As I mentioned at the end of my last blog post, I want to scan some of the parts into 3D digital models. I'm making almost everything from scratch on the build, including the small tip inserts for the slats and flapperons. Normally these come with a kit and are made of either fibreglass or more recently are blown plastic molds. I could just purchase these, but Ron has originals from a 701 which shares the same sie and shape of the 750 ones. Purchasing is easy but expensive and doesn't do anything for increasing my learning. Making my own may not be much cheaper in the long run, but certainly equal or less and making my own also means I can learn some practical skills that come from 3D modelling and printing. First step in this process is to 3D scan the original tips. Again, I could just purchase a 3D scanner and get at it, but what fun would that be? When Microsoft brought out the XBox gaming system, they shortly after released a sensor system that can detect player movements and translate that into interactive game play on the screen. I believe this was in response to the Nintendo Wii game system which had already broke ground and was first to market with this type of player interface. Microsoft took the best of what the Wii motion sensor did with infra-red (IR) and expanded it to include camera capable of sensing colour, faces and more refined depth of field. Enter the "Kinect". In this past decade of electronic and programming experimentation, it wasn't long until someone (much smarter than me I'm certain) said "Hey, I wonder if there is a way to hack this XBox sensor and piggyback on what Microsoft developed for other things?" One of the first uses was for robotics control - robots that could see (sense) and recognize objects. This quickly led to 3D scanning for types of objects, both for item manipulation and avoidance (is the obstacle in my way too big to move or is it of a shape I can grab/push etc.) These type of developments often branch out to other things, including 3D printing. Think about the possibilities! Being able to 3D scan a rare car part and print a replacement for example. Scanning and printing replacement bio-mechanical pieces (heart valves). Printing materials are also evolving - industry is now printing everything from concrete to rubber to aerospace alloys. Like 3D scanning, 3D printing has come also come to the home/hobbyist workshop - makes sense, these home hobbyist are often on the leading edge of these things, at least initially. At thankfully for less knowledgeable people like me, they often share their knowledge online - thanks YouTube and Instructables.com! So, where to get started. I picked up an XBox 360 Kinect sensor. It is the most current one being used by 3D scanning hobbyists and has wide ranging support. The hack of the sensor requires 3 items. A 12 Volt power adapter (bottom left and middle), the male end of a USB cable (top) and the Kinect sensor itself (cable end on the right). Normally the XBox console gets power and sensor data directly from the Kinect. As a result (and probably because Microsoft wants to control everything) the Kinect has a proprietary plug similar but not exactly like a USB end. The third party 3D scanning software runs on a Windows computer, so that requires a USB connection. So, my hack requires replacing the proprietary XBox connector with a USB and also injecting 12 Volt into the cable to replace the XBox console power. I found this wiring diagram in one of the online tutorial videos. In this case, the author wanted to dual-purpose his Kinect sensor for 3D scanning and maintain it for gaming use. To do so, he added a switch in his diagram - I won't be doing this, I don't intend on reusing this for XBox, so I can eliminate the switch and the XBox end shown on the left: First step was to clip off the unneeded end of the USB cable (the phone end in the case of my sacrificial USB cable) then strip off the outer jacket of the clipped end: Strip back the outer shielding if there is some and the inner foil shield if there is some (cheap cables don't have these, that's why they are cheap!): Trim away the two shields, leaving the traditional white (data -), green (data +), red (5V +) and black (ground) USB wires: Repeat the process with the Kinect cable (cut off the proprietary plug and strip/trim the shielding: First thing I noticed once the shielding was pulled back was an extra brown wire I wasn't expecting....hmmm.... I was expecting a gray wire. Wonder if the diagram is referring to the outer shield, it's kinda gray? A little further reading in some of the comments on the YouTube videos and some of the instructables pages I quickly discovered that Mircosoft switched to a brown wire from gray at some point. Problem solved. I tinned the wires first after stripping of the insulation - this makes soldering them together much easier when the time comes. I also added thin wall heat-shrink tubing to each connection which once I confirm everything is working, will be shrunk to tighten everything up. Next, solder white to white, green to green, red to red. Next, add in the 12 Volt supply lines. Positive 12 Volt from the wall adapter to the brown wire (gray in the diagram). Lastly, black wire from the USB side, black wire from the Kinect side and Negative 12 Volt from the wall adapter (hard to see in the picture sorry). Next, connecting to a computer and powering it all up - hopefully no smoke escapes! Unfortunately, my laptop doesn't have a graphics card that is supported by the 3D software, so I'll have to wait to get my home server back up and running to test this, but should be good! Not much I like better than wiring projects, can't wait to do more of this on the airplane. I'll file this in tools for now and get back to it soon. Want to get the 3D scanner working so I can scan the flapperon and slat parts I mentioned above then print them. Carbon fibre anyone? :) Stay tuned.
