Welcome to our 15th build blog. Since 2012, we have been documenting our strategy, prototyping, and progress to allow everyone to follow along and learn from our season.
Our pre-kickoff slides have led off with an abbreviated version of this excellent post for some years.
This year is an opportunity to do more for yourself, your team, and the community. Being part of an FRC team is a privilege many people don’t get. Embrace your opportunity and bring your best self to your team each day.
2026 Season Links
- Website: www.Spectrum3847.org
- Public Onshape
- GitHub - https://github.com/Spectrum3847/2026-Spectrum
- Photo Gallery
CAD Collection
Nearly 850 links to FRC CAD models. If you have posted your 2025 Robot CAD please submit the link to the form below so we can include it in our collection.
- Submit CAD to the collection - FRC CAD Collection - Spectrum3847
- View the Collection - FRC CAD Collection | Spectrum3847
Resources
- Robot Exploration Videos
- These videos explain FRC concepts in the contex of the robots that use them.
- Training
- Links to all of our Spectrum training Videos
- Conference Presentations
- These are the talks Spectrum coaches have been a part of at Mentor Workshops
- FIRST $2000 Tool Guide (and archived FIRST $1000 and $1000)
- A guide to purchasing tools for your team.
- The first $2000 is for essentials, and then the next $6000 includes ways to expand your shop.
- Next Steps goes even further with the more expensive tools used by some FRC teams.
- Illuminations - Young Team Guide
- A Guide for young FRC teams to quickly understand key aspects of FIRST Robotics.
- Recommendations for young teams
- Swerve Guide
- A basic introduction to swerve for new FRC teams.
- Cart Guide
- A simple, customizable cart for FRC Robots. It can be easily folded down to allow it fit inside a car trunk.
- Made with standard FRC tubes and parts.
- Inexpensive Build Tips
- 2026 FRC Voucher Checklist
- To help you keep track of your FRC Vouchers, make a copy and add your voucher codes from the Lead Mentor 1 or 2 dashboard.
Robot Plan
For this season, we are changing the structure of our robot development process. In the past we have made versions of these robots, but we have always planned to have our Final robot before our first competition. This season, we are planning to give ourselves time to make design decisions after our first event and after we have seen Rebuilt played at real events.
1. AM (Alpha Machine)
- This is going to be our first true Alpha robot. It’s built in the first 5-7 days of the build season.
- It doesn’t have to meet any of the rules for expansion, height limit, etc. It can have mechanisms that are always deployed like an intake.
- It will be built on top of an already existing MK4i drive base. It could be MK5n if we think it’s required for space claims.
- It can have lots of plywood, etc.
2. XM (Experiment Machine)
- This is closer to the AM robots we built in 2023 and 2024. This will be a mostly plywood-and-tube quick build of the architecture for our competition robot.
- This will run on our spare MK5n modules; these will come off before the competition, as once PM is running, we shouldn’t need XM anymore.
- This gives us a first pass at mechanism design before we design the initial competition versions.
3. PM (Provisional Machine)
- This is the first iteration of our competition robot.
- This robot will play at our first event (Comp Week 1)
- This robot is fully powder-coated and runs MK5n modules
- We will have spares for this robot built that can become the spares for FM.
4. FM (Final Machine)
- This robot is finalized after week 1 or 2 of the competition season to determine whether we need to redesign our robot for Week 5 or State champs.
- We should have the drive base for this robot complete by Week 1 of the competition.
- This robot build is done after week 1 or 2 to allow us to set up for a “rebuild” for States/Champs or a 2.0 version of our competition robot.
5. Photon (Development Team Robot)
- This robot is designed to share multiple mechanisms with PM, but it will include 1 to 2 mechanisms designed by 8515 students to allow for more iteration.
- Will have the same drive base size as PM.
- Has to be complete for Week 3 Competition at Houston District.
Spectrum Design Guidelines 2026
- Guidelines, not rules, can be broken, but don’t do it too often.
- Don’t Do list
- Don’t be that far off the Meta design
- If you’re doing something really weird, stop and think really hard about it.
