599 The Robodox / Build Thread 2024

FIRST FRC Open Alliance
,
1

doxoa2023
doxoa20231920×720 56.9 KB

Welcome back to our 2024 build thread! It’s a pleasure to have you. or, well, it should probably be the other way around, but who’s counting

We’re a student-led team based out of Granada Hills Charter High School in Southern California.

This year, we have 42 students, 15 of which are rookies.

Links:

Website
Github - both CAD and code
Smugmug
2023 Thread

2024 Regionals:

  • Los Angeles Regional (week 3)
  • Aerospace Valley Regional (week 6)
  • and really, this is just a pipe dream of mine, but Hueneme Port Regional? :pleading_face:

We’re looking forward to a great season with everyone!

- Ben

23 Likes
2

Looking forward to this thread! Would be cool to see you guys in hueneme if you manage to make it!

4 Likes
3

Let’s get this thing started, shall we?

Pre-season Preparation

There’s still about six weeks left until the end of our fall semester and when i can finally get some sleep.

We’ve got a lot of stuff planned; I’ll let the other leads say what they’re working on…

CAD

Rookie training is progressing well, if slightly slow. I’m having the rookies each work individually on projects to learn our workflow and get experience with creating subsystems. Right now they’re working on a west coast drive; after they’re done or before, we’ll see we’re gonna try the 2485 cadathon (water game be gonna be really funny). After that, it’ll probably be winter, and then it’s F4 CADathon time~

Mechanical

All mechanical members right now are being trained, with two people assigned to each machine in our shop.

Our plan is to go through parts that gradually increase in difficulty over the rest of the year, in hopes of developing individual member autonomy for build season. :crossed_fingers:

Electrical

All of the electrical members have been learning various electrical engineering concepts and theory that relate to our work in the shop.

First we started with Ohm’s law (relationships between power, voltage, current, and resistance) and basics in power systems (introducing components like resistors, capacitors, diodes, etc and where/why we use them).

After that, we taught testing with multimeters to practice efficient debugging practices and increase the division’s overall knowledge on electrical systems. We then dabbled into simple relay circuits to give further understanding of current/power control, and started integrating schematics diagrams in our circuit workshops.

We are planning to go further into systems understanding and use our learning in debugging/systems drills that members will perform on our 2023 robots which will undergo in the following weeks.

Programming

Programming is breaking up into small groups to work on separate projects. Most importantly, we’re switching to Java for 2024. Due to a dearth of resources in C++ for many of the projects we’re trying to achieve, we thought switching to Java would be most helpful for us.

To that end, we’re working on translating our build season robot’s code into Java, and fixing our offseason robot’s code (also in Java).

We are also working on developing a proper vision solution using PhotonVision and the BeagleBone Black coprocessor.

Strategy

Strategy is focusing on analyzing previous FRC games in order to ensure that Strategy is familiar with game analysis for the 2024 season. Currently, we’re wrapping up reviewing 2017 Steamworks, and are moving on to 2016 Stronghold to make sure that each member knows a shooter and pick and place game.

Additionally, with CAD participating in 2485’s CADathon, Strategy will also work with them to analyze the CADathon game.

We’re also troubleshooting our new scouting app so it can handle the large amount of data input from scouting simultaneously. We were not able to complete this testing before Tidal Tumble, so we are doing so now.

Written by Ben, Rianne, Rae, Ayush, and Leah

Cadathon post soonTM - the kickoff is in a day anyways, after all

6 Likes
4

Looking forward to reading this thread! Interesting to the various focuses of each department of your team. Especially excited to see all electrical-related updates, it’s nice that you are going into the different systems. Also loved seeing the process with the whiteboard in the picture!

1 Like
5

Kickoff

What an unexpected game huh?

On another note, we appeared in some of the FIRST videos and that is pretty hype (those labcoats are paying off).

For day 1 of the 2024 build season, we read the game manual, did the 1678 rules test, and discussed possible strategies. If you are curious about our game analysis document, here is the link:

It is very similar to the one we used last year with some minor tweaks to improve our focus on designs.