2 Comments
A real good, full and productive day today at the shop. I'm pulling together the last of the details about the elevator. Started off today by laying out the access hole for the trim servo: The hole turned out well and deburred nicely. I think I may have to enlarge the hole some, but I've already cut my cover plate as per the plans: The easiset way to make sure the cover plate is centred over the access hole is to extend the layout lines far enough, that layout lines copies to the cover can be lined up: The plans call for A4 pulled rivets to secure the cover - I think I'll add riv-nuts like I did on the fuel tank cover of the 701 wing repair. I want access to the servo during routine maintenance inspections and drilling out rivets each time doesn't make sense to me. Next up was figuring out how the elevator connects to the stabilizer. According to the plans, the centre elevator hinge requires a bushing between the attach bolt and elevator horn and the specs are described in the drawings below. I cut a small length of bushing tube material and confirmed the diameter is 1/4 inch: Initial width was brought close to 2.5mm using the bench grinder. Pro tip - hold on tight with a pair of pliers, the busing tends to go flying across the shop if you don't (this was my 3rd attempt!) Careful hand sanding brings it down to the correct width of 2.5mm This is where things got a bit confusing. The plans clearly show the bushing extending on either side of the elevator horn - 2.5mm is barely wide enough to protrude out each side of the horn? Should the bushing rotate in the elevator horn? Right now, my bushing is almost a press fit in the horn with very little to no movement. The bolt rotates freely in the bushing. I'm guessing the bolt rotating in the bushing is better than the bushing rotating in the horn - easier to replace a worn bolt/bushing than a horn? Also, I've been told that any bolt involved in a moving assembly should be drilled and cotter pinned or safety wired. The plans call for a nylon lock-nut here, but I'm kind of leaning towards a drilled AN3-5 and appropriate castle-nut and pin. An email to Roger at Zenith cleared things up nicely. The bolt should clamp the bushing and the horn rotates around the bushing, not the busing around the bolt. Roger was kind enough to send a reference page that explains all this: So, as counterintuative as this is, now I know what the correct assembly should look like (just like the plans). One difference is that the reference page explains the dimensions better. If you look closely the bushing should protrude a minimum of 0.5mm on each side of the horn when the nut is tightened. To make the bushing rotate freely, I ran the bushing tube on the belt of the bench sander to very slightly reduce the diameter, only enough so that it would rotate freely in the elevator horn, and then cut it to the correct width: I test fit the hinge and everything looks good. I used a non-locking nut temporarily to hold everything together, the proper nylon lock nut will be added at final assembly. Just enough clearance and no slop in the hinge - very good. Next up, I had to devise a way to accurately place the hole for the outboard hinge pin in the wing fences. To do this I made a template for each side using card-stock paper. To make a clean hole that doesn't tear open and also matches the 3/16 diameter of the outboard hinge pin, I placed the card-stock between to thin scraps of wood, clamped it together and drilled through: Placing the template over the pin, I now have a convenient place to mark a reference line (the wing fence rear rivet holes) - sorry, took a pic from each side as I was doing this: Black reference marks on the card-stock where the rivet holes are: Connect the reference marks and voila - when laid across the wing fence, reference marks line up with rivet holes and the drilled hole in the card-stock shows where the hinge pin hole will be. Easy! From here, I drilled out the hole to 1/4 inch, in anticipation of adding a bushing: The plans don't call for a bushing on the pin, but there is room for one and I'd prefer that for smooth rotation and protecting the aluminum of the wing fence. To make the require bushing I added the inside washer then measured out the pin to where the other washer will be inboard of the cotter pin hole - turns out to be about 6mm between the washers when assembled: It's hard to position this completely as the clecos protrude out from the hinge pin plate, but this is close enough for now. Unlike the centre hinge, this bushing is free to rotate in the wing fence and around the outboard hinge pin. It will be well lubricated and I may add some