I Over-Engineered this Machine, cause Manufacturers Didn't

I've built this oscillating Edge sander from scratch and you will never imagine the mechanism going on inside let me show you what it can do obviously it's oscillating and can take different size drums and the belt for sanding inner and outer Contours but what's always been bugging me with these machines is the drums and belts are this tall and most work pieces are not resulting in only using the lower part of the sanding grid sure the drums I can just flip around but the belt is directional flipping it around you could catch the seam and rip it apart so a key feature of my sander is a height preset I can lower the whole mechanism down and use the upper part of the belt isn't that fantastic it also works super smooth changing belts is done easily by flipping a lever and the whole belt attachment comes off with this lever sweet one feature I decided against is a tilting table that's just inconvenient on my machine the belt [Music] tilts much more useful it's effortless with this lever and clamped with a single knot there's nothing on the other side everything done from here but it gets clamped on both sides I'll show you later how that works and the big dust collection part is out of the way in the back that's everything from the outside you you will never believe how I made the oscillation mechanism inside but first let me tell you the stupid reason why I built this machine not because of the mechanism a few years ago I bought this sander with way too many sanding belts now sitting in storage and I just wanted a better use for them really stupid but that's the truth now how do you approach such a project from scratch where to start well I'll try to summarize I wrote down a list of requirements with features I wanted the machine to have like the tilting belt the height preset and so on then when designing at the computer I put each feature into its own assembly so this does the illation this does the tilting you get the idea and wherever possible I tried to integrate multiple features into a single assembly sounds straightforward but all that was the hardest part I spent a lot of time designing but then you can build assembly after assembly and bring them together so let's get back in time when this didn't exist and begin with the oscillation mechanism I've got this Precision ground shaft which needs to move up and down together with the sanding drum a common solution is a geared motor with a crank but to me that's kind of boring I wanted to couple the mechanism mechanically to the spinning shaft with a gear reduction that should be interesting from my machine specs I need a gear reduction of 25 to1 and I could do that with a warm Gear Drive every turn turn the warm advances the gear by one tooth it has 25 teeth and I've I've got the right gear reduction then I can slide this into this housing add a crank pin here and there's the complete oscillation concept it's a great concept with few parts but 3D printed like this it can't survive long term this has to be metal and constantly lubricated so I would need to find just the right one off the shelf or have it cast to manufactured that's not going to happen so another mechanism with a high gear reduction is a cycloidal drive and if you ask yourself what that is take a look at my logo that's a cyal drive conceptually it's simple we have a housing with pins the shaft with an Ecentric cam the cycloidal disc and an output ring the shaft goes into the housing with the cyal disc on top and when I spin it it does this kind of motion the output ring fits on top and turns that wobbly motion again into a smooth rotation the gear reduction is the amount of teeth of the cyy disc so in this case 9 to1 this concept I'm sure I can make reliably lasting with simple off the-shelf parts and 3D prints and so now we've got a fast spinning shaft and a slow spinning gearbox but no oscillation yet on this concept the output gear is oriented 90° to the main shaft making the crank possible and I still need that for this one battle gears for example could do that but again I don't see them lasting a long time so what else could create oscillating motion let's start simple with this ramp shape if I move this back and forth you can see the intersection here is moving up and down and I could pick that up with a bearing now let's mirror that ramp and the result is one up and down motion without changing directions or rather one full oscillation but how is this helpful well when I write W it around a cylinder and make it spin I now get continuous oscillation and what actually matters this cylinder is the same orientation as my gearbox output and a single part no need for bevel gears so I think we have quite an elegant solution here can get serious making the real version getting serious means a lot of 3D CAD work I modeled the entire assembly with every component before I ordered parts or started printing but here we go there are many printed Parts a few bearings pins and screws and machine parts which I need right now so I cut off pieces from some tubing the longer bushing goes on First with a wash and screw as an end stop that I have to play with of course let's put that aside for now now the gearbox housing the pins this uses are actual steel pins that fit in these holes that I still need to re the automatic compressed air comes in real Handy here removing the chips the underside then gets a bunch of nuts pressed in as well as a few on the inside I could have also used the CNC kitchen melt in threaded inserts but what they don't have is a plastic lock ring and since I'm expecting vibrations I wanted that this bearing is a light press fit in the center and this ring goes on top clamping it down making this my fixed bearing the large bearing fits around the housing where it first slips down easily and the last a little bit is again A Light press fit and then insert all the pins next I need to make sure this F dis is sized correctly and moves freely in the housing so when I put it in place it should just drop down but it doesn't and I have to press it down it does move but it's pretty tough I kind of expected this here cuz I designed and printed this with no air gap and when that's a real mechanical part it's basically guaranteed to bind in some way I need some amount of air gap in here but just don't know how much exactly so let's find out I printed five discs with increasing play for testing starting with 0.05 and it already drops down but the movement is still kind of rough 0.1 yeah that's a bit better but there's still some places where it feels like it's binding 0.15 yeah now it's getting better 0.2 yeah that's great and 0.3 is probably no improvement just more play here you can see the play this is 0.05 and this is 0.3 0.2 seemed to be The Sweet Spot let's go with that a bearing gets pressed in followed by Steel bushings in all the hes which came in an interesting packaging and then a second one but before the disc came another ring first it has some holes to re nuts to press in a center bearing and then the output pins because this is part of the gearbox output and it sits below the CYO discs now the parts can come together on the shaft first the housing then one of the spacers and the output ring next come the Ecentric bushings that create the wobbly motion their Center hole needs to be Reed as well then I have a feature in them which positions them opposite to each other like so to fasten them on the shaft there's a pocket for a little square nut and a tiny set screw and now the disc wait oh crap I forgot to file Flats on the shaft and this assembly is not a straightforward yeah at least I didn't forget to Mark where where to put the flats now since the steel is hardened I'm using a diamond file and it bites into the material well but there has to be a more accurate way of doing this I put two square blocks at the ends and this piece which holds the file parallel and at the right height no need to worrying about steady hands but the real benefit is the second side cuz I just need to flip the shaft and the second flat will be at a perfect 180° to the first simple technique but great results the other flights are for later now I'm much more confident in the set screw and the cyal disc can go on top I was really keen on testing so let's do it this looks nice so now the second disc I need two discs cuz having only one wobbling