- Does the advantage you think you are getting matter? Will it matter on Einstein?
- No launching during pick-and-place games
- No scissor lifts
- No pinch/claw intake from the floor
- Always spinning intakes with active eject when possible
- No pink arms (Telescoping arm with center pivot)
- No picking up flat objects at a steep angle
- Did it in 17 and 18, and it was a mistake both times
- No ramp bots
- platforms/forks are okay, teams are awful at driving up ramps
- No Mecanum, h-drive, kiwi/Omni/x-drive, tank treads (Everyone should be swerve if they can)
- No 6-32 bolts (4-40 and 8-32 will almost always work or 3mm)
- No pneumatics (20 motor slots now)
- No small motors, no brushed motors
- no NEO550, 775pro, Minion, etc
- Servos may be okay for some applications
- No motors in rollers/tubes designs (neat but hard to maintain)
- No bumper or frame gaps (unless rules don’t allow extensions)
- This prevents your frame from getting bent
- allows bumpers to take all impacts properly.
- Don’t chase magic numbers.
- If there is only a single optimized dimension in which your mechanisms function, that typically means it won’t maintain that throughout the season. Especially if one direction makes part of the mechanism easier and the other makes another part easier, finding that perfect balance is very difficult.
- This could be an intake opening, a specific hood angle, etc.
- Avoid Gimmicks
- 2025: taken L1 and put higher up, center align for both reef poles
- 2019 & 2020: buddy climbs
- anything you don’t plan to use in a perfectly executed Einstein match, anything that takes time and doesn’t score points.
- Don’t be that far off the Meta design
- Fasteners
- 10-32 socket head is standard
- Alloy steel: zinc, or black oxide-coated
- Never stainless steel (70k psi vs 170k psi tensile strength)
- Possibly Titanium or Nylon if weight is a big issue, but hopefully not
- Vibration Locking on every bolt
- Nylock nuts or nylock jam nuts, no regular nuts
- Loctite or
- NordLock(Wedge Lock) washers
- Button or flat head when needed
- Alloy steel: zinc, or black oxide-coated
- 3/16 multigrip rivets
- Steel mandrel, aluminum head
- Dome or countersunk
- Use other bolt sizes sparingly, only when required by some COTS part or size requirements
- 10-32 socket head is standard
- Use COTS parts effectively
- Besides cut plates, prints, standoffs, rollers, etc
- Materials
- Pre-drilled tubes of any size
- .03”, .04”, 1/16" or .09”, 5mm aluminum
- Can bend flanges for more strength
- MAXcomposite, SRPP
- .03”, 1/16", ⅛" Polycarbonate
- ½" & ⅜" Rounded Hex Shaft
- REV Rounded Hex ½” and 10-32 Spacer stock
- WCP ⅜” rounded hex (7075)
- REV Maxspline shaft and polycarbonate Maxspline shaft
- Polycarb, Nylon, PACF filaments
- Don’t use PLA on comp robot mechanical parts, covers, spacers, etc. It is fine. (It works, but we had issues with cracking last year)
- Multilayer parts okay, polycarb with aluminum sandwich, etc.
- Carbon fiber tubes are used sparingly.
- Swerve
- MK5n modules
- Potentially some kind of wheel drop setup if we have to climb obstacles of some kind, and the 4” wheels and motor clearance is a problem.
- Bumpers
- Something new with the new bumper rules
- Likely using the FRC#5000 air tool quick-release mounts
- Use a mounting rail structure like 3538-2023 when possible
- allows for clean belly pan electronics
- Allows for mounting intake plates, etc all the way to the corners, which is harder with the MK5n modules.
