Build Your Own Thrust

Vectored Rockets For

Vertical Landings Like
SpaceX


 0
 833
 34


If you love rockets, you can’t help but notice that real space
launch vehicles lift off the pad slowly,but model rockets zip up
like darts. That’s how I became obsessed with using thrust vector
control (TVC) — gimbaling the rocket motor — instead of fins to
keep model rockets upright, so they can launch, and land, far
more realistically.
ZERO TO ROCKETEER
I started out from scratch in rocketry; I’m all self-taught. After
graduating with a degree in audio engineering from Berklee
College of Music in 2014, I saw what SpaceX and other
aerospace companies were going for with propulsive landing
technology and I was hooked. I knew I wanted to get into
rocketry to get a job at one of these companies, and I wasn’t in a
position to pay for another college degree. I figured instead I
could demonstrate what I was teaching myself by propulsively
landing a model rocket the same way SpaceX landed the Falcon
9. It was a literal ―shower idea.‖
I started BPS in 2015 with the goal of achieving vertical takeoff
and vertical landing (VTVL) of a scale model Falcon 9 rocket.
This would require me to solve two tough problems — thrust-
vectored flight and propulsive landings — using solid-fuel hobby
rocket motors.
I picked up a few textbooks (I strongly recommend Rocket
Propulsion Elements by George Sutton and Structures by J.E.
Gordon), found a few good YouTube tutorials for coding and
mechanical design, and got to work experimenting.
I naïvely thought it would take four months. My first ten launches
were failures. But the eleventh succeeded, and the successes
accelerated after that. After four years of hard work, crashes, and
iterative designs, I’m now achieving beautiful thrust-vectored
launches of several rockets, including my 1:48-scale Falcon
Heavy — three cores! — that you can watch on my YouTube
channel.
And after some very near misses, I’m confident my Echo rocket
will stick the vertical landing in 2019. Rather than attempt to
throttle a solid-fuel motor, I’m firing the entire retro motor —also
TVC’ed —at the precise time and altitude to enable a soft
touchdown. It hasn’t been easy.
KITS FOR MAKERS
Of course this technology is still not mature, and it’s my hope that
the advanced model rocketry community will build upon what
I’ve done. In 2017, I began selling my Signal flight computer
board and TVC motor mount together in a kit. After getting user
feedback, the computer was redesigned from the ground up to
include Bluetooth — Mission Control from your phone!
BPS.space is now a proper company and a full-time job for me,
funded through flight computer sales, the BPS.space Patreon
page, YouTube ad revenue, and sponsorships.
This kit is for advanced rocketeers. If you don’t have experience
with scratch-built rockets, I recommend you hone your skills first
with an Estes ready-to-fly kit and seek advice from fellow
rocketeers at the National Association of Rocketry Facebook
Group and the Rocketry Subreddit.
And if you’ve got some experience, I hope you’ll give it a try!
I’ve even shared a scale model of the Rocket Lab Electron you
can build using my tutorials.
THRUST VECTOR CONTROL – Ready to Aim Fire
Model rockets have fins and launch quickly, but real space launch
vehicles don’t; they actively aim — vector — their rocket exhaust
to steer the rocket. With thrust vectoring, your model rockets can
slowly ascend and build speed like the real thing, instead of
leaving your sight in seconds.
THE COMPUTER
The Signal R2 flight computer runs a high-speed control loop,
prioritizing functions depending on the progress of the flight.
Thrust vectoring draws considerable current, so once burnout is
detected, Signal centers and locks the vectoring mount. Focus is
then set on detecting apogee and triggering pyro events. It needs
at least 8V; 9V alkalines or 11.1V LiPos recommended.
THE TVC MOUNT
Developed over three years of iterative design, the thrust vector
control motor mount is made from 3D printed PLA material.
It uses two 9g servos, geared down for higher accuracy. The
assembly can gimbal a motor ±5 degrees on each axis, X and Y.
Though the mount will work with up to 40N of force, it works
best with lower impulse motors, especially those with long burn
times.
THE SOFTWARE
The flight software tracks vehicle flight dynamics while the
rocket is powered on. Signal looks for cues to shift system states
at liftoff, burnout, apogee, and landing. This makes operation
simple — once the settings file is configured for flight, just turn
on the flight computer and it automatically enters Pad-idle mode.
Signal can detect launch in under 10ms. Once detected, thrust
vectoring is activated, in-flight abort is armed, and high-
frequency data logging begins.
THE DATA
In-flight data logging takes place at 40Hz. Vectoring output,
vehicle orientation, altitude, velocity, acceleration, and several
other data points are recorded using a custom protocol to a high-
speed flash chip. Upon landing detection, Signal creates a new
CSV file on the microSD card, dumping flight data into it. Once
the data is verified to match, the flash chip is cleared and Signal is
ready to fly again. A 1GB card can store hundreds of flights.
Flight settings are programmable in a settings file on the card.
THE APP
Signal is configured using an app on an iOS or Android
smartphone.
The app helps the user configure TVC sensitivity, parachute
deployment altitudes, the abort system, ground testing, rocket
tuning, and more. My goal is to put Mission Control in your
pocket.
Build your own scale model Rocket Lab Electron by following
the ―Build Signal R2‖ tutorials at the BPS.space YouTube
channel, and try it out!
PROPULSIVE VERTICAL LANDING – Drop It While It’s
Hot
When I started building my Signal TVC kits in spring 2017, I put
my propulsive landing project on hold. This past year I got back
to it: I started drop-testing my experimental Echo rocket from a
drone and began a new YouTube series, ―Landing Model
Rockets.‖
The series explains my entire process, from selecting motors, to
planning the flight profile, to developing a new control board for
propulsive landing. It comes in two flavors: Blip, a DIY version
using breakout boards and through-hole components; and Blop, a
lighter, surface-mount PCB. Both are based on the powerful
MK20DX256VLH7 processor that’s used in the Teensy 3.2
microcontroller, with a Bosch BMP280 barometric pressure
sensor and InvenSense MPU-6050 inertial measurement unit
(IMU). For now I’m sharing these experimental boards with my
Patreon supporters.
So how can you land with solid rocket motors if you can’t throttle
them? It’s all about timing! If the flight profile is fairly well
known and the legs are built to withstand small hops and drops,
the motor can be fired at just the right time to slow the vehicle
down for a soft landing. As the rocket nears the ground, the
microcontroller reads the barometric altimeter and fires the retro
motor.
Four landing legs deploy by rubber band,
and the rocket lands upright. In theory.
Turns out vertical landing is super hard! As the rocket slows to a
near-hover, horizontal drift becomes an issue. I’m already deep in
the weeds programming the math necessary to control for this.
I’m also experimenting with a tiny LIDAR sensor for really
precise rangefinding to the ground.

