5013 the Trobots are happy to return to the Open Alliance for 2024 Crescendo. We look forward to sharing and collaborating with the whole FRC community.
Commitment device to produce better documentation and keep our CAD up to date.
Peer review. Sometimes we have bad ideas, the faster we realize it the better.
Visibility. We want to get our name out and get known better outside of our region.
We think it helped us last year.
** We won Excellence in engineering partially be we had better documentation
**Thinking about what we publish made us more cautions about our decisions.
Over the summer we focused on outreach and STEM education.
Throughout the whole month of June 2023, the Trobots assisted around 50 middle schoolers that were participating in a STEM Camp at our district’s LEAD Innovation Studio. Throughout the camp, we helped provide programming assistance in Apple’s Swift language, supported the construction of 15+ robots, and refereed many robot matches between the students at the camp.
On July 15, 2023, the Trobots partnered with Zona Rosa to present a STEM booth at their Summer Fest. In order to capture the attention of the event’s attendees, we set our autonomous LEGO Robot Snake on the ground, which spun around in circles until someone approached, at which point it stopped and “looked around’’ for what had stopped it. Many people were very intrigued by this snake-like behavior!
On August 4th, 2023, and August 5th, 2023, the Trobots hosted two STEM camps with 20+ total attendees. The first camp on August 4th had Girl Scouts earning their robotics badge, while the second camp on August 5th had a general population of middle schoolers who learned about robotics.
Mechanical
Built a elevator using a Thrifty Bot kit.
Worked on designing and printing TPU wheels for the SDS MK4i. @DChilson will add a post about that in the future after we see how they perform at Cowtown Throwdown.
Programming
Programming is researching upgrading to Phoenix Pro to allow for FOC and more modern code structures.
Still needs to program the elevator mechanical built.
Working on training new programmers with FRC Ladder
The team preordered 10 WCP Kraken motors. We are excited to see how these work out for drive motors in our swerve modules. We have also ordered additional swerve modules hoping for a practice bot for programming and drivers.
This last weekend 5013 The Park Hill Trobots had 2 major events. Our fund raising craft fair and Cowtown Throwdown. The unfortunate collision was due to Cowtown announcing 1 date then changing after we had already locked in our craft fair date.
Cowtown
In Cowtown we drove new 3d printed wheels based on @Joe_Johnson’s design but modified for for FDM printers. WE printed the hub out of Matterhacker NylonX and the tire out of Polymaker Polyflex TPU on a Bambu X-1 Carbon. The wheels held up very well and made the bump easier as well as having extraordinary grip on the carpet. We do have one more geometry change to do on the hub before @DChilson will publish the final cad and printing instructions. He will also publish a comparison of new wheels verses the wheels we practiced some on then drove all of Cowtown.
Between ACT’s and the craft fair our Cowtown crew was light and our normal drive team wasn’t there. We did not perform as per our usual but did get some time developing drivers in a competition environment.
We didn’t get our setpoints fully reacquired after arm maintenance that in addition to different tires made our autos not as reliable as before. The reason we weren’t ready comes down to an odd coms issue. In shop direct connecting to the radio we were having occasional coms losses. We spend a lot of time checking connections and replacing components. We replaced the Brainbox switch, the radio, the REV Radio power module., and multiple ethernet cables. We thought we had it beat as we seemed to have stable coms. When we got to competition and had the radio programmed we could not connect via ethernet from the drivers station. We went though all the individual components and tried a different radio and reprogramming the ordinal radio to no avail. The laptop would connect while going directly to the Roborio though. It was perplexing. The CSA went through all the steps we had already gone through. Then @blturner came by and spent some time poking around the drivers station. It seems somehow our drivers station was set up to be a static IP but had none assigned. We don’t know how this happened.
We ended up fairly low rank before the playoff selection. 1987 The Broncobots came by our pit to confirm that our driver was going to be back for playoffs. We assured them that our driver (@JoeyK) would be back. During alliance selection we were chosen by 1987. We ended up going through the winners bracket undefeated until finals where we lost the first game due to a red card by 1802 due to contact inside the frame perimeter ripping the wiring harness off of 3824s bot. Without the red card that match would have been difficult to win, as the 1987 was unable to field their bot and we suffered a drivetrain issue cause by a CAN wire being cut by our battery.
The next round of finals were a reversal and without 3824 on the other alliance we had a lopsided victory achieving the tournament high score. Then after a monumental effort by all 4 teams on the other alliance 3824 was able to field for the final match of the tournament. The game was close but we won 133 to 128. This is our first Cowtown championship and we want to thank our partners 1987, 1802, and 5126 for this alliance, all contributed and gave their all. We would also like to congratulate 3284, 1710, 1730, and 5119 for the best match of the whole tournament.
