Damian Suarez (RIT Student) Natalia Sokolov (RIT Student)
DeploymentMod
Its the mod that deploys the things
Guiding principles
You cant make all the biscuits at once without burning the house down
Helpful Links
More Info
Updates since presentation
Using stm32f7 - easier to integrate + still has all the features we need
QSPI flash to mitigate data storage rate worries from grim
STLink MINIE instead of STLink MODS. Less space taken up on board technically less features but covers the features we were planning to use on mods. Maintains
Selected LAN8742A for PHY transceiver. Used on the nucleo for the f7 and we’re going to steal their circuit1
Move BME280, LSM6DSL, ADXL375 to SPI (they all have zephyr drivers and have proven to lock up our i2c bus). MS5611 and INA260 stay on i2c bc we either sample them very slowly or don’t wanna write a SPI driver for MS5611.
Presentation Notes
Do we want to make the 2 P-IO pins on i2c pins or PWM pins (routing)
Old temperature Requirements (10F-110F) were silly
0C-50C is prolly good
Pick requirements for sensors more betterly
Gyro: Small angular random walk
How would it work if we lost power? -dean
Altimeter boost detect would probably still work
IMU boost detect wouldnt
Just dont loose power lol
Configurability in field
really vague requirement (telling aaron to reflash them is configurabililty in field)
has a console to reconfigure - better way to say that
Don’t need arming switch
Make boost detect requirement verify by test
Conops was not a conops
wiring was a little nasty
DE9 Pin shenanigans
2x ematch
2x rs48
2x PIO
5v
GND
ematch bypassbreaking out kinda sketchy
but we need to break it out anyways so
BMP388 vs 390. which one is available
board space might be tight
make a harness diagram for a de9
Little plane whats the deal with that
sketchy header thingy to power mod would work
stlink cut first
takes up a lot of space
dont have compatibility with jlink
is convenient tho
Leds should run at 3.3V
Double check buzzer stuff
can it actually make music
History
The first era. Deployment Mod
Originally, we were under the impression that Brain Board was non functional. We know now that it was untested.
Design Goals
Use known working MCUs and sensors from the Grim Reefer to increase the chances of everything working first try. Basically, a Grim with charge deploying capabilities and ethernet connectivity.
Grim had a super set of the features of an RRC3. Had enough inputs to do phase detection but that was not developed as we did not need it.
Be backplane compatible with the ability to act as the charge deployer for a backplane rocket. Have a backplane compatible interface.
Be able to take battery power as a secondary source. This would allow testing on L1s and L2s that could not support an entire backplane. We saw testing as a vitally important aspect because if we are trusting this thing to deploy charges we want it to be proven many times
TLDR; a more flexible RRC3 that could also act as a backplane module.
The second era. Brainsor Mod
At the beginning of the 2024-2025 school year, the project was brought up to the powers that be.
The main points of the first meeting with the wider Avi community brought about the following thoughts:
Brain Board exists, why are we redoing this?
Standalone Test Flights (L1 grim flight) are really important
Since last year, talks of moving Brain Board away from VoCore to one of the bigger STM32s. The idea was that originally, VoCore was chosen because it would be more powerful and flexible compared to launch_core on an stm32f4 like all the other modules. However, the larger stm32h7 series has chips that are quite powerful (some of them exceeding the clock rate of the VoCore even) and Zephyr solves a lot of the complexity that using a linux based board would solve.
A brain board is pretty useless without a sensor module. While modularity is great, we are taking away a slot that could be used for something more interesting (live video board, things like that that)
From there, we decided to sort of combine the ideas of Deployment mod and Brain Board v2 into Brainsor Mod
Design Goals
Use tested sensors - As of Irec 2024, all of sensor mod sensors were working from Aarons assesment and theres a good overlap there with grim sensors we were already planning on using
use an stm32h series chip. Allows us to use new launch libraries w/ zephyr but with greater power and dual cores.
Keep the ability to deploy charges
Drop the dual power sources from the board itself. They provided a lot of complication and with more sensors (sensor mod has more than grim) and the components for deployment, space was at a premium on the board.
In order to still have the ability to do test flights in L1s and L2s, "little plane" was to be developed. It has not yet been decided what form that would take
Biplane: Aaron brought this up as something he'd been working on but it was not discussed much. I don't know what form this takes
Payplane. Still a thought experiment at this point but it also has the ability to fit into a smaller rocket as it will have to fit into a payload
Power Module Passthrough: A double sided board that has brainsor mod on one side and power mod on the other. W/ or w/out a switch. We know powermod works and if we don't require a switch (for a test flight of brainsor mod, we don't necessarily need to record power telemetry), could be incredibly simple - just headers and a little but of routing
Powered Plane: Put electronics to supply the power required by the backplane standard onto the board onto which brainsor mod mounts. Would basically be what the original design of Deployment Mod but on two boards to simplify brainsor mod. Wouldn't need a switch since theres no one to talk to.
We hope to give a PDR at the Avionics meeting by end of September or first week of October