Spectrum Resources
Do's and Don'ts
Spectrum's Do's and Don'ts — design, manufacturing, motors, gearing, bumpers, scouting, and strategy lessons learned across our seasons.
Do CAD everything before you build it
Section titled “Do CAD everything before you build it”- There are very few things on our robot that aren’t CADed
- Most of our prototypes are sketched or CADed before building them.
- Sketches / Block CAD / Crayola CAD, etc
Do consider manufacturing while designing
Section titled “Do consider manufacturing while designing”- What tools will be used to make your parts
- How does your design handle tolerances
- How precise does it need to be?
Do consider assembly/maintenance when designing
Section titled “Do consider assembly/maintenance when designing”- Is it possible to assemble these parts
- Is there another part in the way
- What fasteners are you going to use
- Will they interfere with your mechanism
- Do you have room for a tool
- Design in maintenance holes
- Make it easy to replace parts when they break
Do use the most common motor
Section titled “Do use the most common motor”Now we are all brushless: Falcons & NEOs, standardize on either Before we only used MiniCIMs and 775pros/Redlines.
Don’t use less common motors
Section titled “Don’t use less common motors”Window motors, snow blower motors, PG motors, linear actuators, etc are all off-limits for us. It’s harder to stock enough spares and gearboxes that interface with them.
Do limit shafts in mounted bearings
Section titled “Do limit shafts in mounted bearings”- In general we avoid live shafts running through bearings
- We prefer dead axle or shoulder bolt mounted shafts
- Versaroller System is a good example
- The Spectrum Dead Axle Roller System is also heavily used.
- Allows for easier maintenance.
- Shafts can be taken off by removing two bolts.
- Shafts can be replaced as a complete assembly and swapped quickly if damaged.
- Don’t need to worry about rounded hex shaft tolerances
- Gearboxes are the exception
- Our 2022 robot didn’t have any hex/rounded hex bearings
- Our 2020 climber gearbox had two hex bearings for a rounded hex shaft crossing the
- Those are the only 2 hex bearings on our robot
- Our 2019 robot had only 2 hex bearings as well
- Our 2020 climber gearbox had two hex bearings for a rounded hex shaft crossing the
- A few other gearboxes have round bearings for turned down hex shafts
- Our 2022 robot didn’t have any hex/rounded hex bearings
- If we need to use a shaft through a bearing
- Use a REV Rounded Hex shaft.
- Preferably design in specific mounting brackets for the bearings so they can be removed, to replace the whole assembly.
Do build in ways to adjust your gear ratios
Section titled “Do build in ways to adjust your gear ratios”- Planetary Gearboxes
- MAXplanetary Gearboxes are awesome for high torque applications.
- 8mm/Falcon Pinion swaps
- 10/11/12, easily lets you make small adjustments to speed/torque after construction
- Gear families let us change ratios by swapping both gears and keeping the same number of teeth.
- This is less important for high speed, low torque applications like intakes, ball paths, etc.
Do use springs to assist mechanisms
Section titled “Do use springs to assist mechanisms”- Use springs to counteract gravity
- Our most common springs
- Constant force springs
- All elevators
- Torsion springs
- 2018, 2019, and 2020 intakes
- Constant force springs
Bumpers
Section titled “Bumpers”- Do build full wrap bumpers
- They are stronger and will protect your frame better
- You can get bumper hooked in your bumper gap
- You get bumper hooked worse with a frame gap
- Don’t build reversible bumpers
- Your bumpers can switch colors mid match causing you to be disabled
- Do use sailcloth for bumpers
- Design how you are mounting your bumpers
Do organize systems
Section titled “Do organize systems”- Have separate bins for each system to store parts and any specialized tools.