With the many field elements this year, we started manufacturing and preparing the wood necessary to build the elements. We hope to finish it soon to start testing and brainstorming ideas for our designs.

Priority List

Leading us to discuss our priority list. Inspired heavily by top teams, like 3476 Code Orange, we decided to create a priority list to first focus on the strategy portion of the game and our approach to the various aspects. Following that, we would then discuss designs and how we will achieve the needs/wants of our team.

Here is the list in its entirety:

Some things we want to point out:

  • Ground intake: we decided towards prioritizing a ground intake over getting the pieces off the shelves because of what we observed from last year’s game, Charged Up. Teams were able to have lower cycle times from being able to grab the pieces off the ground and also could have strong multi-piece autos, leading us to prioritize it this year (not making our mistake last year).
  • Size of robot: With the length of chain, we are considering making our robot smaller to gain the Harmony bonus and possibly the Ensemble RP. It may be similar in size to our 2023 Off-season bot, Barry, to achieve this goal.
  • Buddy climb - with only a one foot extension, and much of that taken by bumpers, a buddy climb without cheesecaking seems basically impossible.
  • Like many other teams, we see the traps as… a trap. It seems valuable (and it is) but there’s so much work for it that it doesn’t quite pay off.

Potential Strategies

Our team spent most of our time determining what a strategy and ideal alliance may look like at competitions. Here are some important factors we considered:

  • Defense: With the need to travel across the field and the large Stages in the way, it’s likely that zone defense could be strong in slowing cycle times.
  • Feeder-Shooter alliance: To reduce cycle times for our alliance, we considered the possibility of one robot getting the Notes at the Human Player station, exiting the opponent’s Wing, and firing it across to another robot to score in the AMP or Speaker. We think this may be strong enough to get the 18 cycles (or 15 with cooperition) rather than having robots individually cross the field.
    • Potentially, shooting from half field could also be useful for this, if the shot misses the cycle is not wasted, the piece is just fed.
  • Endgame during Elims: From our math, the endgame may be more fruitful through prioritizing getting more cycles and amplifying the speaker compared to a triple climb with traps (traps may be a trap too). This is still under consideration given the unknowns in cycle time and the skill of human players landing Notes on the Microphone.

Game Comparisons

These are some games we found similarity to in gameplay or robot design:

  • Ultimate Ascent - 2013
    • Game piece is similar and designs may be similar to it
  • Steamworks - 2017
    • The non-symmetric field and area of retrieving pieces is similar, and the stages’ placement on the field make similar blindspots
  • Power Up - 2018
    • The Amplify at the AMP serves a similar purpose to the Power Ups
  • Infinite Recharge - 2020/2021
    • The speakers and scoring method can lead to very similar cycles
  • VEX: Spin Up - 2022/2023
    • Similar to Ultimate Ascent, the usage of discs can lead to possible similar designs

What’s Next?

Continuing our discussions of designs, we plan to brainstorm and finalize a general idea of our robot tomorrow to start CADing.

We are getting amped up here, and hopefully we can orchestrate something together!

Post written by @kungfuwisdom, with just a but by me. A more design focused post coming in a bit~

Inspired by 3847 Spectrum’s Posts

4 Likes
6

Day 2 + 3

This game is so hard to figure out. Especially the trap, and we know it may be a trap, but we are also delusional for these types of difficult elements to figure out.

For these past two days, we have had design talks about what direction we want to head towards based on our wants and needs list we made on Day 1. The main difficulty is trying to stay low and small while making our subsystems fast and efficient, leading to considerations of double-use subsystems and such.

We also have been building our field elements and just planning for the season ahead, hopefully we finish those this week.

Designs

hey look it’s our repo

krayonCAD below of our possible idea


robot arch things~ more or less
 
 

more fleshed out geometry - climb not included because it totally won’t be copied from another team that makes tiny robots

I haven’t quite nailed down what I want from the shooter, but as of now, a “full wrap” on the game piece looks to be best. It’s quite unstable without spin, so we added side wheels for spin.