After several very near misses (watch them on YouTube), I refuse

to not stick the landing in 2019! Follow me on Twitter to stay up
to date.
YOUR ROCKET MOTOR QUESTIONS ANSWERED
Q: Do you make your own motors?
A: Nope, I use commercially available propellants. Usually black powder
motors from Estes or APCP motors from Aerotech. For custom or more
complicated builds, BPS.space outsources propellant work to other
manufacturers, who can achieve higher precision and predictability. I
learn best by experimentation, which works very well for software and
electronics, less so for explosives and propellants.
I also got a C+ in chemistry during high school — designing and making
rocket propellant is not a good idea for me. :)
Q: I want to make my own rocket motors. Do you have any advice?
A: I strongly recommend against building your own rocket motors unless
you have an experienced mentor or teacher to help you. There’s certainly
information about building these motors online, but much of it is incorrect
or lacks enough consideration for safety.
Q: I heard that active control in model rockets is illegal, can’t you get in
trouble for that?
A: Nope, it’s not — active control has been around in model rocketry since
the 1980s! None of the U.S. regulations prohibit active stability in model
rockets. There’s an important distinction here between guidance and
stability.
Guidance is usually about maneuvering in reference to one or more real
points in space. This might be via GPS, GLONASS, RF, or even dead
reckoning.
Stability is just about keeping the rocket upright, which serves the same
purpose as fins on a traditional model rocket. Signal does not carry any
GPS or point-based guidance equipment. The computer is only able to
keep the rocket vertically stable off the launch pad.

