I suspect more of the reason those shafts snapped was either the threads giving a stress riser on the inside or the joint between the shaft and the plate not having a big enough radii on it. A solid shaft is still stronger than a hollow one but I’d imagine it would bend if it was a strength issue with it being hollow Still a very cool build though! Definitely a good move to go to the shoulder bolts regardless, being able to replace them will increase the length of time you can keep running those end caps
Oh I think it was a stress riser from the outside, the part has a step in it to stop the bearing from rubbing against the plate and that step has no radii, as I wanted the bearing to seat against it at a predictable location, this is exactly where it snapped I think the shoulder bolts are going to be waaay better too
Good to see the melty again. Hopefully those upgrades will hold firm. Regarding those pulleys for the weapon motors, you mentioned removing teeth to make the belts fit better. Please could you explain how you do that, and which material do you use, or would rather use, to print them? I've just printed a prototype GT2 pulley for Unmaker, and I'm wondering if this is a good enough profile for this purpose.
I printed the new pullies in TPU, which is the same material I have printed all weapon pullies in for awhile now I use a pulley generator i got from thingiverse awhile back and open it in OpenScad to make the modifications Once i get an STL of the tooth profile I open it in tinkercad to add mounting holes and belt retention features
The pulsation in the weapon speeds is a bit weird, might just be because it was sat on the tape stacks rather than on the wheels, but should also be solved by balancing.
The reason it's spinning like that is that you forgot the squeeky hammers... it's throwing everything off balance... 😂. I'll wait patiently for the electronics video. I would love to see that all spread out across a table instead of the pile of spaghetti shoved into a tiny box.
The electronics here havent changed for awhile, and while I'm updating the brain the escs and things wont change much I'll look at adding a bit more detail to the explanation of it all in that video
Very exciting progress, this bot is getting more legit with every video. Are you going to mill out a circular indent in the top plate so you can both fit a googly eye and not compromise the inverted performance? With the belts being extremely tight but I think they stretch a bit especially when doing hard spinups. I think you're better off having a belt that's a bit tight to put on but settles to the correct length than having a correct length belt that settles into flop and falling off the pulleys mid fight. You can "encourage" the belts to settle to the correct length faster by adding a bit of heat.
OH!!! I knew I forgot something! the googly eye hole! gotta add that back in I think you are right about belts, but they where so tight I'm not sure heat would have fixed it. I removed 2 teeth from the pullies and the belt is snug, I'm thinking I'll add 1 tooth back so the belt is tighter while new, but snug once settled
I HATE melty brains as a rule. I think 9 out of 10 times they should get 0 (out of 6) for control and 0 (out of 5) for aggression if they ever go to a judges decision. That's being said, I was very impressed with your showing at NHRL. Your bot looked so planted to the floor and really appeared to be a class act. It almost made me change my mind about melty brains going to judges decisions.
The problem with melty brains is that when they first appeared, the electronics available where nowhere near fast enough to control them, leading to very sluggish translation and performance. the "sit and spin" style melty that waits for its opponent to come to it But now, with teensy's, brushless motors and esc's with dshot (high speed comms) its possible to build a melty that translates a lot better. But we are in the early stages of this, theres only a few melty builders out there right now so progress is slow, but it is progress.
I suspect more of the reason those shafts snapped was either the threads giving a stress riser on the inside or the joint between the shaft and the plate not having a big enough radii on it. A solid shaft is still stronger than a hollow one but I’d imagine it would bend if it was a strength issue with it being hollow
Still a very cool build though! Definitely a good move to go to the shoulder bolts regardless, being able to replace them will increase the length of time you can keep running those end caps
Oh I think it was a stress riser from the outside, the part has a step in it to stop the bearing from rubbing against the plate and that step has no radii, as I wanted the bearing to seat against it at a predictable location, this is exactly where it snapped
I think the shoulder bolts are going to be waaay better too
Duudee LOVE YOUR VIDS!!! Way to go!
Glad you like them!
Good to see the melty again. Hopefully those upgrades will hold firm.
Regarding those pulleys for the weapon motors, you mentioned removing teeth to make the belts fit better. Please could you explain how you do that, and which material do you use, or would rather use, to print them? I've just printed a prototype GT2 pulley for Unmaker, and I'm wondering if this is a good enough profile for this purpose.
I printed the new pullies in TPU, which is the same material I have printed all weapon pullies in for awhile now
I use a pulley generator i got from thingiverse awhile back and open it in OpenScad to make the modifications
Once i get an STL of the tooth profile I open it in tinkercad to add mounting holes and belt retention features
The pulsation in the weapon speeds is a bit weird, might just be because it was sat on the tape stacks rather than on the wheels, but should also be solved by balancing.
Nice video again :) i'm loving seeing this project evolve over time.
The reason it's spinning like that is that you forgot the squeeky hammers... it's throwing everything off balance... 😂. I'll wait patiently for the electronics video. I would love to see that all spread out across a table instead of the pile of spaghetti shoved into a tiny box.
The electronics here havent changed for awhile, and while I'm updating the brain the escs and things wont change much
I'll look at adding a bit more detail to the explanation of it all in that video
Very exciting progress, this bot is getting more legit with every video.
Are you going to mill out a circular indent in the top plate so you can both fit a googly eye and not compromise the inverted performance?
With the belts being extremely tight but I think they stretch a bit especially when doing hard spinups. I think you're better off having a belt that's a bit tight to put on but settles to the correct length than having a correct length belt that settles into flop and falling off the pulleys mid fight. You can "encourage" the belts to settle to the correct length faster by adding a bit of heat.
OH!!! I knew I forgot something! the googly eye hole! gotta add that back in
I think you are right about belts, but they where so tight I'm not sure heat would have fixed it. I removed 2 teeth from the pullies and the belt is snug, I'm thinking I'll add 1 tooth back so the belt is tighter while new, but snug once settled
I HATE melty brains as a rule. I think 9 out of 10 times they should get 0 (out of 6) for control and 0 (out of 5) for aggression if they ever go to a judges decision.
That's being said, I was very impressed with your showing at NHRL. Your bot looked so planted to the floor and really appeared to be a class act. It almost made me change my mind about melty brains going to judges decisions.
The problem with melty brains is that when they first appeared, the electronics available where nowhere near fast enough to control them, leading to very sluggish translation and performance. the "sit and spin" style melty that waits for its opponent to come to it
But now, with teensy's, brushless motors and esc's with dshot (high speed comms) its possible to build a melty that translates a lot better. But we are in the early stages of this, theres only a few melty builders out there right now so progress is slow, but it is progress.
what do you use for skid plates? you mentioned them earlier in the video
also hdpe, they are shims cnc'd to a precise height, they also take like 2 minutes to cut, so i just cut a lot of them and replace them as needed
No washers for nuts? Bigger ones?
I wanted to, but with the lay out of the parts I had to put the bolts so close to the outer walls that washers wont fit