around would vibrate so having a second one opposite to the first one the vibrations will cancel each out but I have to make sure the second disc is oriented correctly to the first one in the housing otherwise these output holes won't line up with each other since I have 12 output holds and 25 teeth this is not symmetrical and there only one correct position so I put a little Notch at one of the Holts to make this simple right here looking straight down you can now see how the hols of both discs kind of do a circular motion around where the output pins will be so let's put them in that looks promising the remaining three holes are for screws and after reing the pin holes in the lid it goes on on top now goes the actual gearbox output with matching Holts and on the outside there is the integrated ramp which I made a lot smoother and it also fits a bearing as usual by now reaming the HS and pressing in a bearing then the last spacer washer and then assembling all the parts together the screws hold everything together but I have to tighten them by feeling and this is my by feeling face without context this looks really stupid and finally a dust cover with an improvised Labyrinth seal oh yeah that's the gearbox done what a nice little compact unit basically full enclosed and protected from dust which is nice it's super smooth I can't really feel resistance other than the six fresh bearings and what you might not expect is this is back drivable which means it works when I spin it here or here that's cool just to bad you can't really see the inside now working hm let's fix that there you go go I cut it open let's see it in action so for this short sequence I had to disassemble everything again but this then at least allowed me to add grease when real assembling but still can this survive long term cuz it will run a lot well I'm confident cuz all the motion is either done by bearings metal on metal or at least well lubricated plastic on metal contact and even if that's still not enough I've got one more trick up my sleeve cuz thanks to the sponsor of this video jlc 3dp I've got all gearbox parts and some more for later printed from nylon a much tougher and more suited material than pla other manufacturers also know that for example this gear in my planer gearbox is also made from nylon hm while we're in here let's clean it and look at the print quality that's just nice much better than what I could achieve and ordering these parts from jlc is pretty simple if I print them myself I export the 3D model and import it into my slicer software for the order I just upload the same file to jlc 3dp here could then choose the 3D printing technology in my case SLS with nylon the amount and I instantly get a quote for example two Cy discs are just 650 just like my own Parts all these of course also need some reaming before assembly but then they pretty much go together the exact same way and there it works nice the heart of my machine if you need custom parts give jlc 3dp a try they printed service starts at 30 their CNC Machining service starts at $8 and if you use my link in the description you can get a $60 coupon but for now let's disassemble it again cuz until the final assembly I want to stick with my parts cuz I of course want to see if they work the next part that goes on the shaft is the pulley it's made from the pulley and the pulley support and starts out as this chunk of aluminum this first gets a facing cut and then turn to size with a few roughing and finishing passes and and a shamer it's a non-critical dimension but still nice when you hit the 45 mm exactly I used a 3D printed template for transferring the whole locations and then cut the piece off but forgot to record it could have looked like this except the diameter is too large for my tools so the last bit is for the hexa a quick Jack jaw swap to clean up the second face and then heading to the drill press for the threaded holes spot drilling Drilling jamp tampering and threading standard procedure for the set screw hole I installed two screws that position the piece correctly in the viice and with the ruler method I made sure the part is centered then I carefully spot face with an ant Mill and do the rest for now this piece is done the pulley starts as a scrap of polyethylene sheet which I placed on the CNC router to cut the screw hols and a slightly oversized circle cut out it goes in the Chuck where I Mark the rough size of a pocket which I turned to a snug fit with the ply support followed by champas and removing this material which creates very fluffy shavings then the poly support goes back in the Chuck where I fac it one more time and then screwed on the pulley with a dab of Loctite on the bolts I turned the dam to size and Sharpie it like to better see what comes next finding the edge with the tool move it to the center scratch the surface and zero the dial then plunging to a certain depth creating the V GRS for the belt and breaking the sharp edges the last feature is a precise Center hole which I drilled and then carefully turned to size it took many passes but I hit my Target and the result is this ah perfect all right next comes the upper bearing block the upper bearing sees more of the load when I'm pushing against the sanding drum so I made the bearing housing from aluminum here I marked the side centers which help me to position the piece at the correct location on the machine bed there I first machined the holds that align with my hole grid so I can use two bolts to hold it down and then I let the machine go to work the finishing T is really satisfying as well as the shamer turned out really well but right now the bearing doesn't fit because the CNC is not accurate enough for a bearing seat and I want to turn that so somehow I need to hold it on the lathe I've got some round off cuts from a laser that should work I marked the center and the whole spacing of the two hold down holds which I then quickly turned into threads with a MAG maget on each jaw the plate has just a little bit of stick out I face it by leaving some material at the center and after about 1 mm I turn it to a nice fit with the aluminum that puts it in the perfect Center and I can hold it again with bolts I left myself about 1 mm of material so I can do multiple passes and get it accurate I think that paid off the result is spot on then turning the hole to final depth at the drill press I drilled and tap the outer holds to m6 and the inner holds to M3 but since there are blind holds I finish the threads with a drill and a low clutch setting the bearing nose should be a light press fit and it's not a turns out the bearing is also slightly undersized which I should have measured before but at least I can now try out a trick my dad told me a layer of Sharpie it's a few microns thick and should solve the problem oh yeah that's a press fit now nice and the six screws hold it in place the last part that goes on the shaft is the drive block thingy that engages with the sanding drums it's straightforward and similar to the pulley support I face and turn some stock to size so it again fits the template jamp for it and cut it off this time with recording it flipped around I fa it to length and then drilled and turn the hole to the same fit as the pulley then transferring the hole positions drill and dream them to size and depth and Hammer in some pins which will engage with the sanding drums they also lined apart in the Vie for the m6 set screw hole very nice everything is mounted but it's far from being a finished sander this this needs to be mounted in something and I call that the sled the sled basically is a wooden box where all parts have different features cut into them and after a bunch of tool path programming and flattening the CNC bed I can begin the workhorses are these single flute bits in different diameters they allow me to take aggressive cuts and remove a lot of material but first I have to prepare my material some 18 mm thick birch plywood which I cut into roughly oversized pieces and then cut a shallow Groove all around for my hold down clamps this gives me access to the whole surface for Machining but mostly the stock material isn't flat to begin with so I shim the gaps to Not Bent apart with the clamps I then cut the various holes and other features into the part followed by the outer Contour to about half the depth then I switch to the flattening bit to make the Surface