MY LITTLE FALCON HEAVY

MEET THE MOST ADVANCED SCALE ROCKET ON
EARTH.
Myself, the Falcon Heavy model, and Tim Dodd, The Everyday Astronaut

In December 2017, I set out to build a 1/48 scale model of the

SpaceX Falcon Heavy rocket as a technology demonstrator for
BPS.space, to showcase the advancement that’s now possible at
the model scale. As usual, I underestimated it as a four-month
job! I finally launched it in November of last year.
Flacon Heavy model – Flight 1

Each of the three cores carries a flight computer, parachute

deployment system, and TVC assembly. The upper stage flies
with thrust vector control as well, and it carries a 3D-printed
sports car — no Falcon Heavy model would be complete without
one.
An early CAD render of the vehicle

AVIONICS
The Falcon Heavy flight computers are upgraded versions of
Signal (see page 32). Each computer has MEMS gyroscopes and
accelerometers for sensing movement and orientation on the
rocket, the same kind of sensors found in most smartphones. A
barometric pressure sensor determines the rocket’s altitude above
the ground.
On each flight computer, a 48MHz Cortex M0 processor reads the
flight sensors at 400Hz and logs 31 channels of data to a flash
chip, 40 times per second. Every second that Falcon Heavy is in
the air, 4,960 points of data are recorded onboard — never too
much data.
The flight computers don’t communicate with each other during
flight; the sensors and software are accurate enough that it’s not
required for short flights. None of the flight settings are hard-
coded; depending on the flight profile, they can all be changed
using the Signal iOS/Android app.
The flight software is written in C++, the iOS app in Swift, and
the Android app in Java. Flight guidance is computed
using quaternions, a complex number system for 3D orientation
and rotation that’s more computationally efficient than other
methods (see youtu.be/d4EgbgTm0Bg for a wonderful
introduction).
MECHANICS
The attachment points at the top of each side core slide down a
ramp on the center core, giving them a bit of clearance during
stage separation. This passive setup keeps things very simple
during flight, but the cores are usually bolted together while the
vehicle is on the ground. All three cores are also connected at the
base of the vehicle using a slightly simpler thrust plate.
The side cores remain attached to the vehicle by maintaining a
slightly higher net thrust force than the center core/upper stage.
As soon as the center core produces more net thrust than the side
cores, it will pull away and the stages will separate. For
flexibility, each side core also has a slot for a small separation
motor; I haven’t needed to use sep motors on flights so far, but
they may be helpful later down the road.
The center core of the Falcon Heavy model goes through
two boost phases during flight. During the first phase, the side
cores are attached. Right around side core burnout,
the center core lights a second motor, and
the second boost phase begins. These two rocket motors are
mounted on top of each other in the center core’s thrust vectoring
mount. When the second motor ignites, the lower, spent motor, is
ejected. This same technique, called hot staging, is used to control
ascent and propulsive landing motors in other BPS.space rockets.
The rocket motors in each stage of the Falcon Heavy model can
be gimbaled ±5 degrees in any direction. Because the side cores
are not firing directly through the vehicle’s center of mass, they
can be used not just for pitch and yaw, but roll control as well.
Each multi-core flight runs a roll program which usually targets
20 degrees of positive roll, executed at 30–40 degrees per second.
Launching the Falcon Heavy model requires a bit of forethought.
The center core motor has a slight thrust spike at ignition, after
which the amount of force produced slowly tapers off. At liftoff,
the side cores must have a greater net thrust than the center core.
Because of this, the center core is lit at T–1 second while the
vehicle is still held on the pad by the launch clamps. At T–0 the
side cores are lit, and at T+0.25 the beast is released.
Also, do I even have to say it? Of course those boosters are gonna
land! The propulsive landing test program is isolated from the
Falcon Heavy program right now, but the two will merge as the
success rate for both programs increases.
LAUNCH PAD 2.0
After five iterations, Launch Pad 2.0 was redesigned from the
ground up to support both single- and three-core rockets. With 8
total launch clamps, and an easily modifiable iron flame trench,
Pad 2.0 is my most flexible platform yet. My Impulse launch
computer, based on an Atmel ATSAMD21 microcontroller, has
plenty of inputs and outputs to support add-ons like load cells,
wireless communications, and any other peripheral that
communicates through I2C or SPI.