We fear that solely using side wheels for thrust will squish the game piece too much and introduce instability, so we have top and bottom rollers sandwiching the piece to impart most of the thrust.

drive - it’s not quite final, but good enough

26" square, as per usual. 2910’s 2022 bot was actually this size, but for some reason it looked so much smaller…

We’re running kraken drive and falcon swerve, so we can use more krakens elsewhere. Also, using inverted electronics again, to free up some space both for the ele and mechanisms. Wanted to have a “full” (covering motors) bellypan, but alas…


the boolean is false.

Unrelated problem: can’t cut one piece bellypan on the omio, which isn’t ideal. thinking about something funni™ like jank friction stir welding to make it one piece…

We’re also considering using some of the 16 tooth mk4i pinions for extra speed, as it looks like we’ll be able to utilize it.

…sounding a lot like some other team right now, aren’t we…

I really wonder what the powerhouses are doing……

The trap really sucks, to be honest. It’s so incredibly useful for the rp, but at the same time feels nearly impossible to do. Our current plan at the moment is to outsource watch week zeros for other teams’ ideas.

We’re planning to prototype shooter and intake tomorrow, we’ll return with our findings.

Post written by ben, lede and a bit by @kungfuwisdom

16 Likes
7

Days 3 and “4”

school started for us, so no more 2am cadding and now much more suffering. short one today, for much the same reason.

Our field elements are nearing completion, we’re waiting on some parts to come in and some wood to get cut.

We couldn’t get to prototyping the shooter, but we did get a bit of intake:

As it turns out, our old geometry didn’t quite work, we ended up raising the shooter by .25", ending in a pretty consistent yoink. Going to prototype some new geometry hopefully tomorrow as well.

Also, the note isn’t as flexible as we thought it would be, so we will have to adjust a lot of our indexing to make the transitions a bit smoother. We have also decided to mount our bumpers extremely low to aid in consistent and fast intaking.

new (untested) geo:

This does contrast somewhat with many other teams’ findings, but we think this has something to do with us applying less power and having less axle density. Also, we did notice some of the same things we were seeing in other videos - very high resistance at bends and a significant slowdown or power spike, which may end in lower consistency. Since we plan on using our intake also to score in the amp (preliminary testing suggests it is possible), note handling consistency is a high priority for us.

Our plan for the shooter turned out to be incredibly difficult to prototype and make even slightly three-dimensional. Imagine this pipe dream, but, superglue…(this is going to suck, might even need to do two parts)


Hopefully we get some more stuff tomorrow~

I forgot how computer aided despair cad could be :waaaaah:
this robot, i swear… exceedingly difficult to package, conceptualize in three dimensions, multiple things happening at once, it just sucks guys send help or validation, we don’t want to miss meta like we did last year

crying about it

post written by ben

8 Likes
8

hey, it’s been a bit (and this content specifically is a bit overdue but oh well)

the dream is dead

with the dramatics out of the way - i’ll explain.

Remember this? (ignore the old geo)

high arc shot into the amp? we tested it; it’s highly inconsistent and thus we can’t depend on it.

So with the amp scoring half of our robot invalidated, it was back to the drawing board for us.

|700.8265666312626x526.0000000000001

Above are the goals for what we wanted to do with note scoring, below were any possible ways to accomplish that.

First thought was something like this - shooter on elevator, to shoot down into the amp and also collapse to score into the speaker. The climb gets eliminated and placed onto the elevator.

It was quite complex, and we didn’t want to spend precious time with an overly complex solution… so we scrapped it.

Looking at some other designs, though, we quickly decided that adding a ri3d bar would minimize the amount of changes we needed to make, as well as potentially being able to do trap with some modifications.

(trap is still low priority for us, we want to score first)

Eventually we settled on this:

It’s an ri3d bar that extends out via a 254-style linear extension.

Hopefully this allows us greater consistency when scoring into the amp. We also hope that this can score in trap, but that’s a low priority right now.

Design post coming soon (i am struggling) (how do you shooter i am getting concerned for our viability)
-ben, which is probably obvious because of how unhinged this post is

7 Likes
9

solidworks crash counter: 5 (3 on 01/18)

sleep paralysis demon git issue counter: 3

Integration is going to suck.