flat and also cut away all the B from previous operations and then the program waits for me to switch to screws to finish the outer contour and like this I cut all the parts some I also just tapped and of course I also had to crash into the shim destroying this brand new bit at least I can continue so I removed all the bigger chunks flush trimed the edges and made new grooves for the second side I then use pins to precisely position the part for the second side that works cuz I have the same whole layout in Fusion 360 so the software position and the real position match and I don't even have to probe the parts most of the second side operations are Holts shamers and big shamers and flattening to thickness basically the same for every part and there are all of them finished and none of these parts are sand it but by using the right techniques and especially aut of operations I get these clean and sharp edges everywhere on both sides let's check the fit ah beautiful before assembly I inserted a few nuts and some steel pieces that are first needed to make however straightforward they will be used later I'll do the assembly here on the table saw cuz I know this is a flat surface starting with a back and middle piece the bottom Edge needs to be flush and then I can transfer the holes all these screw joints got a clearance hole using the existing holes as a guide then a pilot hole and then a screw that way the drill bit doesn't wander in the plywood Edge then the top plate flush with this Edge the left and right side but they needed to be clamped in position lastly this bottom piece but it seems to be just a hair too [Music] wide so that's the sled skeleton almost complete and because of the way I designed this all the pieces are kind of squaring up each other making this a square block all around with it all together I could remove a few screws again and replace them with the dowels to make it a more permanent connection next came two pieces for the motor mount except they didn't require the CNC router and then they got bolted on with the nuts I previously installed then I made these angled pieces that closed the gaps between the top and back plate and glued them in place with the glue dry I flush trimmed all the Dows and then cut a large bevel which is for clearance when tilting that's the finished sled frame let's add components to it the gearbox with belt goes in here where the housing gets bolted to the bottom the bearing block to the top and the motor goes here yeah what motor I've prepared something for that I picked up this crappy little table saw for free from the garbage where it belongs I mean I'm terrified of this as a table saw but it has a nice little motor I just need to free it from all this terribly dangerous tabl saw crap and there it is freed from the pile of scrap metal nice and quiet 700 ws and 2800 RPM that's great the capacitor and wiring are Mount elsewhere so the motor really is as compact as possible something strange show is this a significant rubbing noise coming from the back so I took off the back cover and couldn't believe it there are two pieces on the fan that are intentionally rubbing on the housing why was this supposed to be a motor break for the table saw I don't know let me know what you think in the comments but this will go even spring loaded now right here on the shaft needs to be a flat spot for the pulce set screw I could hold the shaft in place with vice grips and use a similar fire Contraption as before only this time the shaft isn't hardened so I could use a normal fire the rest of the shaft is too long and I also don't need it so off it comes I then let it spin to clean the end and shamer it before I started working on the motor mounting brackets I cut them from some steel flat stock another job for the big ass file getting the length right until they fit in the sled then I transferred whole positions punched them and drilled them for the slots I drilled multiple holes that slightly overlap each other so I could fit a jigsaw blade to remove material so I could fit a dremo to remove material so I could fit a file to finish the slots not my best work but the slots work and that's what counts then I transferred hold positions again drilled pilot hols and added screws I couldn't reach the top screws not even with a driver bit in pliers so I open up these holds and used hex head screws now that's a solid mount to bolt the motor to well I've got a problem I cannot get these back screws installed there's just so little space and I knew this would be difficult I can get them in place but then just can't get the thread started and the screw jumps around and the pulley isn't even installed yet I don't want to drill excess holes in the top cuz I wanted that closed off yeah but I have an idea that might work with a threaded piece I can hold the bolts in the lath Chuck and work on the threads I had to shorten them a bit and then remove the threads of the first couple millim so now the screw can fit in the hole and stays in place making it much simpler getting the thread started then I also realized I could remove the front support for more aess nice and the threads are in all right now I know how I can mount the motor and can continue with making the pulley It's relatively simple and starts from this piece of aluminum from before first facing turning the major outer diameter and the step two size then I again marked the surface to easier locate The Edge and the starting depth and then turned the three slots just like before here I also broke the sharp edges with a file and lastly comes the hole I first removed most of the material with a drill bit then measured the actual motor shaft and turned the matching bar checking with the motor and that's good flipped around I could part off the rest or at least that was the plan luckily only aesthetic damage and I could continue everything thing left is facing to length turning the Second Step deburing the hole and making the threaded hole for the set screw for the wiring I 3D printed this custom lid it fits one of these cable clamps the capacitor has its place here with the wiring going internal through this hole and when it's assembled the hole will line up with this one that I need to open up not only that but also cutting through the top then I finished up the wiring gave it a test Spin and closed the lid that turned out nice another addition before testing the gearbox with the motor are two bearings on top of the gearbox output they are the axial support and will take load away from the other gearbox bearings but having just a plain bearings on top didn't work so well and was wearing the surface quickly so I printed these cover parts from t puu turned and shered them and press the bearings into them like Wheels that's better and back to the motor I spent some time installing it back in the sled got the belt on the pulley and pensioned it it's also important for the pulce to be aligned to each other which I could adjust and check with a Bard gauge and now it's ready for the first Power test it works woohoo what a strangely soothing and satisfying motion very nice so far now I will remove everything again cuz moving forward it's easier when there are less parts and weight for the sliding Hardware I'll be using a linear rail with two blocks cuz this set is with just about 40 bucks pretty unbeatable the blocks have two precise surfaces one where it gets mounted and one here that runs parallel to the rail and to get them aligned on my sled I machine two holds for Pins where I can slide them against mounting them to wood isn't really ideal but at least I have all the surfaces machined and here I thought about access holes that's the slide finished next it again needs to be mounted in something which I call the Carriage it basically is another wooden box around the sled where it can slide up and down in it also holds the mechanism for raising and lowering the sled and later the components for the Tilt feature just like before I prepared all parts for the CNC router and let it do some magic like this or this how about this and that or this and then all the other sides as well as some manual work these are all the parts and the assembly is quite similar to the sled all the hexagonal Holts receive some nuts two holes get threaded for a screw that will later be an end stop the smaller parts are bushing and bearing housings and then there are