- Allows for more cross bars if needed
- Side plates can be enlarged for 2337-2022 style mounting
- Handles
- Handles should be integrated into the robot so that it’s easy to carry on and off the field, and allow more students to lift the robot easily
- Handles like these - McMaster-Carr
- Power transmission
- High torque/low speed
- 25 chain, turnbuckle tensioners, or other tensioners
- MaxPlanetary+motor
- Max90s work well
- Shim tape the hex input and output, and the bevel to the hex faces
- Use the assembly wedges when assembling
- High speed
- 5mm HTD belt
- Motor pulley pinion (normally 12t, 11t if needed)
- 15mm wide when possible, 9mm if space is needed
- 15mm can transfer more torque and is harder to tear
- 3D printed pulleys work well
- No Pneumatics
- Support the end of motor output shafts with a bearing when possible
- High torque/low speed
- Shafts
- ⅜”, ½” hex and maxspline shaft when possible.
- Snap rings are good
- Light, easy, reliable
- It is better to have snap rings and tapped 10-32 with a washer in the ends
- Bearings for all fast motion
- Launchers, intakes
- Bushings for slow motion are ok
- Arms, wrists, pivots, etc.
- Electrical
- Battery
- Flat-mount battery on the belly pan, if at all possible
- Battery strap with metal buckle - Amazon.com: Sea to Summit Accessory Strap with Hook Release - pair (20MM / 3/4" Webbing by 1M Long) - Color May Vary : Sea to Summit: Sports & Outdoors
- Zip tie battery connector every match
- Leads always point up from the battery, NEVER to the side
- Main breaker
- NEVER OPTIFUSE!
- Nordlock and Locknuts ¼”-28 (or 6mm for the yellow handle breaker)
- 3D printed breaker shroud
- Wire Gauges
- 8-12awg for all motors (exceptions can be made for specific reasons, but 12awg is usually small enough for all motors)
- 18awg for device/aux power
- 24awg for sensor wire
- Inline Wagos for power
- Kraken ring terminals direct to PDH
- Inline dual lever nuts for can and signal/low current wires
- 2 to 6 for swerve
- 3 lever for LEDs, other sensors
- 8-port Ethernet switch
- Slim Ethernet cables
- Power regulator instead of the barrel jack
- Mount the radio above the bumpers and away (6in +) from the motors
- Use a thermal pad behind the radio to help with heat.
- LED strips should be visible from all sides of the robot.
- Cameras
- Limelight for Pose/aiming
- Limelight+coral or PhotonVision for game piece tracking
- Fisheye for the driver camera should be able to see the intake
- Sensors
- Use internal encoders whenever possible, zero at boot, and on a gamepad button
- CANrange for game piece detection in the robot
- Swerve absolute encoders
- CAN absolute encoders, if needed for pivots, etc. Thrubore where possible
- Color sensor??? - something over network tables, probably photonvision or limelight, maybe (would be an expensive color sensor)
- Only the carnivore is connected to the RIO-usb port
- Battery
- Aesthetics
- Primary purple powder coat, some white powder coat
- addressable LEDs
- Sponsor panels
Things to consider
- Robot starting configuration
- Generally, being close to the max frame perimeter is preferred.
- Handles and carrying the robot
- Electronics layout, gyro, Rio, PDH, etc.
- Wiring paths
- Tether location
- Practice full matches, from cart to cart
- Avoid shock loads on any systems, even if they work during testing, they will eventually cause damage over the whole season.
- In reach games, however high you think you need to reach, make sure you can reach 4”+ more than that. (2018, 2019, 2023, 2025)
- Building in more reach, you can always limit it in software or with added stops, but adding more reach if needed is way more complicated.
Idioms/Quotes
- “Steal from the best, invent the rest.”
- Copy everything you can from other teams, use COTS parts when possible, past designs, etc.
- “Don’t reinvent the wheel.”
- “The Best Ability is Availability” - Bill Parcells (NFL Coach)
- Your robot has to be able to play and work in every match. It needs to be consistent.
- “KISS: Keep it simple, silly/stupid/etc.”
- “Everything should be as complex as necessary and no more.”
- “One is none, Two is one.”
- always have a spare
- “One is a gamble, two is a plan.” “ If it’s critical, it needs a twin.”
- “One is a fluke, ten is a trend, fifty is a feature.”
- One success isn’t the goal; consistency is the goal.
- “Don’t test it until it works; test it till it can’t fail.”
- “Slow is smooth, smooth is fast.”