Week 1

mmm, CAD deadline at the end of the week~

Honestly though, we’re further ahead comparatively to this point in the season last year. The game is much more complex, and our robot is actually pretty far along, especially in two dimensions.

We’re quite far from done, but we hope you’ll be able to see the dream.

Packaging has been difficult - the game piece is wide, and the chassis is comparatively very narrow.

Drive

20fps swerve and inverted electronics, and a bellypan that only crashed solidworks once!

We’re planning on revising our swerve covers to cover the cancoder wires, so if you’re interested in that, please wait warmly~

Intake

We took a lot of inspiration from 4414 on this one. We’re probably going to have to suffer through putting silicone on rollers again… but it should be better this time (we hope)

Indexer

Our first indexer revision if you can even call it that is considered an Unmentionable™. This is the second version, which is significantly less cancer:

(for some reason, solidworks ctd’d three times working on this. i think it’s retribution)

Shooter

we are about as decisive as a metronome, but we’ve finally settled on a 3255-style shooter with kicker wheels to feed from the indexer.

This is probably the least complete mechanism but it’s getting there it has to

Climb

Do ignore the offset clamping plates at the top

but this is our nearest to “complete” subsystem.

I wonder who we took inspiration from, hmm?

s t i c k

as a late addition (and the application of washed awards cadder) this one is ah… going. Somehow? It’ll work out eventually we swear

Post written by ben

CAD brought to you by commits at 4 am ahem @Leahthedragonite

this year in a nutshell:
image

15 Likes
10

I was lucky to witness the Solidworks crash counter go up by two in the span of no less than 5 minutes
(My first CD reply/post :star_struck: )

2 Likes
11

We’re hoping so. We’re planning on a linear rack intake!

2 Likes
12

Can’t wait to see more of this!

4 Likes
13

new pfp who dis~

been a while. Hasn’t been too much to post about - we don’t have the bandwidth for an alpha. Also, we’re behind!

how is it week four already h o w is it w e e k f o u r alreadyyyyyyy

It’s mostly just grinding cad, now. This highk really really sucks

omio crash: 1

solidworks crash counter: 32

For some better news: drive (and some other parts)

and a mostly finished intake:

The good half of our robot:

now for :waaaaaah:

Shooter is a box of pain and regret and we’re probably going to change the design again but for now we just need to get stuff moving.

and the stick is proving to be much harder to package than we originally expected.

It’s gotten to be so bad that we might just change the architecture. It’s week four. This game sucks.

Send help.

Please.

Much of our parts cannot be finalized because the bot is integrated so tightly (subsystems depend on other subsystems for mounting and etc) and this makes CAD deadline misses have a very significant cascade onto everything else down the chain.

hoping for a bit of engagement here - shooter suggestions? help? alternative designs? aside from adding more metal because it’s already an overbuilt massive nightmare

post written by ben

distilled suffering quarantined within our shop

10 Likes
14

Remember when I said we’re rearchitectureing?

I did not sleep tonight.

Heavily inspired by 111, and slightly by 95; thanks guys

still in very much development; i’m going to sleep now though… hopefully you can see the dream.

We’ll see tomorrow if we do go through with this.

Post written by ben
distilled suffering has broken containment

16 Likes
15

long overdue post

here is your OA solidworks crash counter: 60

KRAKENS

after many weeks, we received our Kraken order. thank you to WCP for this product, we are excited to test and compete with this motor!

CAD

as we finished up the CAD for our robot, we realized it was about to be overweight considering the electronics and wiring, and how it was also not very accommodating for the battery

so we aren’t done with the CAD just yet.

this bot will end up being our “beta” bot for the season as we just want to ensure that we have at least something in these coming weeks.

with this now in consideration, we made a large modification rearchitecture for our robot, leaning more towards a 3847 architecture as it helps us find a more consistent amper scorer and climb without compromising our weight. our shooter and intake are undergoing minor changes so we don’t lose as much on time.

here’s what we have for now

in solidworks

and krayoncad

Mechanical

we’ve been makin parts!! we finished rails for our drive train a while ago and now it’s been shafts and polycarb tubes :+1:

Electrical

Prepared all new comp batteries and been preparing wire for drive. Mostly took time to train skills, learn more about electronics, and side projects with non-FRC circuits.