again alignment pins for the linear rails the assembly starts with these two alignment angles in the side piece and like before everything gets a pilot hole and a screw there are the alignment and square reference for the front and now this is nice and square next is the the back and the other side on top so I can clamp at the bottom then the bottom which also squares up the back and finally the second side which now has lots of alignment surfaces all right that's the carriage assembled and I now need to check if the sled with rails attached fits inside here and oh it fits nice just a little little bit of friction and it's not getting tighter oh yeah seeing this work is so satisfying and quite a relief cuz there's already a lot of time invested here I slit the rails against the pins and Scoot them down the two rails should now be parallel to each other but we can also check that with a dial indicator mounted on a block about 1/10th of a millimeter variation and against the face it's a little over 210 if I'd be building a CNC router for steel that would be terrible but in this application where also the frame is wood and not that stiff that's perfectly fine and knowing that I could replace some screws again with Dows holding it together for good while that's drying I can start working on the assembly that will go here that interacts with the slot in the gearbox and also allows the raising and lowering of the whole sled and when you look at the available space left this had to be very compact conceptually there needs to be this bearing mounted rigid so the gearbox can push the sled up and down from it and additionally I need to be able to move the bearing up and down the very first concept I had in mind I guess could have worked but I'm glad I scrapped it for this one there are some steel pots that are took care of right away first a pin so lath work relatively simple part first facing the end and making the M5 threaded hole with just a little bit of lubrication H then turning the outside the overall biggest diameter then the 12 mm part a bit less for the M12 threats and then the 8 mm section for the bearing then adding a leaden champer for my tail stock threading die holder I built last year what a great little tool and after that I could cut it off but something seems wrong this doesn't match the drawing uh I forgot to add the thickness of the cut off tool starting over is so much fun but now I could check at the other way to face it to fin length and cut the M8 threads in device I filed the wrench flood and this part is complete next is a plate for which I could use the offcut from before I just needed to bring it to length Mark the whole location and drill it lastly another plate that I cut and sand it to size so it fits the cutout in the carriage for the slot I used a step drill removed the Middle with the jigsaw and filed the rest perfect okay I need one more part from the CNC router to demonstrate the concept and like the gearbox output it also has a ramp this material is not wood it's called hpl and it's kind of a mixture of wood fibers and resin compressed together and with a pretty hard and durable surface it's also weather resistant and dimensionally stable but I can easily machine it with woodworking tools thus making it a very suitable machine part this gets bolted together then in the slots and the ramp on top and the bearing and now as I slide this let to right the pin slides up and down all this not just needs proper guides so it's solid and works smoothly my material Choice are sheets of polyethylene which are dimensionally stable with low friction I could use the same holdon techniques from before and also use my woodworking tools just a bit less aggressive great camera placement in general quite an easy material to work with but the chips can get pretty annoying sometimes now with this cut and still clamped I can check if the depth of this step is the same as the thickness of the slider and I can already feel and tell that's not the case 6.34 for the thickness and 5.98 for the step let's fix that ah that looks good now and with the lid on top it can still slide freely with little to no play Summer features for the other side and onto the smaller guides which are basically the same but simpler and with additional manual work and there I've got six nice parts and one glorious fail one of the guides for the metal plate was supposed to be aligned square with pins but something obviously went wrong there and this won't be helpful however the hole in the backing plate is accurate which I can use as a pivot and then align it with a square did I just say one fail I meant two that's better nice and smooth and very little play the pin got an acrylic Glide bushing which I machined to a perfect slip fit and reamed the hole the acrylic is hard has low friction and had just the right thickness then I lock nut to secure but not tighten it the other guides received some nuts before the installation what oh great the holds don't line up and a drill doesn't fit so guess who spend some quality time drilling holds by hand me having to watch this again is just as painful okay now um the next problem the slider doesn't fit at least that's an easy fix with some precise tabl so Cuts perfect then just adding the lids and the bearing wow that took quite a while but now it works smoothly the bearing feels solid with little play that's good next is the hand crank coming out the side to control the height which involves all these parts this 3D print receives some nuts and gets screwed to the the slider this little trapezoidal nut needs to be flat on one side and gets mounted to the slider as well the little bearing Flinch from before gets screwed to the outside with the threaded Rod right through the nut is just a little snu and still moves which allows me to now align it with the spindle axis and then lock it in that position next I turn a tip to one end which I could use to transfer the right hole location into the guide rail it's hard to see on camera but there needs to be a hole for a tiny brass bushing so turning some brass to a light press fit and then the hole which I started within 6 mm anend maill and then used another anend maill in a bushing in a bushing in a mount in a tool holder as a tiny boring bar and that worked pretty [Music] well the remaining wall thickness here is just 25 mm but it worked with everything back together the threaded Rod should now just fit in the bushing and it does nice this now allows me to pull the slider towards me and to push it I have this adjustment ring against the bearing with access to the screw here very nice and lastly the hand crank it's a 3D print and I'll mount it with one of these clamping rings that I drilled two threaded holes into the shaft is still too long but I'll shorten this later because now I want to see how and if this works together with the sled that's an extremely important test since the whole sled weighs 10 kg so as an afterthought and just in case I'm installing some gas Springs all right and there I should get oscillation oh yeah it works okay and now with gravity on lifting 10 kilos it's pretty smooth nice isn't this amazing that was without the gas shocks all the weight just on the gearbox so let's hook them in and try again well I can't tell much of a difference here but the hand crank oh that's now so easy to crank up that's without the gas shocks and that doesn't sound nice and now of course with the motor oh yeah nice now that's it's a huge milestone but next this once again needs to be mounted in something and I call that the Tran assembly yet another wooden box and it's what allows the tilting motion and it's also the machine base that holds everything together and will later also carry the table and dust collection the most interesting parts to start with are of course the trunions which I again make from polyethyl and sheets but thicker it wasn't flat to begin with so I shimmed and and flen it before the actual Machining but then it's not much different from other parts so far so cutting holes chamfers flatten it to thickness and mount it with screws for the Contour once done I check the TR in width and since it came out right I could already cut away the remaining stuff position it with pins to remount it and cut the second side to finish it up but eventually the camera was tired a clever detail is that the Holts I used for positioning on the