The toy lightsaber I always wanted as a kid.
- KyberBoard Rev 1 sent out for fabrication.
- Found Luke Skywalker lightsaber CAD for 3D printing by CaseStudyno8.
- Circuit design complete and subcircuit interfaces frozen.
- Subcircuits passed breadboard feasibility tests.
- First pass on firmware architecture complete.
- Build a lightsaber that extends/retracts, lights up, and makes sound
- Demonstrate "full-stack" electronics engineering skills + some mechanical design
- Stay sane for the rest of the pandemic (everyone needs a hobby)
- Maybe have a Luke Skywalker costume by Halloween 2021?
Written for the STM32L051K8T6 microcontroller. Code should work on any STM32 micrcontroller, and most interfaces are abstracted into header files.
TODO: state diagram
Edit project settings to include the multi-project libraries and make the compiler optimize for size. TODO: change saber-libs to holocron-libs
A mildly overdesigned PCB to allow flexibility in mechanical design. Designed for test, assembly, and reconfigurability.
Keep It Simple, Stupid. Stick to driving circuits that have readily available and easily solderable components. If necessary, target components that have in-stock evaluation boards and take the scrap.
| Component | Amperage | Reasoning |
|---|---|---|
| motor | 10A | stall current of on-hand dc motor |
| solenoid | 1A | 12 V/13 Ohms rounded up |
| 3V3 bus | 1A | typical LDO rating |
| waste | 0.5A | safety factor |
| ----------- | ---------- | --------------------------------- |
| total | 12.5A |
Mostly concerned with battery protection I2R losses. Target < 0.5V peak drop (instinctive spec, not computed)
Major concerns are
- falling below the solenoid turn-on voltage, although I haven't really settled on a solenoid
- throwing off motor control
Current resolution: use ST's VNHD7008AY eval board example design. It seems to work fairly well
Use the built-in protection of the hbridge. Need to characterise I2R losses. This will help prevent the Big DumbTM since I plan to use terminal blocks for my main connector.
Grocery store batteries can't drive the actuators, so I'm looking at using a LiPo. This simple Zener-NMOS-PMOS circuit I designed for a student club will should prevent battery over-discharge.
When the voltage is above Vthresh, the Zener conducts, so the NMOS conducts, which makes the PMOS conduct.
LDO; I do not want to deal with buck converters. I am not being paid for this design.
For that sweet, sweet zwoom. Use I2C/SPI to make routing a little easier. I have a still-in-stock breakout board for the no-longer-produced MMA84252QT.
I may use torque control to retract the saber. Currently planning to extend it with a spring, so the "retracted torque" is theoretically better characterized than the "retracted position".
I found a high-current hbridge IC that does not require complementary PWM (thank you ST). I have limited PWM capabilities due to only low-power ST micros being in stock.
Low-side switches that do not require inrush-limiting resistors to reduce component count. The SSM3K337R,LF stays well under the maximum drive current and well-preserves a 20kHz square wave.
- Cost: 4 layers. 2-sided board gives me the needed density, and routing isn't that awful
- I'm not sure I want to hand-solder QFN packages
- Size: Maximize density to maximize mechanical design flexibility.
- Power budget: Meh, I can increase resistances after fab if it's a battery killer
Found a Luke Skywalker lightsaber CAD for 3D printing by CaseStudyno8. Was also pointed to Disney's new lightsaber, which shows some novel mechanical design for retraction. Although I would prefer to be able to fence with my design, the tape measure strategy does show promise.