I have them secured by bolts and some of those bearings pass over each other,, with the first designed usage I had all sorts of head room and did not think about the clearance,, but it turns out that when you lock the bearings down and close those clearances the bolt heads can collide,, and that breaks things after a while of hitting,,, my bad on missing that.

[webby2]

I need to stop with the time estimates,,

So I found a flaw with the build design, that is it allows for movement in the wrong plane.  If the levers were moving in the horizontal plane what I get after a lot of run time is an oscillation in the vertical plane,, this comes from the stupid fact that I used a single single row bearing for the pivot of the arms,,  I have changed this and am re-printing the whole thing due to the increase in sizes of all the parts.

[webby2]

I ordered some parts, for the system and for my lathe.  I broke the belt for my lathe and I need it to cut the cir-clip grooves in the pivot shafts.

I had to modify my printer a little bit so I could print a part 4mm bigger than what my printer would do,, the bed was to small.  I got that done and am trying to clean up the large part and get it within a reasonable tolerance.

It is taking longer to get parts from overseas, whereas they used to be a stocking item they seem to be more of an on demand thing,, more time waiting.

I have most of the other parts printed and cleaned up, one small print flaw to fix, a small warp in one of the parts, if I can not fix it nicely I will just re-print it again.
I also made a final decision on my ancillary parts, that is how they are going to be setup and what I decided will take a lot of fiddling around,, lots of small parts that need to be arrange within a bigger part and then a few bigger parts assembled into the final part, I need to make 2 of those.

[webby2]

I gave myself permission to get sidetracked over the December-January time period, that and I ordered some wrong parts.

What was fun with the sidetrack is in what was not the same as my main build, looking at things a little different so I can have a more positive feeling about what it is I am doing.

So I am going to repair my lathe and get back to building, I have a lot of tedious parts to make and have not been looking forward to it.

[webby2]

I have been playing with the sim.
I have a scaled version of the system sim'd up and was working on my run topology for the controls,,,, well I messed up one the changes I was looking into and this is what I got as an output,, a rather large cost for an ideal device

I laughed at the output graph, it is such a large difference that nothing needs to be done to see that it is a large loss.
anyway, I suppose you can see that I inverted my topology by mistake.

[webby2]

Here is a pic of a planetary thingy I got sidetracked with,, this is just one of many variations I was looking at.

Some interesting stuff with my main build,
First the control that I need for the control unit is more complex than I can make, so I simplified that part and went back a few steps in design.  This means that I can no longer have very fast load control or anything like that, the system will take the time it takes to adjust to any change in load,, so I will try and make the load match the system,, I was hoping to be able to make the system match the load.
Second is an observation,, my input adds energy to the system as a whole but at the same time reduces the angular momentum of the system while maintaining the system rpm.
The trick there is in locating the mystery storage area

[webby2]

In my opinion this is what a successful simulation of an output gain might look like

[webby2]

This is just the motor power from the same run

[webby2]

and this is the total power only

[webby2]

I have the sim set to 100 degrees per second of main motor rotation, I wanted to increase the system frequency up but chose to keep it slow so that I could get a reasonable amount of data points for the time it takes to run the simulation.

The other input power is not so important really, it is the torque of the main motor that matters and with the motor set for a constant 100 deg/s the torque follows the power curve with a peak of approx. 1000 N-cm (positive) and -2000 N-cm (negative).  Positive torque means an input cost a negative torque means an output gain.

The system add and subtract to\from the main motor is not relative to the actual motor RPM, so the motor output POWER will go up with the RPM of the main motor but the other external inputs stay pretty much the same.  So in other words if the system is making an average out of 500 N-cm and that is at 100 degrees per second then if it is making the average out of 500 N-cm at 2160 degrees per second my input is the same relationship and so with the main motor spinning 21.6 times faster there is 21.6 times more work out.

If I increase the other control inputs so as to increase the frequency of that cycle then the torque values go up as well, in both directions of course, right now that frequency is close to 1/2 cycle per 5 seconds, there are 2 input pulses per 360 degrees of the control input system.  The higher the control frequency the higher the torque swing and I am using more external input to drive higher, but that stays with the same relationships.

What this means is that I can increase the system output by either increasing the control frequency or the main drive RPM or I can increase BOTH.

This sim is actually setup to run in a more safe mode than needed but the plastic parts can only handle so much force and so I tuned it down to reduce a few force interactions to hopefully save from ripping the testbed apart.

With all this then I do believe that I can now finish the build to the specs I have in the sim.

[webby2]

here is some stuff with the motor set to 1080 degrees per second

[webby2]

you need to look at the scale on the graphs

I am re-running the 100 degrees per second to grab the motor torque for it as well as the all others,, I will post it when it is done.

The all others is without the motor, this is to show what my external input is besides the motor.

[webby2]

That went faster than I thought