K'nex Build for The Works Museum 2023 - Episode 4

[@Distanc3]

A strange but welcome suggestion by the KZbin algorithm. This was a very enjoyable video! I will queue up the previous parts of this series. Hopefully the algorithm continues to work in your favor!

[@5MadMovieMakers]

"Shifting gears to work on another part of the machine" no pun intended!

[@danielfisher1515]

Fantastic descriptions of your build process!

[@Thibault_Art]

These episodes are awesome!

[@TGHstudio]

Those animations were great. Amazing job, looking forward to the next episode.

[@AnimilesYT]

Every time you drop a video it is a very pleasant surprise. This is starting to look great!

[@HarrisonStaab]

Wow this is incredible. I vividly remember getting the Big Ball Factory for a Christmas present and spending the next two days building it. I'm amazed by your creativity and ingenuity!

[@Austron]

The Big Ball Factory remains one of the best and most impactful Christmas gifts I ever received as a kid! Best K'nex kit ever made!

[@tommaw3204]

Looking great! I love the cables to straighten the beams. True engineering knowledge showing through. It must have been such a wonderful feeling to have 3,000 white rods just appear on your doorstep! Just in general, the amount of support this community exudes is amazing!

[@Dust1218]

This video inspired me! Thank you!

[@spacegames4685]

I love the update videos! I've been watching your channel since I was a kid. Please keep it up!

[@GuitaristSJH]

Awesome work :)

[@Jumbo1907]

Awesome!

[@videoboysayscube]

I'm loving the progress. Is that rod lock an official piece? I've never seen it before.

[@Austron]

The rod locks are real, but pretty rare. You can find them in a few sets, like Son of Serpent or the Nerf guns. You can buy them on places like Knex Replacement Parts, Knex User Group, and eBay. I have a few hundred of the real things, but I also 3D printed a few thousand of my own several years ago.

[@FygerFyre]

Curious if you got a chance to run a distribution test of your diverter, and if you considered employing the same tension idea to induce a reverse bend in that platform for a more even distribution (ie, flex up the middle section of the pegboard where the highest distribution is to bias the balls to the out side paths). Additionally, what are the effects of the failure states of the lifts on that distribution? I'm presuming a normal distribution but skewed in those areas, but I could be wrong hence the question I love the project and look forward to future updates!

[@Austron]

Reversing the flex on the ball diverter is a brilliant idea! I may have to steal it if uneven distribution turns out to be a problem. I haven't run a distribution test for the diverter yet, mostly because I haven't thought of a good way to do it short of sitting there with a pen and paper staring at the machine and trying to manually tally off each ball. For the change of distribution for different failure modes, that is indeed an issue, and I'll be talking about it in the next episode. I can give a slight preview of how I'm hoping to deal with it though. The paths will more or less be separated into the 3 sections of the machine. To keep things more or less evenly spaced regardless of failure modes, the 10 outputs will round-robin through the 3 sections. So going down the line left to right, it will be 3 2 1 3 2 1 3 2 1 1. This will mean that certain paths will be taken less or not at all during specific failure modes, but at least balls will be rolling in all 3 sections of the machine so the idea is that it will hopefully still look like the whole machine is still full of rolling balls. It's possible this won't work out in the end, and I've been considering something like motorized spinning arms in the ball diverter to randomize them even more if it becomes necessary. I think in the end I'm really going to have to wait until most or all of the paths are built to do see how the machine looks in different failure modes and see if adjustments are needed.

[@FygerFyre]

@@Austron dunno how good you are with electronics, but you could set up a simple circuit with a sensor at each exit and have it tally how many times the sensors are touched by balls running through them. But i do agree in the end it will likely require testing to figure out what is working and not. Looking forward to seeing what you come up with in that regard ^^ Oh, and absolutely go ahead and use that idea if it becomes necessary!

[@Austron]

@@FygerFyre I'm reasonably good with electronics, I plan to add sensors to monitor the status of the lifts eventually. I considered using sensors to track the frequency of the pegboard outputs, but I'm kinda starting to run up against the clock here, this machine needs to be installed and delivered in September, so I'm probably just going to leave it untested for now just due to timing. It would be cool at some point long-term to install an array of permanent sensors and have it log the frequency over the lifetime of the machine, though.