more washers on final assembly to tighten up any slack: With the Arduino trim control pretty close to final set up, I wanted to double check the power source as it will be in the plane. The aircraft will have 12v DC nominal power from the Corvair engine, so I need a way to regulate this down to 5V for the Arduino board and the servo. Welcome to the "Buck Converter" - it takes up to 48V DC input and brings it down to a flat regulated output of the users choice. The brass set screw on the potentiometer allows the user to set the output. The beauty is the output remains a constant flat voltage, regardless of changes in the input (say when the alternator is charging the battery circuit). Once set, I'll add a drop of LocTite to prevent it from changing due to vibration. For testing purposes, I borrowed a 12V utility battery from Ron and hooked it up to the converter, and the converter to the Arduino test board: The Arduino trim system runs exactly as it would powered by the USB cable - perfect! I might use the converter below for powering the trim system. The trim system doesn't draw a tone of current (very little actually). This converter does the exact same thing as the one above, just not as bulky as the other one - reducing the footprint even further. The converter above might be better suited for powering LED landing/nav/strobe lights. Arduino will be used for lighting control too. Next up, I started working on the mounting bracket for the trim servo. I grabbed a previous incorrectly bent bracket to re-purpose as my mount: Once cut down to size, I laid out the lines for the servo mounting hole: A cutoff wheel in the Dremel and the air-saw made cutting the mount hole easy: With the bracket built, I can work on getting it mounted inside the elevator. I haven't got the push-rod or connectors yet, so I'll wait until I have those to mount the servo. I'm not sure this will be enough support for the servo either. The plans call for the Ray Allen trim servo to be riveted directly to the elevator skin - I'm thinking I might do a rework on my mount to attach it to the elevator spar. What I could do in the meantime is fabricate and mount the trim control horn. I cut the piece a long time ago and bending such a small part was challenging but I got it done fairly quickly: The control horn will be riveted to the trim corner. The horn end will be drilled once I know the size/type of push rod connector I'm using. It took a long time to get to this stage and I'm almost ready to.... pull it all apart for deburring, priming and reassembly! It's nice to be able to move the elevator up and down now that it's attached to the stab - no binding, no grabbing. Nice and smooth! Really like how the fences turned out too! With the horizontal tail group getting close to completion, it will soon head to storage. Next up will be the slats and flaps - both parts of the control surface systems on the wings. I've been spending a lot of time considering creating a 3D digital model of the slat and flapperon tips. Normally these are blown plastic inserts or made from fiberglass and way to expensive to buy from Zenith. My theory is that I should be able to print them on a 3D printer - possibly even out of carbon fiber!! To create a 3D model easily, it should be possible to scan an original (which are the same as the 701, which Ron has several in stock). Wouldn't that be cool! There are many commercial grade (read expensive) 3D scanning tools available to buy. But what if I could hack a 3D camera sensor that used to belong to a Microsoft XBox gaming console? Yup, we can do that thanks to YouTube tutorials. I picked up an almost brand new XBox 360 Kinect sensor bar for $15 dollars in Facebook Marketplace! Next up, hacking it to interface with my Windows laptop, scan example items, and print them! <Insert evil laugh here> Thanks for following along, more to come! Happy New Year everyone! With the Christmas holidays over, its time to get back..... WAIT!! Failed well pump at home - replaced/repaired. Failing clothes dryer - replaced. Updated home network infrastructure.... started and functionally complete, but need to do more. Check engine light on the car and new front struts installed. Back to work... shift shuffle meaning an extra day of work to balance out the hours....yay me. So, unfortunately the shop has taken a back seat for a couple of weeks. However, I AM BACK! First up, finishing the centre section drilling out to A5. There is no question the number of rivets make this centre section solid: Drilled out the rest of the elevator skin and nose skin holes to correct size: Caitlyn came over with me to the shop on boxing day to shoot