- “Take your time in a hurry.”
- “Be quick, but don’t hurry” - John Wooden.
- It’s better to be well practiced and consistent at a lower top speed than be jerky and inconsistent at a higher top speed.
- Rushing causes mistakes. Even when you’re in a rush, do things the right way and stay in control.
- "Pixels are cheaper than aluminum.”
- "If it’s not in CAD, it’s not on the robot.”
- Everything is done in CAD before it’s done on the robot.
- “Doing it right is faster than doing it twice.”
- “Measure twice, cut once.”
- “Do it right the first time.” or “If you don’t have time to do it right, when will you have time to do it over?” - John Wooden (Basketball Coach)
- “Nothing will work unless you do.” - John Wooden (Basketball Coach)
- The best robot design will fail if the team members aren’t putting in the effort needed to make it work.
- Also means the opposite is true if you work too hard to the point where you harm your physical and mental health; the eventual outcome won’t be good either.
- “Design for the field, not the shop”
- Murphy’s law: Anything that can go wrong will go wrong.
- Needs to be built robust enough to work after wear and tear from a tough season. Needs to work in non-optimal conditions.
- “Trust, but verify.”
- Double-check that things are actually done correctly. Quality control is essential.
- Never Assume
- “Planning is Indispensable, Plans are negotiable.”
- “Failing to prepare is preparing to fail” - John Wooden
- Plans will always change during the build, match, event, etc.
- The act of planning and thinking through all possible options is more important than perfectly following the plan.
- “The exception does not make a rule.”
- “The exception is not the example,” “This exception is not the model.”
- “Little things make big things happen.” - John Wooden
- Attention to detail makes the larger project succeed.
- “Set screws sucks” - Woodie Flowers
- “Fail fast”
- “If you’re not making mistakes, then you’re not doing anything” - John Wooden
- The goal of early development should be to identify failures quickly, so you can iterate quickly to avoid them in the future.
- Communication
- “If it isn’t documented, it didn’t happen” - Write it all down, Google Docs, Slack, etc.
- “Hard on problems, soft on people” - failures are on the team and processes and not on any one person.
- “Listen to understand, not to respond.”
- “Don’t raise your voice, improve your argument.”
- “Too many cooks”, “Too many cooks, spoil the broth”
- If too many people are trying to do a task, the outcome is worse.
- “Bad news doesn’t get better with age.”
- Tell the team about failures/breaks/problems as quickly as possible.
What to do during the first week
- Maximizing Week 1
- Our presentation from the 2024 FRC Mentor Conference
- We also added some tips to the end of the slides
- Robot Architecture
- An overview of the process we use to decide our robot architecture each season.
- Prototyping Presentation
- An overview of the various types and styles of prototypes and some of the tools and resources we use to prototype quickly.
- KrayonCAD Examples
- Example KrayonCAD robots that some of our students created over Winter break.
Team Organization and Communication for the Season
- Part Sheet Example
- This is the sheet we use for generating part names for our robot.
- Based on 1678, but we should share a public version of ours.
- Task lists in Slack Canvas or Lists, or in Google Sheets
- Every year, we end up having a to-do list somewhere.
- We have been using Slack Canvas and Lists since 2025 and will continue to use them in 2026.
- In previous years, we have used a Google Doc. Task Google Sheet Template
- Mechanism Design Leads
- We split our robot into 4-5 mechanical mechanisms (we try not to have more than that)
- Then we have 1-2 student leaders serving as project managers for each mechanism.
- Other students are assigned to these mechanism teams/pods/squads.
- Our controls/software team is also divided into groups.
- Weekly Design Recap Video Calls
- We moved these to Slack Huddles, but previously used Discord or Zoom.
- These calls are where our mechanism leads students to review 1 to 3 slides they create that describe the progress made on the mechanism, the tasks to be done in the coming week, and to discuss anything delaying progress, such as waiting on orders or other mechanisms.
- Mechanism Design daily update threads
- The mechanism leaders create weekly threads in Slack that students update daily with their progress, including what they completed.
- This makes creating the weekly slides and our build blog posts very quick.