Programming

more focus on just testing basic concepts of subsystems on test boards we have in the shop while utilizing Carter to test drive and vision.

here’s a video of our auto-align testing using odometry to what the robot thinks is the speaker

see you when we see you again! as always, ask any questions below

post written by @RisenFalcon , Rianne, Rae, @kungfuwisdom

9 Likes
16

CAD looks really cool! I know you’re already into the build process, but I have some questions/suggestions about what I’ve seen

  • If you’re looking for ways to save weight, try dropping some of the polycarbonate rollers. The amount you have looks like it’s way overkill.
  • I recommend incorporating some chain tensioning system for the intake’s pivot mechanism. It can be as simple as a slot cut out which can fit a delrin rod. It doesn’t look like you have enough range of motion to put a spartan chain tensioner in there, but I might be wrong, depending on how much it has to turn. It couldn’t hurt to check if you haven’t already.
  • Is there a reason the indexer plate isn’t polycarbonate? This looks like a part that doesn’t require the strength/weight of aluminum. Maybe you haven’t pocketed yet, which is a good design principle to not prematurely pocket things. But pocketed aluminum is still usually slightly heavier than unpocketed polycarb.
  • What role do the vertical posts serve? Is there a reason you can’t attach what you’re attaching on them onto the climb to save the weight up top of the posts?
  • Your robot is dense. You probably still have space for bumper mounts, but just remember to account for them.
    • I understand that the statement “just make the robot less crowded” isn’t helpful, but believe me, every small thing (such as removing rollers, or simplifying transmissions) will make you/whoever has to maintain this thing’s life much easier.
    • While your intake looks solid, kudos to you for making it easy to replace. It looks like you just have to unbolt the polycarb plate from the sprocket, and rebolt the replacement on. Good choice switching from pocketed aluminum to polycarb. We used pocketed 1/4" aluminum for our devbot, and while it handled ok it did get a little bent through abuse.
  • I think the robot rearchitecturing was a good choice, all things considered. I’m sorry that you folks have had trouble meeting CAD deadlines. 1622 has been suffering from similar problems, albeit we’ve mostly overcome them.
    • I don’t know if you’re a senior or a junior, but I recommend looking into reading books/essays such as The Mythical Man-Month by Fred Brooks to get insight into how to make it so you can better manage your team and work with what you have. I’ve found writing like this very helpful as a leader of a CAD team.
3 Likes
17

the gap between each post gets longer and longer

we lost track of our solidworks crashes, its up in the 80s right now probably

robot progress

in the past couple of weeks, we have been building our alpha bot which consists of the shooter, indexer, and intake. given the timeframe, we decided on the CAD below to be our competition bot just for our first regional:

here is our assembly progress so far to allow some testing to occur:

|657x874.8274359420639

intake and the limelight tower will be going onto the robot in the upcoming days due to some issues with incorrect belts/incorrect pulley distances. we opted to not include our climb for this upcoming regional given the lack of materials present, so we hope that we can present a stronger case in the other aspects of the game to maintain a strong strategic position still.

here is a video of our robot serializing a note for the first time (courtesy to 696 for letting us practice at their field):

(video)

testing

with indexer being one of our first subsystems being built, we have done some testing to ensure that the note could pass through. courtesy of our team’s testing, we ran the test between a good and bad note:

running the note through the indexer 10 times at different speeds, we started to see the success rates of each variation. realizing the friction in the indexer was too high especially with the horizontal compression, we added teflon tape to the polycarbonate pieces to make the travel more smooth which allowed for the most successful tests as seen below:

here is a video of the indexer with beam breaks included to allow for the stop:

(video)

we will continue to test indexer and the other subsystems in a similar fashion to make sure our notes are getting the best path and generate the best shots.

obstacles

as we assembled and started testing our robot/subsystems, we faced two main issues:

  • belts/pulleys were the wrong ones or incorrectly spaced or not in stock
    • addressed by either:
      • changing distances between pulleys
      • printing different sized pulleys to fit
      • temporarily tightening the belts using zipties/bearings
      • ordering correct belts/pulleys
  • possible belt skipping in the indexer
    • rollers may not be concentric
    • plan:
      • straighten out rollers
      • possibly loosening mounting bolts for tubes to decrease compression on the overall tube

we plan to continue testing in all subsystems to see if these issues are present elsewhere and hope to find a durable solution for them as well before we compete soon.

