mini swerve modules and chasis

Programming complete! PCB time. Today I fixed the issue of module 3’s wheel being offset, and tuned the PID loop. Now the robot finally works well enough to begin pcb design! Designing a pcb will make the final push of testing and programming much easier. I probably wont have may devlogs here for awhile because Im documenting my PCB as a seperate project on hackclub horizons so I can qualify for polaris, but once the pcb is done ill do a final post for this project. This has been very fun, i will definitely do another project before summer ends on stardance.

pid tuning + finding some problems! found some issues in the code varriable defenitions that did not relate to the actual robot, ex some pins were switched that made modules steer backwards rather than the direction intended, mod 3 displacement was wrong. still some issues, like modules 4 and 3 are pretty tense, will have to open them up and see if anythings wrong. probably going to start pcb design about now anyway because I think the code is fine, and I need 20 hours in time to buy a ticket to polaris. anyway super super tired, havent slept in 20 hours :’-)

pid tuning + debugging. i did some bug fixes with the kinematics code, such as fixing radians to angle conversion that i forgot to add, added a step that accounts from slight added drive speed from steering that happens in coaxial swerve drive, and spent 3 hours trying to fix my multiplexer only to realize i accidentially changed the adress. oops! then, I finally began tuning the pid loop, only to discover that turning is weak because the gears that are attached to the steer motors have bad grip on the shaft, because when I made them i forgot to add wiggle room, so during assembly i would sand the insides, making them too loose. so fixed that and reprinted, now in my next time log i will add the new gears and finish tuning the pid, if thats done then this projects time on stardance is complete! when i finish this project, I defenitley want to do some cool stuf with RFs. defo gonna do another project here on stardance.

BIG devlog tonight! swerve drive kinematics code + some testing! today, I pulled through and did all the kinematics code for the swerve drive. it required using the offsets of the modules from the center, the drive joystick inputs, and the steer joystick inputs to calculate vector length and magnitude, then use those to derive a target angle and speed for each module, then speed is scaled down to fit the (0,4095) PWM range for my motors. the target angle is optimized to switch to a smaller rotation if the angle is above 90 degrees, then it is plugged into a PID loop that helps the steer motor achieve the angle. im super tired its like 6 am over here, so Im going to sleep. tomorrow I will have to replace an encoder that went bad on module two, and tune the PID loop. after that, this project will be complete, and I will begin the second half, which is pcb creation, and I will be documenting that on horizons as a seperate hardware project. :-)

code and elecronic fixes + new weird issue I dont really remember what I did last night on the robot, I do remember though that one of my encoders went bad so I had to buy a new one, that should come soon. also, all four of my drive motors have stopped reacting to pwm inputs. after peering more, the motors themselves are fine, and even the h brigde is fine because the motors will sometimes twitch abit during startup from floating values. I think It is the code, will look into that tonight. on a good note I have daisychained power to all 4 of my motor controllers, and begun planning out how I want to program my swerves and learning the mathematics needed for that, so Im pretty excited to do that. hopefully I can wrap this project up soon so I can start working On the ISEF project I wanna do with radio waves :-)

booting difficulties + figured it out + swerve kinematics planning! I was having alot of trouble getting the code to flash to my ESP32. the board would dissapear, loose connection, etc. this began happening after adding alot more components to the board, which eventually led to me learning that some of the pins on my board are completely unuseable due to connections to PSRAM. I need almost all the pins on the board for my project, so I am switching to a different ESP32 model that doesnt have psram, as my project doesnt need all that processing strength. after that, I began learning the kinematics needed to program a swerve drive train, and began taking notes and making plans for code structure, etc.

Soldering electronics to breadboarded system + some coding! Today I soldered all the electronics to the breadboarded system that I haven’t yet, so the breadboarded system is complete and ready to be programmed for 4 modules! I also ran out of I/O pins on my esp32, so I added an I/O extension chip breakout board to the system so I could have enough pins for my drive motors’ encoders, which will be very important for autonomous programs later os For coding, I changed my functions and rewrote some of the code to be suitable for 4 modules, such as changing the drive and steer functions to take an int variable argument of 1 through 4, to control different modules. I also did the same thing for the encoders. I’m tired, so I will sleep, and tomorrow I plan to program the I/O extension chip and the drive encoders, then hopefully get my hands on some 18AWGg wire so I can daisy chain the battery to my four motor controllers, so I can see how just the basic code without any trig math works on my swerve drive train, then just go from there.  hopefully this coding stage doesnt take so long, so I can design a pcb! the messy breadboard is stressing me out >_<, also I wanna make my first pcb, working on this project has helped me learn alot more about data protocols and chips, also how to read datasheets, so Id love to figure out kicad and make something using that knowledge, Oh and I really wanna make the pcb cute with decals n stuff lol. Wow, that was a long devlog prob my longest. Oh, one more thing: I’m pretty sure one of my modules steer motors is broken, so I’ll have to open it up and see what is wrong; probably will have an update on that tomorrow.

completed chasis + starting code! I completed the chasis, and I will begin programming. I just noticed my code isnt working for one module and radio system, so im debugging that, then ill probably go to bed and solder the other 3 modules to the system, then program it. I want to have a working system and code before I try making a pcb, so ill have to deal with this wire mess for now :’-)

assembling final chasis began! I finished the first swerve today and screwed it into the chasis as you can see in the picture, and I would have done the other three tonight, however my printer (my moms technicaly, shes a teacher and she bought it for her class using a grant, but she lets me use it lol) is clogged! my mother doesnt want me to touch it so she will fix it soon, so until then I cant print anything :’-( should be fixed in a few days though.

assembling some swerve V6 + one minor issue + soldering some components!: I assembled about two swerves, before running out of filament, so ill keep printing once I get my hands on a new roll. as for the meshing issues, they have been fixed now! the gears mesh perfectly. however, there is too much wiggle room for the 48t gear in the bottom shell, which causes meshing issues when the 48t gear shifts out of the center, so Ill have to fix that, but after that the design is perfect! yaaay! other than that, I disasembled the older swerve model ive been using for programming, and soldered its components plus the steer motors.

Finished swerves V^! should be the final deisgn, as I went in and fixed all the design errors I was avoiding. everything looks perfect! hopefuly this is the last iteration, ill print it ASAP and see. (ignore the missing leg thingy, in the redesign both are identical!) these files will be on the repository soon.

swerves V6! began working on (FIINAALL) swerve iteration. V5 wouldve been final if I hadnt ignored super obvi meshing issues.. anyway as you can see meshing is perfect. as you can see by the second photo I got a whole lot done. moral of the story, dont ignore obvious problems until they become a hassle to fix and have you awake at 4:41 am :’-)

trying to fix meshing issues. after some research, its likely due to innacuracies of the shaft holder for the drive gears, and misalignment of my bevels. its super late, so ill make some cad fixes tomorrow.

greasing modules + meshing issues! idk why theres meshing issues tbh. im gonna look at the cad and see what I can do :’-(

today I worked on assembling swerves for my final assembly, and I also soddered some encoders for the swerves.

moving project to stardance from forge, and code updates!

today, I moved my projects logs and info to stardance as best I could, because shipping ends today and Id like to ship. my logs from forge are on my repository, and I now have a demo in my stardance if anyone is intrested in viewing the cad. also, last night I worked on the code alot, specifically the system that will use nrf pa lina modules to transmit controller commands to the robot. its not working, and after working around the code abit, I suspect it is because I havent soddered 10uf capacitors to the gnd and 3.3v pins of my nrf pa lina modules, so I will try that soon and update how it goes. the updated code with radio transmission is posted in my repository however, if you would like to see it.