100% lol
3 Likes
Turns out it wasn’t an easy fix.
I noticed the motor mounts appeared bent away from each other and figured it was down to the m5 crossbar being in the wrong position.
After disassembling and reassembling the whole assembly 3-4 times, tightening different things in different orders, whacking bits with a mallet, and covering my floor in tyre marks right before a property viewing, I realised it was down to the fact that the mounts can’t get close together enough due to the shape of the hangar. Half of the reason I put so much time and effort into designing the motor mounts was to avoid having to file a flat nearer the centre of the hanger on each end but because I had miscalculated the thickness of the mounts, it had to be done.
So here we are now, and they fit perfectly.
I took the time as well to drill out the truck mounting holes on the deck. Where I had to drill extra holes to make the Apex load spreader plates fit, I drilled them at a slight angle. This meant my plates were starting to delaminate where they were forced in place, the riser pad is cracked and worn, and damage is being done to the baseplate. I’ll admit I’ve been a bit heavy handed with it out of laziness, so I fixed it by making them 6mm to allow some wiggle room, and everything fits perfectly snug.
2 Likes
Crimped a JST-PH2.0mm onto the receiver, hooked an inline fuse up to the charging circuit, and couldn’t help but take her for a spin!
Please excuse the pyjama bottoms and sliders…
Need to fine tune the throttle control but I’m just glad to finally be riding 
2 Likes
I have a plan now for the enclosure but would like someone with experience to sanity check it for me before I go ahead.
Firstly I will use photogrammetry software to create an accurate 3D mesh of the deck. Ive needed to upgrade my AMD Sapphire RX 580 GPU for a long time, so needing an NVIDIA GPU with Cuda support gives me the perfect excuse for that (assuming compatibility, been a long time since I was into PC building). Given the semi-reflective nature of the finish, I’m going to first coat it in blue painters tape and draw in some witness marks to hopefully help the algorithms do their thing.
From there, bring it into some sort of mesh refinement software, trim the excess, remove non-manifolds, reduce the facet count etc. until I have something I can work with in Fusion 360.
Set my datum planes and design an enclosure that can be 3D printed in multiple parts and joined together. Need to figure out what material to use for maximum adhesion to the epoxy. I’ll leave out any holes and cutouts to be done by hand at a later point.
Roughen 'er up with some sandpaper, give her a clean with some IPA then a coat of laminating epoxy. Fibreglass until satisfied with the thickness then sand, prime, paint.
Drill out holes and dremel out cutouts. Fit rubber grommets where possible on circular holes (e.g. phase wire cables) and use epoxy sealant - something like JBWeld I assume - to panel mount and seal XT connectors.
Rubber tape between the enclosure flange and the deck for waterproofing. Drill out mounting holes in enclosure.
Temporarily mount enclosure to deck with some tape and mark out mounting hole positions. Using one of those portable drill guides/jigs to ensure the hole is straight, and a depth stop on the bit (or just tape…) drill out threaded insert holes. Apply some epoxy to the insert then fix to deck.
The battery will be deck mounted with heavy duty velcro. The BMS will be jankily taped down somewhere or held in position with thoughts and prayers. The ESC, which is a MakerX DV6 Pro, has mounting holes on the back so I’m going to ignore them and probably just velcro that too. I’ll use some thick foam to provide some protection between the enclosure and the electronics (creating somewhat of a seal between the bottom of the battery and the enclosure to help retain its position) but not enough to prevent passive cooling.
On the topic of cooling, I’m a little concerned about things overheating but equally, I don’t intend to put this board through hell - at least to begin with. If it turns into a problem I’ll make some cutouts in the enclosure and fit some fat heatsinks.
Thoughts?
4 Likes
What software do you use ?
I’ve been working on photogramtery as a researcher for a bit, in my experience the more important factors to get an accurate model are the choice of the pictures you feed the algorithm and the consistency of your camera.
For the pictures I’d go with a rectangular pattern (pictures parallel to the object sides, always try to get 1/3 of each preceding picture in the next one for horizontal array, and 1/4 for vertical ) then a circular pattern all around the object.
For the consistency, if you’re using a phone or a smart camera, be sure to deactivate all stabilization, autofocus and auto exposition before you start shooting.
Then you can use CloudCompare to refine the point cloud.
Never used it on such a scale, always worked with buildings or large monument, curious to see if you’ll get something precise enough for your use-case
3 Likes
Yeah you just want to make sure that the focus and exposure doesn’t change from image to image because that can cause issues when your pc is referencing points between images. The painters tape is also a good idea because in my experience reflections always cause artifacts and an uneven surface on the model where it should be flat. Little markers are mainly helpful for big flat surfaces where there is not a lot of detail to reference from, so I would put a bunch on say the top surface of the deck. You will probably need to go back and clean up the geometry after anyway but this stuff helps with getting a cleaner result to start with.