- Allows for quick maintenance and repairs in the pit
- Gives each system a storage location only they use
Do Scout matches
Section titled “Do Scout matches”- Scouting is useful for match strategy and alliance selection
- We currently use iRaiders QR code system
- We have used 180’s Google Sheets scouting system
- Find what works for you
Don’t Scout unnecessary data
Section titled “Don’t Scout unnecessary data”- If not used in scouting meetings or match strategy, don’t scout it
- Make sure to iterate your system accordingly
- Champs 2019, we scouted HAB points, hatches scored, cargo scored
Do define team standards
Section titled “Do define team standards”- Anderson connector setup
- White/Red on the Right when looking from the top
- Zip tie through the holes on every one
- Pneumatic tubing
- White extend, clear retract
- Snap ring standards
- .06“ from groove to the end of the shaft
- Fastener standards
- Bolts are 8-32, 10-32, and ¼-20, Rivet holes are 3/16“, 6-32 bolts are never on our robot
- Programming standards
- CAD standards
- Brand standards
Do prioritize game tasks
Section titled “Do prioritize game tasks”- Prioritize which tasks are important, know what tasks to avoid
- Evaluate goals and make decisions accordingly
- In 2020, going under the trench was not a priority for us
Don’t use weighted decision matrices
Section titled “Don’t use weighted decision matrices”- Especially in a large group where scores are inconsistent
- But even alone, weighted decision matrices don’t make sense
- Make decisions based on consensus
Don’t vote - come to consensus
Section titled “Don’t vote - come to consensus”Don’t build a ramp bot
Section titled “Don’t build a ramp bot”- Good ramps are really hard to build
- Ramps rely on partners to drive on and that is very difficult.
- Platforms/forks are normally better
- If you don’t think champs elims alliances will need to lift a partner don’t build a way to do it.
- We shouldn’t have built the buddy lift in 2020
Do plan for Championship Playoffs
Section titled “Do plan for Championship Playoffs”- If a feature on your robot won’t be used during Championship Playoffs, it’s likely not worth your time to invest in it.
- Get better at the most crucial aspects of the game that will be valuable in every match.
Don’t fall into the sunk cost fallacy
Section titled “Don’t fall into the sunk cost fallacy”- Just because you have put effort into something doesn’t mean you need to continue down that path.
- Re-evaluate based on the goals and weigh options accordingly. Sometimes, this means scrapping designs that have had effort put into them.
- We have removed many things from our robots or completely redesigned things.
- In 2015, completely redesigned our robot from the drive rails up
- In 2017, scrapped our original shooter
- In 2019, scrapped our 4 bar climber for the suction climb
- In 2020, never mounted our buddy climb design
Do document your progress
Section titled “Do document your progress”- Common photo gallery, we use Smugmug
- Easy public uploads
- Every student can upload photos and videos during the process
- We have a media/awards mentor that reminds people to take photos and takes photos
- Weekly Design Reviews
- Each subteam leader is responsible for ensuring photos and information on their slides
- Daily blogs
- Students can send Allen photos and tell him ideas for the blog
- Allen publishes the blog before he goes home every day
- Share your progress
- Use social media, post on CD, or start a blog
Do make handouts for judges
Section titled “Do make handouts for judges”- Robot handout
- Magazine
Do follow common software practices
Section titled “Do follow common software practices”- Do use JAVA
- 80%+ FRC teams program in Java, your team should too unless you have a very good reason not to. (Python is going to become more acceptable in coming years)
- Do use WPILib Command Based
- You don’t need to create your own state machines and things to get your robot to work. If you start with Command Based you will be able to write teleop and auton code quickly.
- Don’t Reinvent the Wheel
- There are dozens of sample projects on Github and for most things you want to do on a robot they have probably already been done. Use that code as an example to get your robot working how you want it to.
- Do be Robust and Simple > Don’t be Cool and Complex
- Use spacers where possible instead of standoffs
- Spacers have one bolt all the way through, standoffs are threaded with bolts on both ends
- Use flat plate construction
- Lets you test quickly with wood, then move to plastic, and finally aluminum if needed.
- 3D print timing pulleys and buy cheap timing belts
- Many mechanisms can be done cheaply with 3D printed pulleys and cheap timing belts
- Have sensors on every motor system (brushless makes this easy)
- Use a design calculator to get a rough idea of the correct gearing for every mechanism
- Use wood on your robot (every Spectrum belly pan since 2015 has been birch plywood)
- Clean your robot - Vacuum, wipe down, simple green
- Use bushings for slow moving parts and pivots
- Do use JAVA
Don’ts
Section titled “Don’ts”- Don’t say that you “Graciously accept”
- Don’t pick up flat objects at an angle
- Don’t have exposed main breakers
- Don’t worry about optimization before it actually works
- Working is more important than being light and fast. Iterate to get it lighter and faster
- Don’t control more variables than absolutely needed.
- Do you really need an adjustable hood, turret, pass through, many degrees of freedom?