WHAT’S NEXT?
» Fine-tuning the TVC — I continue to launch the Signal R2
board on my new Scout D1 rocket to improve the tuning feature
in the software.
» Slow-burning propellant — Most model rocket motors burn just
3–7 seconds, not long enough to simulate real space vehicle
launches. I’ve been working with Aerotech to develop long-
burning motors such as their new G11 and G8ST that burns for
almost 18 seconds!
» High-powered rockets — I’ve been getting involved in the HPR
community, where amateur rocket flights go miles high and
safety is a lot more important. So I’m now developing the Arc
dual deploy board: 2 pyro channels for firing main and drogue
chutes, barometric altimeter up to 100,000 feet, multi-speed data
logging, and Bluetooth for smartphone configuration.
» Reaction control system (RCS) — One more thing: Rather than
attempt TVC on high-power motors, I’m developing a reaction
control system — six little cold-gas thrusters near the top of the
rocket to improve stability
— and of course a new Relay RCS board to run it.
 ADVERTISEMENT

By Joe Barnard

By Keith Hammond
October 25, 2019, 6:00 am PDT
ADVERTISEMENT

RELATED TOPICS

 MAKE69

 SPACE
ADVERTISEMENT

ADVERTISEMENT
How The Heck Do Eink
Displays Work? Let Collin
Explain


 0
 0
 0


Eink displays have been getting more and more readily available.
Now is a better time than ever to include one in your project.
They come in lots of different sizes and even now have multi
color options.
Collin, in the conintuation of his ―Collin’s Lab‖ educational
series, tackles this display technology. He explains in simple
terms exactly how these work, demystifying the curious power
saving abilities of the technology. As usual, fantastic work from
Collin and a must-see for anyone who has considered using these
panels.

By Caleb Kraft

@calebkraft
 19 hours ago

RELATED TOPICS

 EDUCATION

 EINK
ADVERTISEMENT

ADVERTISEMENT

Maker (Faire Producer)

Spotlight: Ian Cole


 0
 0
 0


Ian Cole has been at the helm of Maker Faire Orlando, and its
parent organization The Maker Effect Foundation, since its
inception. While this spectacular, volunteer run event did not take
place in November as usual, Ian has been no less busy
manifesting maker experiences for others (and acting as a board
member and advocate for makers across the U.S. through Nation
of Makers).
Florida has more Maker Faires than nearly any other U.S. state
(New York wins that prize) and Ian has been deeply involved in
building that community, both within Orlando itself and by
supporting other Maker Faires across the state. This ethic and the
spirit of collegiality behind it shows in the choice the MFO team
made this year: Rather than pushing the event virtual, as many
Maker Faires have done, Maker Faire Orlando decided to step
back and keep things small and local under the logic that it is the
community and the experiences it generates that make the event
special, regardless of scale. And, also, that there was just an awful
lot of Zoom going around. As such, they have have held several
pop ups where makers share projects in an outdoor space (the
parking lot of the MakerFX Makerspace) and a Maker Takeover
at the Orlando Science Center on December 19th. MFO have
also fed the need of the strong local community of BattleBots
enthusiasts—which included organizers & competitors from
BattleBots teams Sporkinok, Extinguisher, and Kraken—with a
modified (and live streamed on Twitch) Robot Ruckus that you
can check out here.
Producing a Maker Faire, especially one the size of Maker Faire
Orlando, builds some serious production chops. Earlier in the
year, as uncertainty swirled around events across the globe, Ian
and his team (including his wife and Maker Faire producing
partner, Candy) spearheaded an organized maker response
network to provide medical supplies across Central Florida, a
region that by virtue of its large population of retirees, was caught
in the crosshairs of the global PPE shortages. Revealing his
genius at networking and motivating people to collaborate, Ian
chose to focus on the distribution and supply chain aspects of the
process, as opposed to producing PPE at scale within the
MakerFX space. The civic response, ingenuity, and commitment
of regional coordinators like Ian under Open Source Medical
Supplies has not only served many frontline workers and at risk
individuals, but also revealed the wellspring of distributed
manufacturing and rapid prototyping that is one of the great
success stories of the year (see our Plan C Live program for more
maker civic response stories).