CNC bed are now also used for positioning the part on the carriage everything has a purpose then just making the second one and this one has a little bit of a flexible section there's this tapered hole where for example a screw head can spread this slot open which then wedges the Tron in place the actual piece of course won't be a screw head so let's get that made a slightly more interesting lifee part starting out simple with a threaded hole then I turned down the major diameter and the smaller ends to finish dimensions and scribed where the tapered section needs to end then I set the compound to 15° and whittel Away the material up to the Scribe line going by ey is accurate enough since I added enough clearance on other parts and it seems to work so I smoothed it out a bit cut it off and went to the Disc Sander for the flat that's necessary to probably start the hole for the pin that gets hammered in place and will prevent the cab from spinning in the hole these are done so next are there matching Parts which are also the machine based sides and the most complex parts to machine that's because after Machining the first S I needed to take it back to the bench and glue on this additional part well in terms of complexity that was mostly the programming there are just lots of operations on this part that was the roughing cycle where this roughing bit made a super nice job of creating tiny shaving stad basically never clog it up all right the Machining is done and if everything went well this should now not fit in the slot and it doesn't I intentionally made this too small at first so I could now sneak up on a good fit and a little problem cuz I didn't surface at the end there's a lot of fuss going on so I run the champer bit along these edges to clean that [Music] up and that is pretty nice and there's the second one Let's test if that clamping feature actually works so now it's loose I give it a little spin and it does oh yeah that's tight I have not tested this before if that wouldn't have worked I would have been kind of screwed but I also couldn't think of a way why this shouldn't have worked but I'm still glad it does oh yeah Ah that's great you've probably also spotted these integrated gear teeth which in combination with a small gear allow me to precisely adjust the angle with tail lever and in addition to the gears I also turned some gear mounts very similar to the pulley support with a simple Reed hole three threaded holes in the face and one more for the set screw but to try all this out with the carriage I first need to assemble the rest of the base and that is the bottom plate and two support drips they again got nuts inserted and one threaded insert where I tried the drill press nice and then again it goes together like a simple puzzle the two ribs go into the bottom plate first providing alignment and square reference for the sides which can then be clamped and screwed together and there's your five-part 3D Puzzle complete only thing left now is mounting the trunions to the carriage with the gears already in place the two bushing blocks inside and temporarily securing the taper Cap all right and now for the first time I can try to fit the carriage in the base which I have to do like this oh and it fits nice it's also very smooth there's also a small air gap here and here so the carriage doesn't rub on these surfaces and only Glides on the plastic the gears are still free to move that's good and the whole carriage by design has about 1 mm side to side play for clearance and when I lock it at some angle it will always get pulled towards this side removing the play while this TR then just wedges itself secure so next I can get Gears functional and that's done with a shaft going all the way through to the other side tightening the set screws to it and remove the subtle gear backlash with the bushing blocks the handle is similar to before a clamping ring that I cut threads into and a 3D print but this time it's more of a lever type handle cuz you do less than a full rotation with it and there it works nice I wasn't sure but this feels like a good size for my hand and with the gears I can be super precise with the adjustments that's nice but how to lock the angle now well you've maybe seen that this shaft is actually a tube where I can now stick a threaded Rod all the way through the end will find the threads of the tapered cap you can't see but hear it when I'm turning left there the last thread skipped and now I can thread it in now I just had to bring the parts to the right length so all this can work I also reduced the diameter where the set screws mangle up the surface so the tiny bur doesn't protu past the main diameter with two nuts I could tighten the rod into the taper cap then the handle with a washer another washer and the knob so now I can set an angle give this a few turns and it's locked on both sides all done from here that's what I think is intuitive and easy to use later in the build I made this even better because with the weight of all components setting an angle was getting pretty hard so I installed two gas sprinks that made a huge Improvement much better of course we also can't forget about an angle gauge and a pointer that mounts to the trunion with a little stud the base as open box with the two ribs right now is pretty strong already but obviously not enough so to act on that I glue together a solid hardwood beam as a structural member which I then M square cut the width and set up the bevel Cuts this may look a bit sketchy and I wasn't 100% comfortable with the offcut sliding down the blade so my dad hit the emergency stop as soon as the cut was done then only cutting it to length which I screwed up and had to go through all this to give it another try that's more like it the piece also got some holdes for mounting the table and with some setup work for the counter boards and a counter Boo Hole in the middle for the end bolt yeah don't ask I cut a few threads with a tap and forc the bolt through the rest making it very tight now let's mount it I transferred the whole location but will drill the pilot hole with a slight offset so the screw will really pull it in place there you can see the offset with that in place I can install the 90° end stop and and a 45° end stop the bolt in the front Beam with both of them now being metal on metal contacts and there I've got a big structural piece with integrated end stop and it will also hold the table which is next and as you might guess it also has features and won't just be a simple board it's made from multiple pieces and depending on which sanding drum or belt is used has different table inserts first of all the subtable that holds everything together another CNC routed part and after the first side is done I check the group for the c channel rail which needs some [Music] adjustment difficult to see on camera but it's a better fit now then with the second side done I can align the c channel flush Mark the other end cut it off and sand it to the line would be a perfect job for an edge sander you know I also drilled two more mounting holes because only three seemed not enough the material below the c channel got quite thin so it got a reinforcement piece that also gives the SE channel screws a lot of meat to bite into the actual table surface plates are made from the same hpl sheet like the slider or gears that killed yet another bit God damn it I was too optimistic again let's better go on with some manual work like reaming two Holts and champer the edges next I needed some steel brackets but the right width was out of stock so I cut a wider piece in half I really don't want to file these to final WID so I have an idea the pieces are small enough to put them in a tool holder I just made sure the tool post is trammed in square plac two spacers under the part to raise it to the center clamped it down and just went for it I mean the L theoretically is a mill after the first pass I also added power feed the results speak for themselves not bad the longer piece I could do in two setups and I could also take much deeper Cuts 8 mm depth no problem with the same setup I could then also cut the parts to length then just marking and drilling the Holts then rounding the corners and they are done they now all fit into pockets on the underside of the table and will later hold various insert plates of the table some of them can also swivel but it