- Robot Reports Channel to document things that break and need to be repaired.
- This channel is specifically for documenting anything that breaks or malfunctions on the robot. This way we communicate to everyone that there is a new potential problem or failure mode that we need to address with future improvements.
- Voucher Checklist 2026
- This isn’t really communication, but it’s helpful to keep the organization on the Kit of Parts vouchers.
Spectrum Robot Cart 3.0
We have produced the 3.0 version of our robot cart. It only has some minor upgrades over our 2.0 cart.
- The cart is 24 in x 36 in so it is legal to take onto an FRC field.
- The front and back are open so we can put a StackTech 3 drawer box on the cart and easily access the drawers.
- The underside of the robot is easily accessible for swerve maintenance, etc.
- Wooden blocks can be added to the cross rails for the robot to sit on.
- Onshape CAD
Allen went to India
RFL (Robo Fun Lab), the parent organization of teams FRC#9692 Sigma & FRC#10428 KAIZEN, invited me to come to India from Dec 18th to 28th to help prepare the teams for this season. They flew me out and hosted me so I could help answer all their questions and speak with their students about the robot design process, the fundamentals of FIRST, and much more.
They were fantastic hosts, showing me around their cities, taking me to excellent food, and just learning from each other. Thank you, @Virendrasinh, for being a wonderful host, guide, andfor treating me to a lot of wonderful food.
We are planning to have weekly video calls to continue building this relationship.
Working on robot things
FTC Scrimmage
Sightseeing
Food
Shop Upgrades
X-Tool Metalfab
We preordered an xTool Metalfab 1200W in April, received it in early August, and built our off-season robot using it. It is wicked fast, and we have cut up to .25” aluminum with it. This will allow us to cut aluminum much faster than we could with our Wazer Pro or CNC router.
Here is our xTool Referral Link, if any other programs are interested in buying a MetalFab, I believe we both get something if you use our link: referral-program
Tapping Machine
We have replaced our pneumatic tapping machine with an electric one. This new model improves ease of use and enables automatic tapping. We also adjusted the clutch to further prevent tap breakage.
We purchased it from Amazon it was only $699 when we purchased it but it’s now listed at $900.
3D Printers
The 3D printing area has undergone significant improvements. The area’s physical configuration has been completely overhauled. The new, optimized layout is designed for maximum efficiency, allowing for substantially easier access to all printers and auxiliary equipment for people not as tall as Allen. We have substantially enhanced our hardware resources with the introduction of the new Bambu H2D high-performance 3D printer. This acquisition is critical as it expands our in-house manufacturing capabilities to include a much broader range of advanced engineering-grade filaments. Specifically, the H2D’s high-temperature capabilities and robust design now enable us to successfully process sophisticated materials like PPS-CF which give us more options in our exotic material 3d Printing.
Powder Coating Booths
We’ve put together a custom all-in-one powder coating station that’s been a total game-changer for our workflow. We started with a deep-profile shelving unit that fits perfectly in our shop space and built out a plywood enclosure to keep the mess contained. For filtration, we went the DIY route by mounting and sealing a standard box fan to the back to act as a homemade dust filter. We also rigged up some super bright LED strips from Amazon so we can actually see what we’re doing, and we’ve got a regulator hooked up to our 120 PSI house air to keep the spray nice and consistent. With the booth, our curing oven, and new safety gear all in one spot, we’re ready to churn out clean, professional-looking parts all season long.
Spectrum Small Parts Organizers
This fall, we have begun reorganizing all of our small parts cases and many of our fasteners, etc.
The full write-up can be viewed in its own post: Spectrum Small Parts Organizer
StackTech
We have started transitioning some of our toolboxes to StackTech, but they haven’t released the new drawer modules planned for this summer.
No More CNC Router
We recently got rid of our CNC router to make room for a new machine. (We donated it to our school’s theater program.) We used it once during the 2025 season, and only when we needed more part output than our Wazer Pro alone could produce within the required timeframe. With the X-Tool, we shouldn’t have those issues again. More to come soon…
, yay