As if all this wasn’t enough AND because his recent projects

capture the generosity (and sparkle) of the holiday spirit, we
reached out to Ian to share with us some of the other work he’s
been doing. In his day job, Ian works as Chief Innovation Officer
at the whimsical nonprofit Give Kids The World Village, an 89-
acre resort in Kissimmee, FL, that provides cost-free vacations to
critically ill children and their families from around the world.
This year, GKTW decided to do something new, both as a
fundraiser for the community they serve, and to bring some
holiday joy at the end of an undeniably challenging year. The
result is Night of A Million Lights (actually 3+ million lights), a
display across the entire GKTW campus that is ticketed and open
to the public. Again, those Maker Faire production skills came in
handy. ―We’ve been closed since March due to the pandemic, and
Night of a Million Lights was a way for us to fundraise while also
driving significant awareness of our mission. This is the first time
in the history of the Village that we’ve opened our gates to the
public. My past experience with Maker Faire Orlando was a huge
benefit as we designed our ticketing and entry process as well as
other aspects of large events that are open to the public.‖ Check
out the aerial feature photo to get an idea of the sheer scale of this
undertaking.
Here’s a behind the scenes look at the making of the light display:
Ian has also been spearheading the Designer Desks project with
partner the Boys and Girls Club of Central Florida. The project
makes art desks for underprivileged youth in Central Florida at
the MakerFX Makerspace. Each desk is stocked with art supplies
and the intention of giving kids who don’t have it a space to
create. Local artists have also decorated some of the desks before
delivery. If you’d like to make and give desk of your own, the
design is open source and available on GitHub. This project has
an active fundraiser with all funds going to purchase wood and art
supplies at a discount from local stores—you can donate here. Ian
and the team are also available to share project details with those
interested in setting up a chapter in their own city.
And because Maker Faire would hardly be Maker Faire without
making, Ian and the MFO team have been bringing Learn-To-
Solder badge making into the community. Through this new
partnership with Boys & Girls Club of Central Florida, the MFO
team created a new branded Learn-To-Solder badge for Boys &
Girls Clubs, and have equipped the clubs with soldering stations
from Adafruit, cutters, solder, solder braid, safety glasses and
more. ―We brainstormed how we could still teach more than
1,000 kids to solder without having our big Maker Faire Orlando
event in 2020. This partnership came from that determination that
we were still going to make an impact, even during a pandemic.
With such an amazing partner in Boys & Girls Clubs of Central
Florida, even more kids are learning to solder while not
increasing their COVID risk – and we are already discussing how
we get those kids to Maker Faire Orlando in the future!‖ If you
would like to replicate this program with your local Boys & Girls
Clubs, reach out to the Maker Faire Orlando team
at [email protected]
Not to be outshone by a mere 3 million lights, Ian has also been
developing the BrightBikes project that inspires makers to create
bikes festooned with interactive lights and meet up for rides.
Want to make your own? Or start a ride group in your own
community? Here’s the Bright Bikes repo on GitHub. Follow
their Facebook page. You can find the process videos below
on Maker Faire Orlando’s YouTube channel.
As with all community projects, although Ian’s leadership is a
secret sauce to getting them started and seeing them through, he
notes that ―they are only possible because of the ingenuity,
generosity, and dedication of our amazing volunteer network,
partnership with other local community organizations and with
Orange County’s continued support of Arts & Culture through the
pandemic.‖ It seems fitting to add here a standout statement on
The Maker Effect website that embodies Ian’s collaboration (and
the Maker Movement as a whole) with his community and the
energy and heart he devotes to his many projects. ―The Maker
Effect is the sum of the personal growth, professional success,
community development, and continuous innovation that results
when makers learn, educate, share, and create together.‖
Feature image from @bioreconstruct on Twitter, who shared
night helicopter views of the GKTW Village on December 6th
that really helps give you an idea of the scale of Night of A
Million Lights.