seems like I forgot to cut the pocket for this one here but that's no problem thanks to the whole grid I can just remount the part and let the pocket magically appear the last parts missing are the insert plates but they are just some more CNC parts so let's skip that and go on with the assembly like before there are a bunch of places for nuts and bushings then I screw on the support beam and drill the holes for mounting the rail M5 threats right into the wood if you wonder if that's strong enough well it stalls my little drill in first gear so yes now the top plates I didn't want any screws through the top to have this nice and clean look so they're going to be screwed on From Below there I have all these holes for short M6 bolts and I'll cut the matching threads right into the tail plate with a bottoming tap nice and clean without any screw heads or holds except for these two but you'll see later what they can do next are the steel brackets the ones that can swivel have a bushing a brass washer and a spring washer so they are always tensioned nice and tight movement without play however a little bit of flex and that's what these slots are for I marked and drilled a hole in position for another screw that supports the end of the bracket and removes all the flex only problem these screw heads collide with the sled at 45° so I have to draw some clearance holes except for the inserts that's the table complete and I can mount it with five bolts three in the front beam and two in the back in some blocks that are screwed and later glued on now when I was about done with the table the second package from jlc 3dp arrived caina only sent 3D printed Parts but also CNC machined table plates from aluminum that are hardcat anodized wow everything is cut to spec the cut and surface quality is beautiful the threads are already m in place that's really nice and even the slightly complex 3D shaping that's required for when the spindle's Tilt is there very cool what an amazing upgrade except that the holes in the back are missing cuz I added them after the order so I have to transfer their position and drill them now with this tabletop I could now also finish all the inserts and make them a perfect fit and here's how they work first the insert plate gets screwed to a sub insert plate to make it longer and it also houses for leveling screws three different kinds of insert plates this one is for the biggest drum this one for all the other sizes together with smaller inserts for each size and this one for the belt for the big drum insert I need to flip out this metal bracket and this one so I have four positions for the leveling screws and the weird shape of course is for clearance when the spindle is tilted then for this insert I don't need this bracket but the other one in the corner and the individual small inserts sit on some more leveling screws just in there lastly the belt insert doesn't need any of the flipping brackets and just sits there in the corner but to prevent it from sliding around it gets a pin that fits a hole in the long bracket oh perfect now I've got all this done but still can't sand so it's finally time to make the belt tension assembly and the sanding spindles the tensioner has three rollers two of them are stationary which already give me two different diameters to sand against the third one is for belt tensioning and all controls are toolfree and easy to reach on top quite a complex assembly and the base is again some CNC cut Parts but I think we are all tired of that so let's just grab them right from the screen not the belt yeah that one very convenient and I also don't think you need to see me making simple bushings much more interesting is the main drum for which I have an offcut of a huge plastic sheet it's going to be made out of three pieces and since it's plastic that means a lot of lath fun Machining this is just easy so first I turn the middle piece to length to a roughly oversized diameter and then drill and bore a precise hole for the shaft well it sounds kind of wrong but a slip fit is what I'm looking for the outer pieces are a bit thinner so more turning fun facing is easy drilling is easy and boring a precise fit is also easy but I still made sure to remove the same amount from both sides so internal material stress is still even ly distributed and then I also shaved the ends to Dimension next I made a template for the whole pattern that once again just fits over the pieces to transfer the positions with a drill the middle piece then got the M8 threats and because it's plastic and nothing is sharp I can do that like this never attempt that with aluminum or steel the other parts got M8 through Holts with counter BS and another template for the pin Holts that match with the drive block and now I can assemble the drum on the shaft itself and screw it together M8 is definitely Overkill but I already had the bolts so I used them now I've put another shaft on the lath to finish the drum on I'll use one pin resting against the jaw to transfer power and with the help of a few bellw washers the tail stock will support the shaft Center and press the drum against the Chuck then I could finish turn the biggest outer diameter and turn the channel in between here the chips can be really annoying but at least removing them is quick now just touching up this side and it's done now I remove the outer piece again to add the grippy material a simple solution would have been heat rink tubing but it being relatively thin I didn't have too much trust in it so I printed these thicker rings from flexible material that gets stretched over the cylinder which then gets bolted back together and I can turn turn the flexible rings to final diameter really funny shavings could also be my back hair wait what my target was 75.6 close enough it's not as grippy as the one on this sander but since it's larger and also used as a backing drum to sand against this should be just fine the other rollers are making from 30 mm aluminum tubing with 3 mm wall thickness they both get bearings so back to the late it should have been straightforward but after checking them for run out I decided to take a light pass from both sides to fix that and the arrow I get from flipping in a Chuck is less than the run out before then turning the bore this time to an actual press fit with the bearings now the bearings the part on Top make sure I only hit the autter race so hammering is fine with the bearings in place I took the tension roller back to the lathe to turn a little bit of a crown to it which should help with belt tracking he's going to regret that the matching shafts are simple with one threaded end that's the roller done so up next is the belt tensioning mechanism which involves quite a few parts but most of them are simple and I won't show them in detail we have metal plates with holds aluminum plates and angled with holds a shaft with a cross hole and one that's a bit more complicated turning the ends is simple but the difficult part here now is to cut the flat along the whole length I again just try to fit this in a tool holder and as luck would have it this is such a perfect fit that the screws are slightly off center and when I clamp it it will clamp it down and back then I slit the tool post forward set the height and trimmed it in square you can see about as much as me but it worked without any problems my target Dimension at the flat was 9 mm looks pretty good the surface finish is also nice and the clamping marks don't matter then just the threaded holds and it's done so up to the real challenging part this cam disc with an eccentric and a tangential hole facing turning and parting to size is the easy part the tricky part now is to get this this hole 2 mm from the center I could just mark it and go to the droal Press but I want it a little more accurate I only have a three jaw Chuck at the lath which doesn't allow me to put it off center but when I shim two of the jaws with a 3 mm spacer like so this will put the center axis of the Chuck exactly 2 mm of Center for my [Music] part perfect and the next tricky part is to get a cross hole for a clamping screw in there eyeballing the position is good enough so back in the tool holder I line it up with the center lock everything down and give it a spin I'd call this success while set up I also drilled the remaining hols and cut the threads by hand only for the slot I don't have an