Sever Issues, Tv2 Flight Updates
It's Loud but Promising!
A slower/larger prop will stop the balloon acting like an echo chamber
I reversed the motor polarity and stuck the prop on upside down to make it push the balloon upward. That tiny motor can lift ~9g which is easily enough to get everything airborne. The issue is that the small prop spins very fast which causes vibrations. I could add some foam dampers but the balloon just amplifies any vibrations from the motor. The real solution will probably be using a larger prop with a rubber band or gearbox drive. This will reduce the prop and vibration noise and hopefully give me some more thrust.
This prototype has just the bare minimum to get in the air because it’s best to fail fast. World record rubber band airplanes use thin wire propellers that are covered with mylar. I got some spring steel and titanium wire to try this myself.
It's a Blimp!
This thing is getting closer to the Hindenburg every day
Just going with the blimp design makes the programming and build process much easier. I don’t have the skill to design stabilization systems so I think this is the right move. The Tile will need to move laterally now and I’m sure it won’t be as responsive but that wasn’t really the point. It just needs to hover quietly at eye level so it can answer questions and express emotion. I want this robot to be friendly and approachable. Many humanoid robots look like something to be afraid of but balloons are fun! 45 won’t be able to move quicky but I was planning on having a backpack dock anyway.
Power Control Issues
It took some work to make that N2222A survive
I haven’t been able to properly automate my electrolysis reactor because the Arduino has no way of turning off the PSU. I really need to put a 60A SSR in there so I can automate switching between solar and wall power but I’d still need a way to turn either source on and off. The buck converter that outputs 5V for the reactor had a switch on it so I thought it would be easy to just put a transistor there. It ended up being much harder then it should’ve been because I was trying to have the fan and the buck converter controlled with a single pin. The fan is controlled by a relay which needs a HIGH signal so I chose an NPN transistor (N2222A). This was wrong because turning on the transistor turned off the power supply because of the strange way the buck converter was setup. The transistor also needed lots of resistors and a flyback diode to survive the harsh environment. This should be an easy fix though as I can just use a PNP or maybe change the switch wiring.
Solar Distractions and Lost Posts
I tried to automate my solar but the battery is my main issue
I built a circuit, programmed an AT Tiny 13A and designed a gearbox so I could automate my solar panel but I don’t think I’m going to continue. The solar panels ended up needing more torque then I expected and I bent the two 3mm steel rods in the output shaft. This was a sign to me that I’d need a very expensive motor and gearbox to get this working reliably. The batteries are the weakest part of my system so I’d spend money there first if I had the budget.
I lost two posts that talked about this solar automation side project when my server went down on Wednesday. My server runs Proxmox as the hypervisor and I run all my services in VMs or containers. My Truenas server is one of these VMs and it works fine until I want to shut it down. For some reason the stability of my server is tied to this virtual machine because shutting it down crashes my entire server. This isn’t really an issue because my server is constantly on. On Wednesday I wanted to test some hardware so I disabled “start at boot” for all my services so I could swap hardware faster. The Truenas VM was one of these services and disabling it made my server not boot at all. I was able to enter an emergency shell mid-boot but I wasn’t able to access the configuration file to undo my changes. I was forced to reinstall all my software which took a full day and a half of non-stop “work” (watching progress bars and finding commands) and I’m still not entirely done.
So now I know there is one checkbox on my Proxmox dashboard that bricks my whole server. Thankfully I didn’t loose anything too important so lesson learned I guess.
This Week's Gallery:
Plans For Next Week:
Stay focused on Tv2 hardware, Avoid server-bricking checkboxes
I need to put in some work to finish setting up the AI server and make sure the V1 code is still working. I made some new sensors for the electrolysis reactor so I’ll try to install those if I have time. I really need to get the V2 hardware in a better place because if flies but its not very good at it. Building a wire frame propeller might help solve my noise and weight issues.
Todo List:
- Fix the AI server and test V1 code again
- Build the wire frame propeller (possibly a gearbox?)
- Rework reactor if I have time