I love the look and feel of the older games and toys. I think it is high time for me to remake another... I went big again!
This time it is a wall mount etch a sketch.
Over 3 feet wide yet still light enough to shake to clear. Yes it is designed to be filled with the same grey powdery stuff as the original!
I'm enjoying the internal mechanics too much to put the powder in at this time...
UPDATE March 31/2019: I have added the magic powder and started to play with this. My observations and recommendations are in the last step
Step 1: Minor Changes
I took a little artistic license with this one and I also added some electronics.
The original has a screen ratio of about 0.7 so I began there. No plans just a rough shape.
Further I have integrated some electrical and electronic components to facilitate computer control. I made use of a welder, and a 3d printer.
The files for printing the frame are here, be warned you will need a machine capable of printing 10 inches by 10 inches ... The numbering system starts from the top left and finishes bottom right with 3 rows.
For my printer this was most laborious task. It had to be cut up into sections to fit the printer. The print time was several days of continuous run.I had 2 failed parts due to using ABS+, I switched to ABS and had 100% success with all the bezel parts.
Other files for printing are in the appropriate sections.
Assembly of the bezel is covered in step 5.
Step 2: The Frame.
I wanted the internal structure to be non magnetic so I chose to make the frame from aluminum.
I had a 39 inch long 1 inch angle which I used for the top of the frame. Several smaller pieces were fastened together to create the height of 25 inches.
I made a mental calculation error which resulted in the drop down sections being too shallow for the drive components so I had to add section made on my 3d printer, more on this later.
The 1 inch turned out to not be deep enough for the internal structure materials that I had on hand. So I fastened more angle sections to the original so now I had 1.5 inches to work with.
I used JB weld to seal all of the seams, this is very important if you want you place to remain clean...
With the rough frame now available I went about designing then printing the drive components, knob housings and internal cable guides.
Step 3: Drive Components
The knob housing were printed from ABS material
The housing sections consist of a base pan which fastens to the frame, A cover plate which houses a bearing and the Drive pulley.
The base has a support section where the lower pulley drive shaft fits.
The drive pulley is a 2mm pitch with 62 teeth for a standard GT2 belt profile.
Press fit inside the top cover is a standard 6201 bearing(overkill but it has the inside diameter that I wanted). The bearing provides support and sealing for the shaft which will be in an absolutely filthy environment.
The housing is bolted to the frame and later sealed with silicone caulk to prevent material leaks.
This created some unique challenges since I wanted to incorporate a CNC type drive as well as have direct manual control.
Step 4: Back Panel
I cut a section of 1/4 inch thick hardboard for the backing of the frame.
The frame was cleaned with acetone for a secure hold
This needed to be attached with a tight seal so I again used JB Weld for this. The epoxy slumps in its moist state and sealed the board to the frame.
Some weight was added for a tighter level fit.
I needed to make holes to secure the knob mechanism housing this was drilled and countersunk to accept 10-32 machine screws. I uses a hand tap in a drill with plenty of aluminum cutting fluid.
I printed glass retaining clips which were equally space around the frame leaving enough clearance for the knobs.
The holes for the clips were drilled and tapped to accept 6-32 machine screws
The frame was then sent to the glass shop for... well... glass.
For those safety minded persons you can either go with safety glass(1-2 weeks wait time locally) or use a clear plastic cover on the glass.
Step 5: Front Bezel Assembly
Printing large flat panels with abs is troublesome, my machine works 99% with standard ABS, with ABS plus there is no hope whatsoever.
Since the bezel was printed in pieces I now had to assemble it then make sure the joints were all no longer visible.
I designed a peg system to align the panels and uses liberal amounts of ABS solvent cement for the joints. clamping was also necessary.
Step 6: Make It Pretty
Sand, clean, Bondo, prime, repeat...
Since the prints have different grain and texture, not to mention slight warping I had to hide the seams with a lot of finish work...
I chose to prime with red primer for a better colour. This was sanded with 400 grit then the paint was applied.
Locally available rattle cans have limited choices I went with Colonial Red because I found the Apple Red to be a little ostentatious.
I applied a second coat of red a day later.
At this point the bezel is still a little wobbly due to being printed hollow but this will be corrected once the previously made frame is made part of the internal and the rear cover is in place.
Step 7: Do Not Use Rustoleum!!!
I chose to use the old standard... Rustoleum rattle cans...Carefully sanded after priming the first coat was applied.
All looked good...
light sanding with 600 grit between coats and the cleaned with dry cloth...
Second coat and some minor bubbling of the paint occurred around the interface holes., should be OK since it will be hidden by the knobs. I waited 48 hours then no sanding and sprayed the clear coat.
Immediate destruction!!! The gloss clear coat RUINED the entire finish.
Extensive bubbling and cracking of the whole surface occurred.
Nothing to do now but wait 48 hours then sand it and try again.
All the bubbling removed and a coat of primer over the whole mess tried tried, no luck the primer now bubbles.
It would seem that the clear coat chemically affected the remaining paint
The Lesson here is I can no longer trust that the rattle cans will work. NO MORE SPRAY CANS for finished work!!!!
For places that the finish does not matter, these will work, but still...
Step 8: Prep and Paint... Again!!!
Stripped it off down to the bare plastic and began the whole process over. This time the primer and paint was changed to acrylic and applied with a proper sprayer.
I used an HVLP sprayer...
The paint is standard acrylic red diluted with air brush thinner. I had to do a 1:1 mix to get a decent finish.
now it works great.