Craft Fair
The craft fair had a disappointing turn out. A lot of Trobot parents who may have attended were at Cowtown. However the chili that was served was a resounding success. The good thing that came out of the attendance was we have some fired up students who want to do more and better PR.
Starting Tuesday we will dissemble Goblin so that its greatness may live on in the COTs parts that are transferred to future bots. May we have a moment of silence.
Quick question for you guys on the programming side of things. I’ve seen talk about FOC with Phoenix Pro, but what does this affect in the code? Is this to do with swerve, or something different?
We haven’t used PhoenixPro yet but looking in the documentation it appears that enable FOC is a flag on a control mode. So if we were using voltage control we could use the enable FOC flag to use FOC.
In this case Field Oriented Control is about the magnetic fields in the motor not the game field. We would use it where we think the additional power is needed. Most likely swerve drive motor and climbers. Though it does bring down top free speed by a little bit it increases torque by up to 15%.
We do also use field oriented control (game field) for our swerve drive. That involves using the gyro to convert driver input to field angle . We do that in our GamepadDrive command class. Our Gamepad drive also uses gas pedal approach where the stick does not provide the velocity but the trigger does to separate direction control for speed to allow for finer resolution on direction at low speeds. This complicates the GamepadDrive with a little (not too much) trig.
Trobots participated in our local offseason event the Cowtown Throw Down and ran our 3d printed wheels. Our driver reported a significant improvement in overall performance across the field and we still manually balanced fine on the charging station. Our auto balance routine had issues but we aren’t sure if it had to do with poor wheel performance on the station or the fact that we never re-tuned the routine for the new wheels. We especially noticed a difference when crossing the cable cover. When we cross it with these wheels, it performs almost like we have a suspension. It’s a much smoother hop over the cover and drivers reported we were knocked less of course.
Cowtown is a shorter event with only 7 qualification matches. We were initially planning to change wheels in between qualification and elimination matches but upon inspection decided to stay on the same set. We ended with a total of 13 matches driven on the set shown below. The nubs are slightly worn down compared to a new wheel, but it hasn’t seemed to decrease the overall grip of the wheel. You can see in the image that the leaf springs on the outer edges are starting to separate and there is noticeably more squish to the tire. This squishiness does not carry over past the outer couple of layers, but it’s something we need to keep an eye on. Only 2 of the 4 wheels saw this separation and it was in opposing corners of the robot, so I wasn’t able to attribute it to weight distribution or anything else on the field. We did use 2 different printers to produce the wheels and did not keep track of which wheels were installed where, so that could possibly explain it.
The leaf spring separation is visible here. It does not carry beyond the outer 2 layers of leaf springs and only occurred on 2 of 4 wheels on opposing corners of the robot.
This is one of the wheels that did not experience leaf spring separation.
Conclusion: The wheels exceeded our expectations and we will be running these going forward. We intend to do a complete wheel change after the first day of qualifications as a precaution even if the wheels still appear serviceable. The wheel and hub set have a relatively low overall cost and it’s not worth being cheap on such an important part of the robot. We will still travel with a set of stock black nitrile wheels in the toolbox just in case something catastrophic occurs, at least for this coming season. We printed our hub and tire with an 85% gyroid infill and didn’t experience the same failures seen by several other teams. I support having as much weight as low in the robot as possible and I don’t see a benefit to going light and cheap on such an important part this low to the ground.
I will edit this thread to add a link to CAD. I experienced an issue and lost the most recent version of the files and a little re-work is now in order. I will also add print instructions for the tire itself as printing TPU was a learning experience for me and I’d like to prevent anyone else from spending the better part of an hour separating the tire from a build plate. That was a painful lesson.
Build season is fast upon us. Our schedule for this process is as follows. (Thanks to @MrKrabs for putting this together)
1/6
Doors Open 10:00am
Season Goals 10:30am
Kickoff Video 11:00am
Grilled Ham & Cheese for Lunch!
Read the Manual 12:00pm
Game Analysis Worksheet 1:00pm
Whole Group Game Analysis 2:00pm
Important Robot Actions 3:00pm
Winning Alliance Composition 4:00pm
End 5:00pm
1/7
Doors Open 1:00pm
Game Analysis Updates 1:15pm
What will win at OKC? 1:45pm
Required Robot Actions 2:00pm
Mechanism Show & Tell 3:00pm
What will integrate well? 4:00pm
So what are we building? 5:00pm
End 6:00pm
Our Robot CAD Link has been updated. Our robot is currently named Gene Eric Robot until it takes enough form and personality for the students to come up with a more permanent name. It has become a tradition for the mentors to give the robot a bad name at the start of the season to encourage the students to think of a good name.
