A little detour to fix a door problem

With all the time and work I have invested on the gripper claw, it quickly became obvious the door panel had a problem:  It wouldn't open all the way.  The door would rub against the skin and not open as widely as it should.

What happened?  When the door and hinge were epoxied together, it looks like a clamp slipped off, pushing the hinge away from the door opening.  As a result, at first glance, it looked fine but when I went to open the door and check the radius, I could tell the hinge was crooked.

Further complicating things was I had used a different epoxy.  I used GOOP instead of JB Weld on these, since I have a love/hate relationship with JB Weld.  For somethings, it works really well.  For others...not so much, especially the Qwik version.

That said, I had to find a way to remove the GOOP adhesive.  Being an epoxy, I found that toluene will dissolve and is suggested by GOOP to remove it.  I went to my local hardware store since no one at the big box store had any idea or personal experience with the solvents.  It was suggested I use a paste/gel type solvent, lightly brush it on, let it sit, pry a bit with a scrapper, then do a bit more in a few hours.

This is really, really nasty stuff so I did this outside, in the garage, with the door open.
(Incidentally, it seems EVERYTHING is bad for laboratory mice)

Once I had the door and hinge off the skins, I let them dry in the sun for a few hours.  I need used the mouse sander to remove the larger residue.  For the rest, I used the Dremel sanding flap to bring the surfaces to a shine yet rough enough for a good bond with JB Weld (later today).

As you can tell from all of these side tracking tasks, my R2-D2 hasn't moved to the new frame.  With 4 events planned for May, I thought it best to leave things alone and working on the new accessories for my second R2.

So, if you have lost track of what I am working on...me too!

Progress will be focused on R2-D2 #2 now, maintaining my first one until #2 is ready to take over.  Once #2 is up and working, I can come back to the first one, paint the blue pieces pink and do a complete upgrade and rebuild into the new frame.

(That plan may change in an hour tho!)

More on the gripper claw, prepping for the CPU Arm version 2.0

Tonight I spent some time reviewing how to fit the gripper arm into the body so that it opens straight ahead.  After a lot of adjusting, pondering and re-thinking....it looks like the best solution will be to remove the breadpan entirely.

The sides of the bread pan either rub the arm or the servo on the claw.  The next step will be to attach the front skin and see how this alignment works without the breadpan.

 The front skins are being worked on right now...trying to remove the crooked hinge still, so I'll have to wait a bit to check out the fit with and without the breadpans.

Meanwhile, I wanted to get the cpu interface arm dismantled so I could examine it...

Here you can see the original setup i had using plumbing pieces.  In the foreground you can see the linear actuator I want to use.

 It took a few good tugs with a set of pliers to get all the plumbing and tubing I had inside the cpu arm.  

Next it was time to visually size everything up.  It looks like once I hollow out the cpu arm a bit to fit the actuator, everything to go together fairly smoothly.

The trick, of course, is making a square piece fit in a round hole.  

Always a challenge, this one.... (Yoda voice)

More work on the gripper arm

Over the weekend I continued to make some progress on the arm assembly.

One thing that bothered me was how close to an inch of space was wasted trying to make a 1/4 20 thread fit into a 10-24 thread actuator.  A coupler was required but I wanted to find something better.

By accident, I think I did!

I found a 10-24 thread grommet that had potential.  I just had to add some holes for a 4-40 screw to pass thru.

 The grommet is too small to reach the existing mounting holes, so I used a small, square-shaped piece of aluminum that would.  Then it was creating two countersunk holes for the mounting to the claw, then two threaded holes for the grommet to be secured to.  (No room for a lock nut where those screws go in)

As you can see, with the old coupler and mounting piece in the foreground.  In an area that already has tight clearances, this helps a lot.

The next obstacle is the attachment of the linear actuator to the servo.

Ideally, you want the servo horn to be screwed into the servo.  However, with the attachment pictured below, that's hard to do...

The L16 Linear Actuator includes mounting hardware, a U-shaped bracket and 10-32 threaded screws and lock nuts.  I drilled a hole large enough for the bolt to go through the servo horn.