The only thing is there is way more over-spray... I'm now inundated with red dust. Looks like a paint booth will be an upcoming project.
Step 9: Make It Work
This may seem a little convoluted at first since I usually work on the fly and make parts as required. I really should plan ahead more...
I used Wikipedia Etch a Sketch for information on the internal workings. The functional drawing is from there.
I designed and printed 2 parts which would serve and a backing plate for the drive components and also a mounting point for the knobs.
I also wanted to install and hide 2 stepper motors for computer control of the sketching.
The drive shaft uses a 3mm wide GT2 belt to contact the cable drive pulley. The drive side is an open ended belt looped along the one side and I chose to use cables for the following side since I had several available.
The left side contacts the horizontal mechanism and the right side contacts the vertical. This is the same as seen inside the actual toy. however I had made some mental errors and since I had already made the bezel I adjusted the pulley positions.
There is a loop of cable for the dive that travels from the rod around a pulley P1 (lower left) to another pulley P2 (lower Right) to P3 (upper right) then P4 (upper left) then P5 (lower left again) then P6 (lower right again) back to the display rod. The looping is similar for the Right (vertical) drive.
I went through many variations of rods until I found the right ones. Everything from steel to copper to aluminum in various sizes and shapes.
The best rod that I could find that had the minimum flex and lightest weight was a quarter inch aluminum plywood edge strip.
There are 2 rod to cable connectors for each axis that were printed. One end secures just the ends of the cable and the rod, the other end attaches to drive belt ends and the cable.
The pointer needed to be able to contact the inner glass and handle rod and cable flex. It is fed over the rods before the rod ends are put in place into the rod cable supports.
Step 10: More Internal Details
So with the inner working built on a flat surface, all looked like it was going to work, however it did seem a little flimsy.
On edge it was a totally different story, the cables flexed, the belts folded and there was no way to get precision out of it.
The only solution this late in the game was to make new cable pulley supports and place some sort of track to help the support rods stay in place.
I did have several sections of 1/8 welding rod, they should work as guides.
The tracks are aligned with the corner cable guides and the whole thing is glued in place.
The rod cable and belt supports now needed to be changed to accommodate the new track arrangement.
This insides were painted a dark colour with a substance that can supply the most evil and blackest of surfaces... a Rustoleum rattle can. I know that earlier I had issues with this product, however this surface finish was not critical and most likely would never bee seen again. I just wanted a dark uniform tone.
The cable and belt rollers tend to flex a little so to prevent the belt from slipping I added a thin printed cap/guide to the end and glued them in place.
At this point I added limit switches to the end stops of the display area. These are water and dust tight IP67. totally necessary for this application.
All joints and seams need to be sealed completely with at least a silicone sealant.
Step 11: More Internal Drive Details
The belt retention blocks have been updated to include a locking line for the belt and another screw to clamp to lock the cable for the other end of the pointer rod. This is to preserve alignment if the belt or cable should slip, now they must travel together.
The pointer and dive rod supports have very tight tolerances to one another. this is to prevent binding. The horizontal bar is below the vertical bar.
I installed the pointer head on the drive rods and rapidly moved it back and forth many timed to eliminate sticking points.
The ends of the cables are locked onto the rear side of the non belt drive block.
There are 4 internal switches, two for each axis, positioned as shown for the end stops of the pointer, these are for the CNC operation which has not been fully tested yet.
Step 12: The Magic Powder
I found a local supplier of powdered aluminum, seems legit...
This is mixed with some Styrofoam micro beads to prevent clumping. This is what the patent says, so why not?
More Safety: Don't breathe the dust or you run the risk of forgetting your own name...
On first try this does not work at all. First of all it is not sticky. And Second there is no clumps... Just super fine dry powder.
It is supposed to stick to everything and have a greasy feel. Not at all like aluminum.
The powder in the jar is airborne and does not coat the glass.
I think it is missing one key ingredient...
I began adding a little bit of mineral oil an shaking the jar. The dust immediately cleared.
I added more until I saw... Clumps.
Then the Styrofoam balls went in.
Much shaking later, the inside of the glass jar is completely coated in a familiar tone.
No dust, no clumps, sticky and greasy.
It is sitting on a jar on my workbench waiting to use...
Step 13: Final Assembly
I also printed 2 alignment plugs so the bezel would fit to the body properly.
The body is fastened to the bezel using several printed frame clips.
The knobs need to be big enough for an average size hand to grip and manipulate while having an integrated GT2 pulley as well as kind of look like the original.
Previously I have use a white ABS printing material and found that it kind of sucks but it did have the look that I wanted plus I have a lot of the stuff so that when it gets dirty and ugly I can just make more.
A cover back panel was cut from hardboard and screwed to the points inside the bezel.
The electronics are just electronic drop in place like so many of the 2 axis CNC instructions out there..
I found it almost impossible to get the stepper drive belts in place so I added a separate drive pulley and height extender to the knobs as shown.
Step 14: Aftermath
As a functional proof of concept prototype, this works brilliantly.
I do need to tweak the magic powder formula with smaller Styrofoam balls and I'm thinking more oil. It still clumps and takes forever to coat the glass.
The pulley blocks flex quite a bit and will definitely have to be metal or supported on both ends in the future.
The body could use an extra 10mm thickness for the internals
The drive elements extend out into the visible area so a wider bezel will have to be made
The pointer/scraper was made from a pencil eraser and a more robust rubber will have to be sourced.