Kickoff 2024.
Apologies for the lateness of this post.
We had a good kickoff meeting with some grilled cheese provided by @DChilson 's wife. After the video we got to work. This is an exciting game with many scoring opportunities , available strategies, and design challenges. It should be fast paced and fun to watch as well.
After much discussion the team came up with this list of robot priorities. Priorities
Necessary to Compete
Drive
Pickup and Hold Note Source
Score in Amp
Competent Human Player (Source)
Antonymous - Leave
To Be Competitive
Pickup Ground
Competent Human Player (Amp)
Speaker Score from Podium and Subwolfer
Score Note from Autonomous with multiple positions.
Under 20 second Climb
Harmony Climb
Vision Target Speaker
LED Status
Vision Note Recognition
Reach
< 10 second Cllimb
Shot Validation / Range (Pose Estimation + vision)
Trap (Appropriately named)
Center Line Autonomous Pickup
Frisbee Golf Master Human player
On The Fly Trajectories
Turn To Target Auto Shoot
The students played the game in roller chairs to get an idea for cycle routes and pinch points. Video has yet to be posted to our media area. I will share when it is available to me.
Some prototype/material testing took place as well.
Students were sent home with the homework to think about how to accomplish the above. Inspiration to be found in locations listed below.
Sunday we returned with new energy and with robot ideas. Some preliminary cad has started to take shape. Gene Eric Robot may not stay a generic robot too long.
One idea we came up with was an intake with over the top handoff that also become the shooter feeder. It seems it can also be used for Amp placement with a flip and eject. We called it the Flippin’ Intake we were happy to see later that Cranberry Alarm came up with a similar solution.
We are leaning strongly toward a vertically actuated shooter . We initially thought Rev Linear Actuator but it was sold out. Then remember the Playing with Fusion arm from last year. Also sold out. Though with our new DRO drill mill we thing we can make that happen.
For the climber we will probably fall back to the old reliable Thrifty telescoping arm. We were able to get those ordered.
We still have quite a few questions to answer
What range of motion you need for the shooter?
How far can one practically shoot because of the Speaker hood?
If you use the wrong wrench and round off a bolt is it just a rounding error?
How fast can we safely actuate the intake?
Programming is looking at their challenges. Many of the challenges are similar to what we have faced in the past . What is new that we have identified a a lynchpin piece for advanced robotic functionality is visual Note recognition and drive to target. For that we plan to use the Limelight with Google Coral edge. This will allow for more reliable autonomous pickup as well as faster in game ground pickup. If we add automated intake deploy and retraction to our pickup command we would minimize intake exposure.
A very exciting time and we are ready for the season.
The REV Linear actuator is a new product so it isn’t sold out, it just hasn’t been made available yet. They are making their way through QC checks and then will be available for purchase (very soon).
Our local weather has slowed progress, we’ve had extracurricular activities cancelled two days in a row preventing our typical build meetings. School today is already an AMI day, but we’re still hoping to be allowed extracurriculars this evening.
We’ve been working hard on the robot CAD with OnShape and it’s developing pretty well. Our initial sketches from Saturday and refined on Sunday have been fleshed out pretty well at this point. We’ve had several deviations, but better week 1 than week 5 when the robot still isn’t functioning. Our biggest change is abandoning the linear actuator for adjusting shooter angle. We tried a variety of designs and nothing would allow packaging in the space we allocated and still give the range of motion that we desire. We have ended up changing to a chain driven actuation at the base of the shooter which has ended up solving a number of other future problems with our electrical panel. This is the most developed our CAD has ever been this early in the season and it’s allowed us to find issues like this and deviate while it’s still a minor speed bump, not a lost week of manufacturing and assembly time.
Another challenge we’ve dealt with was our flippin’ intake would place the note in the shooter with too much control of the note while handing off to the shooter wheels. We had to extend the shooter several inches to space the shooter wheels far enough from the intake feed roller. This additional shooter length means our starting configuration will now need to be more vertical, but that’s no longer an issue because we already deviated from the linear actuator control. We will need to be mindful of our shooter height when we drive under the stage as it’s now possible to have it elevated high enough for a conflict, but programming has some ideas how to guard from that occurring.
Prototyping has taken a back step due to the lost build days but we’re doing the best we can. Our initial concept sketches and development have ended up looking and functioning similar to Ri3D team Cranberry Alarm so we’re matching their compression dimensions from their shooting testing. We did limited verification testing last night in my garage with good results, good enough that we will proceed with these compression dimensions until a better solution presents itself. Our intention is for prototyping to resume with regular build meetings but we aren’t expecting any major changes. As an aside, we would like to thank Cranberry Alarm for the excellent work they’ve done testing and documenting their build. It is reassuring to us that we’re on the right track when we discovered another team has come to similar solutions.
