The search for precision...

"The search for perfection is all very well.  But to look for Heaven is to live here in Hell"

- Sting "Consider me Gone"

Alright, perhaps the above quote is a touch extreme, but only a touch.  After failing to achieve an acceptable degree of accuracy for my block lifting mechanism with laser cut parts I've started down the path of 3d printed parts.  Many thanks to +Duane Johnson for the use of his delta printer and for putting up with my endless revisions.  The first part I designed and printed was this:

This is the carriage piece that will attach an articulated block to the rest of the lifting mechanism.  Unfortunately, my design had insufficient and/or no tolerance so none of the hardware apertures we're large enough.  Thus ensued a bunch of sanding, grinding and filing until the hardware DID fit.  So, back to the drawing board.  Before I wasted a bunch of PLA on another failure I decided to design a calibration sheet (stepping bore holes up by 0.5 mm for a few iterations).  Designed and printed, it told me what I needed to know about tolerance.  Next came this:

  

The lift head, a piece that keeps the power screw and alignment shaft in place and stops the carriage from lifting to far.  Made a mistake assuming that a 1 mm wall of PLA was sufficient to hold the nut for a set screw in place, the part broke as soon as I tightened the set screw.  Back to the drawing board.  Undeterred I redesigned the lift carriage and came up with this:

Success, everything fit and was lined up, the linear bearing even pressure fits into its bore hole (eliminating the need for a set screw, wee).  Emboldened by success I redesigned the lift head and printed this:

A good enough part, it warps a bit when the set screws are tightened, I'm probably going to shrink the bore holes a smidgen so the hardware will pressure fit.

Coming soon, I get too fancy with a part's design and need to go back to the drawing board, again.

Block lifting in action... slowly...

So here we go folks, actual evidence that stuff I design sometimes sort of works.  Im using a SM-S3317S full rotation servo to drive the set up.  I have yet to puzzle out how it "works" with the Arduino servo library, that is to say I can figure out how to get it going, but not how to stop the damn thing.  I'm sure a little more research will reveal the answer.  At the end of the day that's not the servo solution I'll be using, so doesn't matter too much.  As +Jason Hoff pointed out I need to use a power screw with a more course thread to increase the blocks speed, thankfully I already have some different screws for this use.  Anyway here's the video:

Closing in on v0.2 assembly

First let me say that upon assembling block lifter v0.2 I have found numerous flaws in design (I'm sure I've said this before).  My solution for most of these flaws was drill bigger holes and use more glue, this strategy seems to have lead to misalignment and imprecision.  Regardless I think this prototype will basically do what it needs to.  Going to redesign the carriage component so it's easier to assemble and align.

Great Ape Adhesive, et cetera...

Yay, finally a post about progress as opposed to a post about how I dislike national hardware store chains.  In addition to prototyping designs I've also been experimenting with a variety of adhesives, all of block lifter v0.1 was glued up with one version of epoxy or another.  As I'm putting together v0.2 I've used epoxy for some of the heavy wear and tear parts, I'm also seeing how Gorilla Glue works (since I've never used the stuff).  Biggest surprise with Monkey Paste is how much it expands as it sets and cures.  Sure stuff I've read a variety of accounts (including the instructions) saying that it expands (3-4 times), but actually seeing the stuff bubble it's way out of joints is weird.  Here are some pix:

(The assembly for the block lifter)

(Almost everything about this carriage part is rough)

The above pictured carriage brace is probably the most jacked up part of the prototype, it's going to need to be completely redesigned.  The tolerances were too tight and the sockets for power screw's nut was too small by a few millimeters.  This is, of course, what prototypes are for, screwing things up and fixing them with jury rigging parts and lots of glue.

(drive train gears being glued together)

Vivisection of a servo

I'm redesigning the block lifting mechanism, version 0.1 (everything you've seen up till now) is too Rube Goldberg-ish (too many moving parts that rely upon the goodwill of the universe for alignment and proper functioning).  It'd probably work if I were a mechanical engineer, but I'm not, so discretion is the better part of valor.  That being said I plan on finishing 0.1 just to see it work (pictures and video will be taken and posted).  Anyway, I was struck by inspiration while I was daydreaming about a potential future project that involves a power screw to lift portions the sculpture.  Ugh, so blindingly obvious, use power screws to lift and lower the blocks.  Each block can be independently raised lowered with with power screws driven by small continuous rotation servos.  This solution would eliminate the need for a higher voltage power system just for the solenoids (to engage the rack and pinion), springs to counterbalance the solenoids (to disengage the rack and pinion) and the rack and pinion.

To this end, I've started looking at inexpensive continuous rotation servos.  One solution that +Jason Hoff suggested is modifying a Tower Pro MG90S servo (I have a couple laying around).  There are a number of tutorials claiming to modify this servo (they're usually not, people generally modify the TGY-90S which has nylon gears), so I disassembled one of the MG90S to see how difficult modding it would be for continuous rotation.  Here are some shots of the vivisected servo, I want my trusty EE to look at the circuitry before I start soldering in resistors and bypassing potentiometers (the MG90S looks a bit more complex than the TGY-90S).  So, once Jason has his ankle bolted back together and he's off heavy tranquilizers I'll see what he thinks.

(lovely, durable metal gears)

(stops that need to be ground down)

Block lifter V0.2 schematics will be forthcoming soon.  Once I've figured out how to mod this little beast I'll post a full and detailed tutorial (as opposed to a video of the the final result and vague instructions on what's needed and how to do it).