some pics of me working. Here I've taken the nose skin off the elevator to prep the outer hinge plate/pins. The plates are made of 4130 steel, so it's important to centre punch them so the drill bit doesn't wander: A bit of WD40 helps cool the bit, 4130 is a lot harder to drill than aluminum: First holes drilled and plates clecoed in place - back drill to A3 then A4, then A5: Nose skins back on, back on the bench with the stabilizer to start lining everything up, and... .... uh-oh.... some interference between the nose skins and the centre support brackets. Apparently this is a common problem which is easily remedied by trimming the nose skin slightly and trimming the back the rudder support plate edges to clearance the nose skin as it pivots. The fit of the hinge pivot point is still bang on - good! Before pulling them apart again, I used a sharpie to rough out where I need to remove some aluminum: While I had some downtime during a nightshift at work, I rigged up the trim servo and Arduino controller along with an example rocker switch. I had set this aside for several weeks but I wanted to refine my programming code a bit. This is a refurb laptop I fixed up and it took some more work getting the proper drivers for the Arduino installed amongst other things, but once I had those my new code loaded up perfectly and it looks like I've got the system nailed down to do what I want! I brought the mock-up to the shop to show Ron but decided it would be a bit easier to demonstrate by mounting it to a board. I've added some spare LED lights to represent a cockpit indicator and a suplus limit switch to represent a momentary contact switch. I also added a really fancy post-it note flag to the arm of the servo to make it easier to see in a video. Here is a short video describing the components of the system, my reasons for doing so and a demo of what it currently is programmed to do. Be kind, I'm no Martin Scorsese HA! I've been thinking hard on how to bend the elevator trim tab. It's 025 thick and quite long. In addition, it has some complex tight bends that will be hard to do effectively on the bending brake, so I had to really think out an order of operations. Like I usually do, I made up a test piece from some scrap 025 and used it to judge if the plan dimensions accurately reflect the true fit of the trim slot. Unfortunately, they don't quite fit - the gap is too wide to be covered effectively by the piano hinge. I had previously cut a chunk of 025 to the flat dimensions called for in the plans. This the same dimension I used for my mock up piece. This wasn't going to work. The challenge is making the strip wide enough to account for the bends, the depth of the trim tab, the fore/aft distance in the slot and the overlap where they meet the hinge......hmmm.... After much thinking over breakfast coffee it dawned on me - why constrain myself to the flat part size listed in the plans. Why not cut the flat dimensions a bit wider or taller, then trim once I'm happy with the fit? I hate wasting aluminum in this way, but if I oversize the flat dimension enough to clearly cover the trim tab and enough to make use of the cutoff for something else, there won't be any worse waste. At minimum, I'll salvage the cutoff and my original flat piece for other things yet to be fabricated. First bend is the small tab at the hinge joint. The bender does good work on this, but can't bend far enough closed on this radius, so I had to bend it down further by hand. Fastening it the bench and leaning on it with a 2x4 worked. Next I bent the lower angle on the bender - this one is much more open so it worked as expected. I placed the sheet in the approximate position in the trim tab slot. Width is a bit close will trim it down a bit to avoid any interference with the trailind edge. It''s clear from this picture the sheet I cut is well wide enough to cover the return bend back to the hinge. Next I started fitting the hinge. The key here is to align the hinge left to right to take best advantage of a full barrel at each end. The barrel of the hinge alse needs to fit snug up against the trim spar. I used a small spare piece of hinge to mark out the correct length and it fits perfectly to the spar laterally. On top is the spare hinge, length marked in green and my new hinge below it marked in red where I want to cut it A chop saw makes quick work of the folded hinge, cutting though it evenly and easily. For the next step I pulled the pin out and ground down one side of the hinge arms to accommodate the safety wire required at each end to prevent the hinge pin from working it out - that would be bad! To allow for safety wire, the hinge