 Now I can just use electrical tape on the bottom part for now and then the arm slides onto the servo fine.  However, anticipating Murphy's Law, I would like to find something better.  For now it will suffice.

 

And here is where I am at now, R2's arm assembly moves out, extends and the claw can open and close.  However, the next step in this challenge is to have it going straight out, in relation to where R2 is pointed.

I was going to work on this some tonight but decided to address the issue with the door not opening all the way properly.  This requires removing the GOOP adhesive since it dried crooked (clamp slipped off).  As a result, the door is perfect at the bottom but as you go up, the hinge is falling into the body.  I have some toluene solvent, which will dissolve the GOOP adhesive.

The one good thing about all this time in the basement has been my dedicated assistant, "Mokey".  Anytime I get up, he steals my chair.  When I am staring out in space, trying to find a solution to something, he's rubbing around my legs, desiring some of that attention.  Gotta love a great wingman!

Dome pie panel mounted

It took several tries to get right...and I wound up using the trusty hot-glue gun to find the right location!

Here's a look at some of the progress I am making with the Gripper Claw addition to my R2-D2.  Pardon the masking tape and electrical tape!  I still have quite a ways to go to tweak this to run smoother!

 

More work transferring to the new frame - Dome Drive installed

Tonight I did what I thought would be a quick installation of an AP Dome Drive into the JAG v5 frame.

What I am trying to do is instead of just pulling everything from one frame into a new, I am upgrading components I have been wanting to swap out.  The dome drive was on the to do list since a better drive wheel had been developed.

On the JAG frame, there are pre-drilled holes for the dome drive to be installed to.  You simply have to tap the hole to a 1/4 20 thread and bolt the drive in.  The next step is to trim to drive shaft of the Pittman motor, so it will not interfere with the dome plate.  

Easy, right?

Well...once I installed the dome drive, it was rather obvious that the electronics tray was going to have a problem...

I had two quick thoughts...move the dome drive mount a bit further right or move it to the front side.

The problem with the front idea was that there would be no place for the spring pin mount to attach.  That space is taken up with the front door breadpans and then everything that mounts behind the large data port.  (Translated = Bad spot)

So, I opted for moving the position over a few inches.  The only issue with that would place it very close to where the shoulder mechanisms will be.  To allow for maximum space, I drilled and counter sunk a hole for the bolt to go in from above where the piece is closest to the shoulder.  

This looks like it will work out OK.

Next was to trim the Pittman drive shaft and bolt it into the mount.  When that was completed, it was time to slide on the drive wheel and see how it fits with the rockler bearing and dome plate installed...

The good news is that the dome plate clears the drive nicely.  The new silicone wheel material should be an excellent improvement over the older drive wheel that had a rubber band type material.

I like that the drive wheel has a nice engraving on it.  The hex head screws may have to be replace with something more flush in the future.  But, for now, it works.

Building this improved droid is teaching me something rather quickly....space is going fast.  For this R2 to have the accessories it has now and additional one...is going to be quite the challenge!

Installing a hinge on the dome pie panels

Today I have been working on installing a hinge on one of the top dome panels, referred to as the "pie panels".  

When I first put R2-D2's dome together, I simply used some clear silicone to adhere the panels to the dome.  I made the cut outs in the inner dome so I could come back to this in the future.

The particular pie panel I am installing a hinge on is the one that R2-D2's Life Form Scanner extends from. 

With R2's dome removed from the body, I examine how much silicone I need to scrape free to release the panel...

 

 I had to remove the dome power system, which consists of a fuse block and 12 volt battery.  Once those I removed, I am able to slowly, carefully...and very delicately remove enough silicone for the panel to fall out.

 Once out, its slow work removing the dried silicone.
 

The rest will be removed using a dremel sanding flap (120 grit).  To protect the existing paint, I cover it with Painter's Tape and clean the surface.

Next up was to determine the fit and location for the hinge.  I used some double-sided foam tape to find the best position.

The next step is to add some material to the hinge where it meets the dome panel.  Since I threw away the inner dome panel when I cut them out, I had to make use of a scrap piece from the body skins.  The extra material is required for proper alignment.  I decided to use Super Glue, the Loctite brand that has a control gel formula.

