KyberBoard

kyber

The toy lightsaber I always wanted as a kid.

Status

• 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.

Goals

• 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?

fw

Written for the STM32L051K8T6 microcontroller. Code should work on any STM32 micrcontroller, and most interfaces are abstracted into header files.

TODO: state diagram

Configuring STM32CubeIde

Edit project settings to include the multi-project libraries and make the compiler optimize for size. TODO: change saber-libs to holocron-libs

hw

A mildly overdesigned PCB to allow flexibility in mechanical design. Designed for test, assembly, and reconfigurability.

Component Selection Rules

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.

Current (A) Budget

Voltage Loss Budget

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

Subcircuits

Reverse Polarity Protection

Use the built-in protection of the hbridge. Need to characterise I²R losses. This will help prevent the Big Dumb^TM since I plan to use terminal blocks for my main connector.

Low-Voltage Shutoff

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.

3V3 Bus

LDO; I do not want to deal with buck converters. I am not being paid for this design.

Accelerometer/Gyro

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.

Motor Current Sensing

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".

HBridge

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.

Actuators

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.

Non-electrical considerations

• 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

cad

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.