So, first, what is going to make my dragon fly?

As a said before I think that the best mechanism to replicate the movement of the Pegasus will be a crank. A crank is basically a lever attached to a rotating shaft, the diagrams below are from a great book ‘Cabaret Mechanical Movement’ by Aidan Onn and Gary Alexander, highly recommended if you want to do any work with mechanisms!

(scan of crank images, label parts)

I don’t think I’ll have enough time to cut my parts out of brass, no matter how beautiful the end result will be. To save time it would be quickest to draw the components in 2D Design and manufacture them on the laser cutter.

Fist things first, what do I have to make the pivots for everything to move around? Looks like I’ve got 3mm acrylic rod to work with, so what size hole do I need for this to work? I’ll need a tight fit for fixed points and one for moving points but not too loose, I don’t want any rattle:

Using 2D Design it was easy to manufacture test pieces, looks like it’ll be 3.2mm for the moving points and 3mm for the fixed points. Even though the results seem obvious you never know if the laser cutter will melt a little extra away from the cutting line, it all depends on how the machine is set up.

Next, the length of the lever connected to the dragon. This is all linked together with the size and position of the cam for the crank so…

The part of the drawing to the right is an accurate representation of the dragon at it’s lowest point on the crank. There appears to be enough room for everything to move so I think I can go ahead an design the rest of the dragon:

The additional pieces are all to make up the layers that move against each other. So now I have my dragon, the box size and shape and the cam drawn, I think I’m about ready to start construction!

Clang!! first problem, I’ve not left enough space for the vertical rod to fit through the top of the box! Back to the drawing board:

Right got there in the end, time to do a test fit:

Looking good till I tried the mechanism

Crud, he’s all floppy and moving way too far back and forth, I tried fitting some scraps to close the hole in the top up a bit to see if it made a difference:

Hmm it’s a bit better but not by much, I’ll need to alter the mechanism if I’m going to make much of a difference. To reduce the ellipse that the crank describes at the top there are a few things you can do, make the cam smaller, move the shaft and pin points closer together on the cam or to move the guide higher up. As I’ve already cut the box and I don’t want to waste material re-cutting it I’m going to ry a combination of the first two:

The one on the left is my new design, I’ve also included a couple of new components to limit the size of the guide hole in the box lid, time to test again:

Much better! I’ve made the hinges for the wings out of some acrylic tube cut into three pieces, the middle stuck to the wing and the other two sides stuck to the body with a piece of rod in the middle for the pivot, to stop it falling out I ran a little acrylic cement in the end of the tube at each end. It’s still a little clunky but hopefully the linkages will reduce a lot of that.

I did manage to fit the keypad in place and the holes for the LEDs did cut through so I was able to see how it would look when they were lit up. The keypad fit was vary tight, possibly due to the poor cut quality but I have altered my drawing to give it another 0.5 mm. The holes in each corner of the aperture needed to be moved as well, they didn’t quite match up to those of the keypad itself and since there is no wiggle room they need to be as close as I can get them.

On a brighter note the hinge idea does work, after drilling a hole in two scrap pieces of acrylic I can see that the door will swing between the two, this will be much more secure once the safe is constructed it will be locked in place by the case. The back elements of the door all cut through thankfully so I was able to test the fit of the components and the PCB inside…

The board fits! (not that I had any doubts…) once I have trimmed down and neatened up the wiring it will sit better in the case. I’ve included a little notch out of the left hand side of the box to accommodate the battery cable, it will need extending as I’m placing the battery pack on the underside of my safe. The only major change needed is the port for the solenoid (see below), it’s far too small and where it is placed at the  moment the solenoid board will intersect with the main PCB.

So only a few minor changes needed and some testing to get the right speed and intensity for the laser…

The supports would be aliminium bar with the surface pared away to give a good fininsh, these would have a hole drilled into either end to allow the posts on the CNC manufactured ends to fit in and be fixed in place using araldite liquid steel. The inner support posts will have to be made on the lathe as they need to be milimeter perfect to follow the calculations I made so the back of the door would fit perfectly. Right on to manufacture…

(some time and much effort later)

OK I now have 4 exact supports and the 8 tops for the inner case and 6 outer supports and 12 tops for the outer case. The lathe is simple to use as long as you follow two rules 1) measure and 2) measure again. The machine relies on human input to tell it where the end of the billet is for it to work accuratly. Here are some pics and a short video showing the manufacturing process:

All the pieces slot together so I can construct the box and check any fit issues without needing to glue anything.

After extending the battery cable I realised that I have not allowed space for the on/off switch, a bit of a silly oversight but one easily fixed. I produced a simple drawing in 2D design using the layout of the piece the switch is to fit to, by making the paper setup the same dimensions as that of the piece to be cut it was a simple matter of placing the acrylic back into the laser cutter and setting it off, hey presto…

The next step was to test the door in it’s final position. A couple of changes need here as well, as you can see from the first picture below the door would not shut with the battery cable leading down to the pack, the inner case needed a recess for the cable to sit into to allow the door to lie flush to the face of the box. A better solution however was to cut a hole to pass the cable through as this would prevent the cable from slipping out of place as in the second image.

Another problem was that the door was catching on the metal supports on the inside of the green box (see top image). I had already faced them off on the manual lathe to get rid of the bump from the CNC lathe (see below) to fix this I decided to drop the box by 2mm by taking the inner supports down on the manual lathe. This fixed the clearance problem and allowed the door to swing clearly.

All that is left to do now is to finally fix everything together permanently, starting with fixing the inner supports to the bottom of the blue box and the bottom of the green box. I decided to use Araldite liquid steel as I had gotten good results when using it in my other project. Each support was pushed through it’s opening and coated with adhesive, the connecting piece was then pushed down and held in place with masking tape till dry. I also glued a piece of black acrylic to the underside of the inner box to attach the battery pack to, this looked far neater than seeing the glue through the bottom of the inner box.

Once the inner supports were dry I could use the capillary glue to run along the edges of each tab, tape it securely and wait for it to dry. The outer box was a little more complex as the door needed to be in place before it could be completed.  I started by gluing the outer supports to the base, fitting the door, placing the top of the outer box and checking the movement (hmmm… door is a bit squeaky, may need scraping down…) once everything was in place the tops could be glues in place the the capillary glue run allong all edges. The box is finally complete :).