elegant solution so hecka it is but with a little deuring this turned out great and it also so fits holds and the screw head is below the surface and with that I basically have everything ready for assembly except that I had to do kind of an emergency design change let me show you so far the big drum would just mount to the main shaft and the rest to the sled and when I tension the belt I would just pull against the main shaft thinking the deflection would be minimal but when I tested that I noticed the deflection is quite a lot that's a huge problem the big drum needs to be supported within the Bell tensioner and not pull on the main shaft so it also needs some bearings and I need to remake these ones and also change the drum slightly to fit the bearing now assembly time starting with these pieces so the glue can dry the bottom plate get some partial threads so it's easier to start the threaded inserts then pressing in the bearing pushing and a square nut with set screw that's important later with the glue dry I can push it into the next part screw it together and then slide it onto the bottom plate where it gets secured with more screws the top plate helps me to align this next piece which gets screwed to the rest and now I can add the tensioner a sanding belt tensioner should always be spring-loaded cuz it's impossible that two different belts are exactly the same length so by using a spring you prevent under or over tensioning it like it's done here there's the spring and I use this steel plate as my spring it's simpler to explain when it's assembled so this shaft gets screwed to the plate as well as the aluminum angle the mounts for the roller were supposed to be a 3D print but the ones from jlc from aluminum are just way nicer to remove the play I left between the PS I used little orings and this gets screwed to the other aluminum plate they squish together orings now put a constant axial load on the bearings but that's so tiny it doesn't matter besides applying tension this roller is also for adjusting the belt backing so I need to be able to adjust its tilt and that's what this small piece between the screws and the weird notch on top are for this little piec here acts as a hinge I printed it from hard rubber at a slight angle as you can see here and it's also a little bit springy the belt tension will apply force in the middle which will tilt the roller even more so on top now I have a bolt that I turn a tip on and that interacts with the weird shape of the plate when I screw this in the roller tilts the other way that's the tension roller assembly complete I really like it cuz nothing is rattling around everything is nice and solid last part is a spring and then this goes in a bushing now the second shaft with the cam dis attached fits behind the tensioner and moves it well at least when the top plate is in place so the remaining parts are The Idler and of course the main roller very tight tolerances in here but clearance is clearance let's see if it all fits the rollers turn freely and now when I turn here the tension roller comes out through here goes a pin and then a 3D printed handle on top nice this is the end stop screw for the handle and now I need to adjust the cam dis so the tension position is here and the untensioned position is here I left this exess hole in the front plate to reach the cam dis and then it was a little bit of TR error until I got it right like this and now you can also see the springiness of the tensioner although it's quite stiff at the moment I maybe have to cut slots in the steel plate to make it a little bit more springy but now let's try tensioning a belt and that's not tight but that's okay cuz this is just a quick release for now adjusting attention I have these two set screws they push against the bushing on top and botom bottom and by pushing it in this brings to Roller out and thus tensions the belt so with the tension set from now on I only need the quick release to change a [Applause] belt however this one seems to be a bit shorter so it must get tighter oh yeah I definitely have to weaken the spring plate a little but it works and I can already drive it around but now I'm more serious problem because there is twist in the assembly and I canot adjust that I have to remake the front plate and probably the bottom plate as well ideally of course I wouldn't use wood for especially this but I cannot machine these parts from aluminum now let's hope it goes well the second try the new plate is installed and that looks a lot better I am happy with that so in goes the last piece on the back screwed in place and this is ready to be mounted on the sled however when reassembling the components I noticed an issue that needs to be fixed these guides must have shifted under the load cuz there's now a lot of play and I need to secure them better than just the screws so I drilled for a second pin at the left side and made a spacer of exact width for the other side that should stay in place now and I can continue the assembly sanity check it still works and now I can slide this on top but right now it can still swing around freely but on the sled there's this hole for a pin and on the bottom of the tensioner are these two steel brackets as a guide okay that's now secured but it's not fully positioned yet the main drum stops in the drive block but the rest of the frame can still move up and down a bit and that's what these three Holts and the threaded inserts are for the Holts get threads in a bolt to provide a metal contact point and the thread inserts get leveling screws that I adjusted to roughly the right length Okay now everything is nice and positioned but when the table is on I have to adjust these bolts again so when I tighten it down it doesn't get crooked or tries to bend the shaft tightening it down has to be quick and easy so obviously it's a cam lever and the idea is to have a threaded rod with a teut at the end that is pinned in place and the other end is locked inside the cam lever I just needed to add a pin in the cam lever to lock against then I turn it to the right length such that it would also end up at the right orientation and now to get this tight the cam lever has its own adjusting screw and I just need to set that up once and then it works so from now on the installation process is to just slide it on make this tight add the knob and done quick easy and perfectly solid and there we go nice let's see if it can also track a belt uh-oh that's going down let's try again it's going down again so the belts always wandering down and none of the adjustments seem to have an effect I also tried different belts with the same result that's a problem and I think it could be the crown I put on the tensioning drum cuz that significantly narrowed down the section that actually applies tension in the middle it's rather tight and on the edges not really and that's bad because the whole idea of adjusting the tracking by tilting the wheel is to change the tension towards the edges and that's no longer possible with my crown drum told you so so I made a new one but now I need the bearings back H there we go and this little piece made sure I only hit the outer bearing race and don't damage it all right back together let's see if that worked still moving down still moving down still moving down a little better [Applause] ah okay mostly successful and a couple more problems so the tracking now works with different belts so the problem was the roller but adjusting it is incredibly unresponsive but I can try a few more fixes there the other problem is with the main drum and the lower bearing cuz when I last minute changed this to bearings the drum was already made and the slower bit is now pretty short and like a tenth too small small for the bearing and I tried making it a better fit with this piece of tape and hopes and dreams but it obviously didn't work rebuilding the whole drum would at least take half a day so my fix was to turn down the end more then I made an aluminum ring that got pressed on and this can be turned to a proper fit that's better I then also had to modify the drive block but that's no problem and to improve the tracking adjustment I put in a block here that puts that at a steeper angle and instead of hard rubber I this time used solid plastic cuz this had probably too much GI so let's see if all that helps okay the belt stays in place now let's move on the