Too long without an update of any kind, so I'm going to just try to get what I've been working on written out here for now and maybe video later. (Writing this will help me format my thoughts for any video anyway)
My last update I showed off a very curvy forearm design. I decided that would be too hard to produce and maintain, so I went back to the drawing board to make to something more symmetrical. Here's that design:
As you can imagine, I was pretty excited about this! I had a lot of the motion figured out, from the rockets movement to the arm opening up (I'd be able to make a shield like the Mk 3 with it, as well as getting in and out)...
... but there was a catch. A few catches actually.
- I didn't (and probably still don't) have proper modeling techniques for Fusion 360, and despite my best efforts to parametrically design everything there was a lot that would just up and break when I changed a dimension.
- I had designed some things (the mini rockets) incredibly poorly, such that there was no way they were going to be able to be built or work.
- I was running out of space in my design. I'd thought out all the various parts, just not how they'd intersect and interact with each other.
- From my white prototypes you might remember that the breakthrough for the latch was having it hinge directly on the shell. When it came time to model that, I realized that it was going to basically cut off the screw I was using to mount the rocket.
- The mounting holes for the elbow cut into the space I needed for the latches
- There wasn't any space left for new bolts to connect the shell to the frame
- Probably more I'm forgetting about or didn't get to yet
- The whole thing was feeling a bit chunky (37mm from my arm to the top), and there was no way the bulge at the top for the main rocket was going to be hidden in a Mk 4/6 style arm.
- ...I designed it too big to be printed on my 3D printer...
- Remember the old a^2 + b^2 = c^2 ? Well, to maximize the size of parts I could print, I was printing diagonally. I just forgot that if you make the diameter of a circle equal to the diagonal of your build plate then you don't have room to print the circle.
It's fine that this design didn't work out, it's a learning experience. What I'm more upset with in hindsight is that I was so worried about creating a good YouTube video that I didn't stop to document this all at the time. I've given up on enough projects to realize that I might not ever finish a full suit, but I at least want to document what I try, what works, and what doesn't so anyone coming along can learn from it and do better than me.
Sometime in January or February, before I realized the updated design wasn't going to work, I decided to do a basic electronics test on my old cylinder frame. It wasn't anything special, an Arduino, a 16 channel PWM controller, and the servos I already had mounted in that frame. The big thing I learned was that you need to use thin solder if you want to solder small things. No brainer in hindsight, but was having a lot of trouble connecting the header pins on my Arduino before that.
Mid March life got turned upside down, and what use to be normal became impossible... thanks to the addition of a brand new baby in my family!!!
The combination of an infant, a toddler, and working from home has messed with my schedule for working on my suit, since most of my design work up to now has been done while I've been commuting on the bus. No more bus trips = no more dedicated Iron Man time. I've been having a hard time making a regular schedule to replace that time.
Definitely better off than we could be, it's just I'm not one of those people who found themselves with extra free time as a result of Covid-19.
On April 14th I decide start a new file and to take a different approach to this forearm. Being rather dismayed with the prospect of making my previous design look beautiful after the fact, I decided that rather than build from the inside out (like I'd done for everything else) I'd take the opposite approach. So I started with a more elegant/beautiful outer shell and build the inside to fit that.
To help make everything thinner, I decided to purchase some 3.7g servos and see how they compare to the 9g servos I was using. I also picked up a 5 pack of "Longruner Mini Nano" Arduinos (going to need at least one in each forearm, one for the helmet, possibly one to coordinate the other three), and some servo cable extenders (I knew before my basic electronics test that I would want some). I also ordered myself a birthday present - a new 3D printer, the Artillery Sidewinder X1. I got it on sale at Gearbest for $380, and I am loving it so far.
I could literally print a 1:1 copy of my old printer in one go. The print volume is 300x300x400mm, compared to the 120x120x120mm I had before. It's also quieter, heats up faster, has filament run out detection, has assisted bed leveling, is a direct drive extruder... The only drawback is how much space it takes up, lol. I still need to fine-tune it (assisted bed leveling is awesome, but it's still hard to get a good level, horizontal expansion is close but not quite right, and I haven't even touched most of the other settings), but it's nice to finally have a machine that will be able to print as large as I want.
About that redesign? It was rendering pretty nice! I had a nice curve going, I had the shell connecting to the inner frame, and I had the hinge working.
Emphasis on "had". I was so excited to print something with the printer that I printed the frame, and THEN I noticed it was huge near my elbow! I re-measured everything. That part of my arm is maybe 85mm in diameter, the earlier scrapped version I'd given myself a bit of wiggle room and made it 100mm, and when I'd started the newest version I'd given myself some wiggle room again and made it 110mm. Woops! Resizing broke the functionality of the hinges, and definitely flattened the shape a lot. I also realized that because my hinge opens at an angle there was going to be some rubbing against the bracer part. Not quite sure how bad it would have been, whether it would just help hold the arm closed or if it would prevent it from opening, but it is something I'm going to have to be aware of in the future.
During the trial-and-error for this hinge design I learned something
super important. When the hinge in the pin-slot is moving straight towards the hinge pin the two sides move straight away from each other, with no rotation! It's as the pin-slot deviates from that where you get rotation. This means that the best way to design these kinds of hinges is to actually design the holes first, and use those wholes to make the linkages. Also, Fusion 360 can't really handle an arbitrary pin-slot shape, so it's likely that future refinements will have to be tested by printing them out. Having this understanding of what to change to effect which property of the hinge is a nice step in the right direction though.
While this red design is certainly more pleasing to look at than the old design I scrapped, I think I'm going to have to put it on ice, possibly scrap it all together. I can't proceed without knowing if the 3.7g servos can lift the rockets like I want (because the servos determine how much interior space I need) and I really like the look of the Mk 6 forearms.
I did have some minor "Eureka!" moments. One I might apply to the next iteration, and it's an idea that would allow more people to use what I design. That would be to make the hinge it's own separate file, make the shell it's own separate file, and then join the two parts together with something like MeshMixer. If someone has a bigger arm, they can scale the shell while keeping the hinge at the proper dimensions. If they want to use a different shell, they could use the same technique to add the hinge to it. I would probably do the same things with the rockets too, although it gets harder to do that with the flaps that cover them. The other idea was to recycle old Micro USB cables and use them to connect my Arduinos to anything that uses I2C (such as the servo controller). The idea is a natural fit because both my electronics and USB are designed for 5v DC, and both use 4 wires (+5v, ground, data 1, data 2), USB cables are already compact and you can get prettier variations (like braided), and I already have them, so I can quickly use them.
Anyway, that's my update. I'm currently at "man who has everything (hardware) and nothing (design)" stage, so when I can remedy the design part I'll be able to get cracking on it.