pin is slightly shortened to be just long enough to reach the last barrel at each end. I remembered someone online suggesting using the bench grinder to slightly chamfer the rod ends to make sliding it back down the barrels easier. Just a small thing but made all the difference when reassembling the hinge several times over the course of this work. I also plan on using similar hinge to close up the engine cowling which will be open and closed more frequently - more on that later. I flipped the elevator over again. When I was drilling the elevator skin, I didn't drill the holes where the skin meets the trim spar because I was waiting to see how the hinge would fit. As it turns out, this was a good idea, it saved me having to drill twice. I marked out the 40 pitch hole placement, placed the un-drilled hinge between the skin and trim spar and drilled it out to A3 (picture is after layout hinge pin not placed as of yet). With the hinge apart again, careful drill of a small hole through the last barrel at each end to thread the safety wire when final assembly of the trim tab is complete. In this picture, I've already drilled out the 40 pitch A3 holes on the spar side of the hinge. Folding it over on itself, I marked the holes though to the trim tab side of the hinge (black marker dots on the left), then used this as a centre line (black) for plotting my holes on that side. I decided to offset them 20 mm from the spar side (blue tick marks) Flip the elevator back over for viewing and dry fit of hinge to confirm safety wire hole is accessible AND viewable, both for installation and for routine pre-flight checks. From here I removed the hinge again and drilled out the A3 holes on the trim tab side of the hinge as marked. With the bend width confirmed, I used the bender to create the trailing edge bend. It's tight and the bender can only bend so far over. From here, a 2x4 it used to lean on it and bend it further down to match the first. This picture clearly shows the long side of the trim tab will extend well beyond where it should - just like I planned. With it close to coming together I flipped the entire thing over and secured it to the bench. This gives a much better view of the planned overhang. A quick sharpie line down the length gives a good line for trimming away the excess. I drilled both ends of the trim side of the hinge to the trim tab and one in the middle for good measure. With it clecoed together in these three spots, I ressembled the hinge halves, clecoed it to the spar/skin side. Next, I proceeded to drill the rest of the trim side of the hinge to the trim tab using the previously drilled holes as a guide. Took the whole thing off the elevator (again) and used the duplicator to match the holes on the overlap trim tab skin. I love this tool! With everything clecoed together again, I checked the movement - nice and smooth and no binding. With everything good, drilled everything out to A4 final size Voila! (gratuitous happy moment capture) With a few more minutes to spare, I decided to put everything on the bench again and line it up. Before doing that, I made the modification to the centre hinge plate that will allow the elevator nose skins the room needed to move up and down. I also trimmed the nose skin slightly to avoid any interference with the centre hinge support bracket. With everything lining up and measured correctly, and confirming the whole assembly is flat and level, I marked out where the centre hinge bracket meets the elevator hinge spar bracket. With those marked where they meet, I took them off their respective assemblies for drilling. I started with an A3 hole which will be enlarged to the correct size next. Holding it together with an A3 cleco, I confirmed they won't interfere when the elevator pivots up and down. I stopped here, because when I was reviewing the plans, I have concerns about the size and type of the bolt used as the hinge pivot. The plans call for an AN3 bolt and vinyl insert lock nut. Not only does this seem awful small diameter, I believe it would be wiser to up-size the bolt diameter to an AN4 bolt and use a castle-nut and cotter pin to secure it. This mod is an improvement, I'll have to do some research what size bushing that will require. Very happy with what I've accomplished so far. Next up is finishing the elevator/stab connections and fitting the servo and trim actuator rod. Thanks for reading :) |
AuthorHusband, father and 911 dispatcher. Long time pilot with a licence that burns a hole in my pocket where my student loan money used to be. First time aircraft builder. Looking to fly my own airplane. Categories
All
Archives
November 2020
build log
|