2 Likes
I played around with meshroom the other day and only took about 20 pictures. All it returned was a clump of triangles, so thats when I figured I should do some actual research and realised I needed a specific type of GPU and that my sample count was nowhere near sufficient.
My work is looking at some quite advanced 3D scanning software/hardware solutions and I know a guy who may be able to let me play with them, so I’m considering that instead…
Thank you for the advice with the camera settings, it makes sense to keep everything as consistent as possible.
the computing power will do for speed, not for accuracy, the accuracy is entirely the quality of the pictures,
for a medium object with 20-40 pics you should be good
you should give a try to autodesk recap, esay for beginers and gives good results, don’t remembre how their subscription/demo policy works though
lidar will give you good results !
1 Like
I have the educational license for fusion 360 via my Uni so autodesk recap is a great shout. Ill keep you posted
1 Like
I was doing the dishes yesterday and had a bit of a eureka moment. What if I could use a baking tray as an enclosure? Obviously the receiver would need to sit outside but…could it be feasible?
So like any normal, put together, sane human being a put the tray to my deck and realised that the depth and width were perfect:
Maybe I could buy a couple of new ones, cut them up, join them with a bracket and some epoxy, give it a splash of paint and call it an enclosure?
That then led me down a train of thought about using sheet metal instead. I’ve heard that sheet metal can be unforgiving at times, so decided against that.
But then, how about wood? The few wood enclosures I’ve seen so far (looking at you, Hoyt) are sexy, clean and elegant but must be an absolute BALLACHE to build.
Then another eureka moment struck. I must be on the ball today. How about using a skateboard deck?
The length is perfect and it already bends in the right places. Would just need to fill the holes with dowels (the easy part) and make some side panels (the hard part)
Anyway, just a couple of thoughts. The enclosure is a massive sticking point for me right now because money is tight this month and I’m feeling impulsive. If anyone has any suggestions for filling the sidewalls of a wood enclosure I would love to hear them.
Tbh, I think 3d modeling, printing and then laying up glass was a much better approach than this. You could even skip the 3d modeling and just use styro or even wood for your mould.
This problem you’re running into rn is why so many choose a tried-and-true platform to build off of. I made this error in judgement w my first build and it made finishing it exponentially more difficult.
This thread might provide some inspiration.
3 Likes
I reckon you’re right…it would be less work too. My priority now is just to get something I can ride safely, I’m itching to take her for a proper skate.
As much as I’d like to go the photogrammetry & 3D printing route, I think there’s a time and a place, and realistically this enclosure isn’t complex enough to warrant that. It will be time consuming and unnecessary. Maybe in the future though.
I really underestimated how difficult it would be to find an enclosure and in future builds it’ll probably be the first thing I decide on. You’d think given they can be made cheaply with a vacuum former and ABS sheets, Chinese manufacturers would be swooping in with a million different designs on mass production but I guess the DIY esk8 market is too small.
Cheers for the link, had a read through and that was really informative. I appreciate all of your help so far
I don’t know if you’ve seen this but Kami Juins has a video where they used two skateboards sandwiched together as a deck enclosure combo.
1 Like
I remembre seing a couple realy great builds a good few years back made this way,
@DougM is I think one of them, and there was another, a green board I realy liked with one sheet of metal as lid over 3D printed sizes
There was the accoustic board made this way, but I don’t remembre from who it was
Trust me or not, but even this one I have seen XD there was that guy with the snowboards deck
This actualy looks good, again 3D printed parts for the side are the easy way to go
The lower budget one, without a 3D printer would be cutting plywood layers, the same way you make a topography model
that would be a royal PITA though
That’s awesome, I love Kami Juins builds. As cool as it would be, I dont think I’ll have the space with my wide flat battery to do that 
Yeah I think ill stick to the original plan. Sheet metal would be nice but a lot more complicated to design I reckon. 3D printing for the sides gives me the same problem as 3D printing the enclosure, I won’t be able to accurately replicate the curvature without tons of trial and error.
Most of the time you don’t need to, just make it flat and add a moderatly flexible gasket
Plus your deck seems pretty flat
I’m more talking about the curvature of the deck that is acting as the enclosure
1 Like
I meant metal sheet + 3d printed side
2 Likes
Ah oka, that’s actually really nice. Will habe a think.
Wont be working on Nagini until a couple of weeks time probably.