last part missing here is the sanding backing plate but that's just another steel plate cut to size and some holds nothing special but these edges also get a shamer if only I already had a finished oscillating tilting belt sander but I don't SO grinder it is the screws are just a hair below the surface and checking the straightness across and along the plate seems pretty good as well and the drums are just a hair below the surface that means a belt should contact the plate basically everywhere and that looks good okay and with that done I can finally put the rest together and try it out and here's what not to do well it works and Sands straight lines and curves pretty well and square already I hope you didn't expect anything else cuz it's an almost finished sander I still have to make the sanding drums but before that look at the mess on the table and better believe I also have a dust collection PL so let's get that made next the main part is this 3D print that fits a 100 mm hose so I can hook it up to my big dust collection and the other end is close to the sanding action it's a pretty complex shape but that's exactly what 3D printers are good at I printed this in pla and then jlc sent me this one from nylon it mounts to the back plate with some screws and the back plate gets mounted to some strips inside the machine base so let's see how well this works now that dust collection is actually good the machine inside now stays clean but you maybe could see some dust gets launched past the main drum or is left over near it so there should be something that like guides the dust around the main drum into the dust trout and that's what these parts 3D prints and the two holes in the backa for this big part is obvious and you can already see its effect when I just slide it in place but this again has to be easy to install otherwise I wouldn't use it so ideally a snap and place fit and that's what all the other parts are for one custom metal part a relatively simple shaft with this kind of shape yeah it needed to be like that the snap and place feature works with this 3D print two steel bolts sprinks and set screws and it mounts inside the table plate the rest is for another useful feature but just watch for yourself that works well but if I want to sand a longer piece along its whole Edge I don't have to to remove it I can just flip it into another position and it clears the path when tilting the belt it of course can stay in place and when I need to sand against the drum I can just pull it out very happy with this design and that wraps up the belt tensioner the sanding spindles look like this and were complicated in their own way cuz they needed to be a precise fit to slide on the sleeve without slop and clamp it when tiing the knob like this I tried the method from John heist with a foam disc that gets squished and expands but all foam material I tried didn't behave that way so I came up with this the drum has three pieces held together by two rubber sleeves and when you tighten the knob the rubber sleeves slightly bulge out and clamp the sanding sleeve pretty cool that this concept works and here's how I made it I marked and cut circles from the plastic offcut brought them to the lath to face and turn them round flipped around I faced them to thickness turned a shoulder for the rubber sleeve and a Precision bar in the center then the six pin holds with another template before assembling the stack back at the lathe I put it on the shaft and this time I didn't clamp it but only prevent it from moving then turning everything to final diameter making a leaden champer for installing the sanding sleeves and finishing the shoulder at the bottom to install it I just slide it on the shaft and tighten the knob for changing the sleeve to another grit I just loosen it slightly pull it off put the other one on and tighten it and to swap out the whole drum just swap it out I don't know how useful tilting the spindles will be compared to tilting the belt but there are no limitations except for the smallest spindle cuz weirdly the commonly available sleeve sizes are imperial so this halfin sleeve doesn't fit here I tried printing an ultra thin drum but it snapped in half so I just can't use that one last but not least is wiring the machine up and that's simple cuz it's only this switch Supply goes in motor goes out the switch magnet coil too and Earth gets bridged Cable Management as well to keep it organized while it's moving and now all parts are done and ready for finish and paint so I disassembled everything again down to the well wooden boxes and taped all important surfaces so they stay untouched for the reassembly all all nonvisible components and inside surfaces I finished with a clear lacquer the table plate got a dark gray that I already had and all visible components I painted with this bluish hammered finish it makes for a really cool look and a slight texture probably not everyone's favorite but this machine will be in my workshop and I like it a lot everything's ready for the final assembly but so much happened let's summarize we started by discussing oscillation Concepts and for some reason chose the most complicated over engineered one and got that working then we build a box around it with way too little motor mounting space then we build a box around that so we can slide continued with a need lifting mechanism and a great drilling exercise next we built the last wooden box to store the other boxes inside and kept it off with a table to hide them and finally we learned that tensioning and tracking a sanding belt reliably still moving down takes a good amount of determination so now for the final assembly I will also take apart my own gearbox parts and replace them with the ones from jlc once again the heart of my machine comes to life and if there's ever been one place for my logo it's here the machine is complete [Music] stop stop that music what a project over 200 custom parts a bunch of drawings this pile of failed Parts three broken bits and if you think my 3D prints worked on the first or second try absolutely not the machine has one more feature I bet you didn't notice but before that I want to thank you for watching this project has been in my head for over 3 years I already did the cyal research back then and already built an early working prototype finally I could make it happen it's out of my brain I put a lot of brain into this but it's a one-off it's working and it's really cool but it's not cost effective I won't build and sell this machine which makes projects like this very dependent on your help in this scale of project and video it takes a huge amount of time and money but I don't want to make you feel guilty for anything if you enjoyed the video I'm happy but if you can and want to support me with projects like this check out my patreon you get adree versions of the videos and additional content or support me through my online shop check the video description for that please give me also feedback on a video was it too long too detailed I wanted it to be a complete package the package just got big so for the last feature of this machine the front and back are flush with the tabletop which isn't immediately a feature but let's just say this whole project was part one of an even bigger idea one more announcement if you are even more interested in Crazy projects like this or have built something awesome yourself you should consider going to open source this year June 15th and 16th in San Francisco it's the science and maker event where you can meet the creators just take a look at who's coming this year it's going to be wild you can meet me in person we can talk shop chat about projects and I'll bring the open version of the gearbox so you can take a look at it and if you apply to exhibit you get free entry and access to the Friday night party see you in San Francisco

Channel: Marius Hornberger

Views: 1,272,178

Keywords: woodworking, workshop, DIY, homemade, jig, stuff, machinery, making, bandsaw, tools, interesting, smart, great, Marius, Hornberger, Engineer, woodworking tools, edge sander, belt sander, overengineering, overengineering at its best, Oscillating edge sander, engineering, oscillating belt sander, spindle sander, marius Hornberger, oscillating spindle sander, sanding drum, shopmade tools, homemade tool

Id: AHUUXfB0qHY

Length: 83min 24sec (5004 seconds)

Published: Fri May 03 2024