CAD: simplicity v. proficiency

as we make our alpha bot and continue to watch other teams at competitions, we have wondered about the direction we should commit for our changes given the timeframe and proficiency. our efforts in our omega bot may lead us to utilize those changes to make the alpha bot simpler and more proficient, focusing on gaining more practice between our regionals, rather than attempting a large change in architecture.

it draws upon Karthik, a mentor from 1114 Simbotics, states in his annual Worlds presentation that the two golden rules for design are:

  • Golden Rule #1: Build within your team’s limits
  • Golden Rule #2: If your team has 30 units of robot and functions have a maximum of 10 units, better to have 3 functions at 10 instead of 5 at 6.

this decision is still going to await further discussion inside the team, but we hope we can maintain our wants/needs list from early in the season while not compromising our success in the small timeframe we have.

ex. simpler intake from omega (courtesy of @RisenFalcon ):

upcoming

with our first competition being the Los Angeles Regional, we will be spending this next week finishing up the robot, testing, and preparing for the competition.

hopefully robot goes zoom as time is not our friend. as always, ask any questions below and we will do our best to answer

post written by @kungfuwisdom

6 Likes
18

whoops, its been a season

lets run it back a bit

Los Angeles Regional Recap

Results: Ended 5th/6th in Elims

From 3/14 to 3/17, our team competed at the Los Angeles Regional. We would like to first thank our alliance partners: 4201, The Vitruvian Bots, and 1515, MorTorq.

Our team competed with the main season robot, Luna, and ended with a 3-7 record after all our qualification matches with us ranked 38th. We ended up being picked by the 5th alliance and would go through playoffs with a 1-2 record, ending our competition at 5th/6th. In addition, we were awarded the Team Spirit Award, recognizing our enthusiasm in and outside of the competition.

Failures and Fixes:

As our first competition, there were a good amount of failures as we worked to understand this robot.

Failure: The intake gearbox belt distances were incorrect, leading to the shearing of a 3D printed pulley.
Fix: Asked around for a longer belt and utilized a larger pulley inbetween to create the proper tension for the belt teeth to still grip onto the pulleys.


(above is an example of what we sort of did to deal with the issue, courtesy of 3476)

Failure: Intaking the note could lead to jams in the intake.
Fix: Re-picking up the note when returning back to the speaker OR utilizing the note as a possible amp cycle.

Failure: An indexer 3D printed pulley melted and sheared after excessive use (shown below).
Fix: Not really a fix, but couldn’t fix it in time for playoffs which limited us to just playing defense.

Failure: Gyro direction after autos on certain side of the field would be zeroed at a 45 degree angle and mess with driving as it couldn’t be reset.
Fix: Utilizing an inverse gyro button to press right after autos based on which side of the field we were positioned before any huge driving was done.

Failure: Difficulty in integrating varying shooter angles for farther shots into the speaker.
Fix: Limiting the movement of the shooter into one angle that allows us to shoot into the speaker from the subwoofer.

Here is one of our first successful shots (you can see how the gyro issue affected our driving here)

Upcoming:


Above is a before and after of our robot, you can see the number of zip ties we used to keep the shooter and intake stowed to ensure we could play the most rough defense ever (it worked!)

With the result at LAR, we returned to our shop focusing on creating a stronger indexer and intake to allow for more predictable intakes of notes and storage before our shots. In addition, we hoped that we could increase our shot distance and hopefully be more versatile on the field.

Here is an example of the shot we were testing right after we returned from LAR:

i’ll be writing as many posts in these next couple of days to catch it up to where we ended our season, and then it’s off to the new leadership team for off-season

post written by will

3 Likes
19

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.