Again, the most difficult part is the location, height and alignment.  When I was happy with the location and moved the hinge around, I used the Super Glue gel, placed it on the hinge and gently laid it into position.   I would have used JB Weld for this part but due to the awkward location, I wanted something that would set faster, allow me to test, then clamp into place.

Once the hinge was in position and had dried (I gave it a few hours to be sure), it was time to position the panel over the panel opening.  With the hinges, you need to position the panel up a bit, so it can open smoothly and not rub the edge of the dome lip.

I used one of the cat beds to cushion, protect and hold the dome in position.

Now for the attachment to the hinge to the dome panel, I opted for JB Weld.  Why?  Because if I should need to remove the panel and re-position, I can do so.  JB Weld is easy to remove, sand and clean up.  

Tomorrow I'll check and see how well the fit is.  If it works well, I can work on the servo installation!
 

Building a Gripper Claw Assembly for R2-D2

A year ago, one of the R2 Builders was offering a replica of the gripper claw assembly.  It was a really nicely done piece but I just didn't have the resources at the time to purchase one.  

If you are not familiar with what that device is, this is a screenshot from The Empire Strikes Back, in a scene where R2-D2 is trying to grab a small lamp from Yoda...

Like most things with R2-D2...what a movie prop can do for a scene is very different for a real R2-D2.  What's that mean?  Well, finding the parts and pieces that can extend out that far, motorized and work as seen on the movie is a big challenge.

Several months ago, I stumbled upon an aluminum gripper arm & wrist combination.  It strongly resembled the gripper R2-D2 has.

I found this Dagu Gripper Arm at RobotShop.com  It has the servos pre installed and has one in the open/close mechanism and another for the wrist motion.  Due to the fact the tall doors in R2 are 12 inches tall, to fit with an actuator, the wrist portion would have to go.

I found a Hub Adapter made my Lynx Motion (also from RobotShop) that would fit on the end of the gripper.  The key to all of this working is the Firgelli L16 140mm actuator.  With just under 5 inches of stroke, in the closed position it completely fits in the door way.  Needed to make everything work together was something to go from the actuator to the claw. 

 The opening is a bit too large to thread with a 10-24 tap.  The L16 actuator comes with a thread-on piece for fitting into 10-24 threads.  Too get a snug fit, I had to drill this out with an #7 drill bit (13/64) and tap it with a 1/4-20 tap.

It was a real challenge to secure a round piece into a vice....but using a lot of WD40 and patient work, it eventually worked out.

 
The coupler I bough from McMaster-Carr.  The female end is 10-24 and the male end is 1/4-20.  I shortened the female side by 4 threads to shorten the distance up for a better fit.

With the coupler installed into the hub, the L16 actuator can thread in now...

Here is how it fits in the door-breadpan that mounts behind the front door...

Total cost is less than $200 and all you have to do is drill and tap one pre-existing hole!


Dagu Gripper Arm ($57)
Firgelli L16 140mm Linear Actuator ($80)
Lynxmotion Aluminum Hub  ($8)
McMaster-Carr M-F Coupler Part # 92499A127 ($9.82)
$155 + shipping/handling/taxes

Uh-oh...Softpot fried...

The Softpot Rotary potentiometer was working so well...then...poof!

It looks like the plastic stylus may have been forced too close and the ripped the sensor pad.  

For the past several weeks I have been working on the software to read and work with the data.  The progress has been slow but steady...then when I went to resume work one evening, noticed the data being output was very erratic.  After removing the dome and checking the Softpot...well, you see the rips!

I did something a little different this time.  Instead of ordering from SparkFun, I bought two replacements from TrossenRobotics.  I was able to call them on their toll free number and immediately speak to a tech.  The way they sell theirs, you can purchase it with a female connector already attached or bare, with the 3 male tabs exposed for soldering.  They also include the coupler pins, so this time around, it will simply be a matter of plugging a servo cable in...no soldering tiny tabs!

Link:  http://www.trossenrobotics.com/store/p/5604-Softpot-Rotary-Potentiometer.aspx

Replacements should arrive next week, then its a matter or carefully removing the old one and putting the new one in!