We assembled our programming practice bot last night in my garage and hope to have code loaded on it this evening and driving. This is our teams first ever dedicated practice bot, a monumental achievement for us. We are hoping that programming will be able to achieve many of our lofty goals by giving them actual time. They had approximately 2 build meetings worth of time with the robot last year before our first competition and it seems logical that more time is more better.
Our goal for tonight is to present the current state of the CAD to the entire team for feedback and final comments so we can enter into the manufacturing phase this weekend. We are also going to be going over electrical requirements and routings. Several of our mentors were able to participate in Team 1538-The Holy Cows electrical conference in December and we are planning to incorporate as many lessons learned into this years competition robot as possible. This will be the first time we’ve ever sat down to produce and review an electrical plan and we’re hoping actual forethought will prevent many of our typical headaches.
Below is our current to-do list and some images of the robot CAD as it exists now. The OnShape link should take you to the robot CAD, please message me if it doesn’t work properly.
-Electrical is a mess, will be resolved this evening (hopefully)
-Sensors need added based on electrical meeting
-Motor mounts, motors and gearboxes for shooter and intake need placed
-Chains and tensioner need located
-Climber arms need swapped to 2-stage and a hook added
Last night we made a list of needed powered components to plan wiring and verify we have enough power ports.
Motors
4 Drivetrain Steer - Falcon - 12awg
4 Drivetrain drive - Kraken - 10awg
2 Shooter wheel motors - Kraken - 10awg
1 shooter actuation - Falcon/Kraken - 12awg
1 intake actuation - Falcon/Kraken -12awg
1 intake roller - Falcon - 12awg
2 climber - Falcon - 12awg
1 amp retention bar actuation (maybe servo not full motor)
Other Powered devices
RIO - PDH Low output - 18awg
Rev RPM/ Radio - PDH Low Power - 18awg
Rev Mini Power Module PDH - 18awg/14awg
** 2 Limelights (1 port each) custom POE
** Ethernet Switch 18awg
** LED Strip - Through VRM 5v
** 4 CANCoders (serial off of 1 port) 18awg
** Playing with Fusion Time of Flight Sensor 18awg
That uses 17 of 20 PDH high output channels and 2 of 4 low output channels . Plenty of room to spare right?
There is still quite a bit of discussion about how to package the electronics in our spacious 24x24 frame. Also there is a little concern over the hex shaft power transfer for the launcher.
Programming
We have have a working swerve drivetrain for the practice bot we affectionally call 50Brick*. We went through the standard pains of bringing up a fresh can and IDing all motors, flashing firmware, realizing CANCoders were on 5v … you know all the standard things.
Then we used the Phoenix 6 Swerve Generator to bring up the drivetrain. It was the easiest swerve bring up we have had in the 3 years we have been doing swerve. We did pay for Phoenix Pro this year. Phoenix Pro combined with an all CTR drivetrain and the generated code allows for fused sensor settings and a very fast background odometry update.
Work continues. As always ask any questions you have we will try to get back to you as soon as we can.
(*) In the early days of the team we used old stencils to paint the number on the cart. Some students use the B stencil because it looked like a 13 . So we named the cart 50B. The practice drivetrain is without mechanisms and thus pretty much a brick so therefore 50Brick
1/14/2024
We had a good meeting on a cold day yesterday. Mechanical
After many discussions and arrangement attempts we have finally figured out how to package the electrical. It will require a vertical battery counter to our earlier hopes of a horizontal battery. We really thought it would be easier with our huge 24x24* frame. Our mechanical lead will post an more details in a coming post.
(*) Huge compared to last years 22x22. There is a whole 92 extra square inches.
Manufacturing of parts started. Our new lathe and DRO mill have made this so much better. This is the fastest pace we have moved that I can ever remember.
We did find we have access to a small material science oven that we could use for powder coating. If manufacturing goes as well as is is now we may have time for good a good powder coat.
Manufacturing of field elements continued. We have an amp, the speaker is nearly finished, and the microphone ring toss carnival game is planned. We don’t currently plan on making a stage as we really have no where to store it as we practice in the school library and we can’t store huge elements in there without upsetting a few people.
Programming
The seniors on our programming team have been on a self led journey into training tensor flow models for vision recognition. They started a training run on my laptop yesterday afternoon and it is still running at 4:00 am here… I will let it continue as it is generating heat and it is -10 here this morning.
The rest of the programmers worked on driver controls. They took changed from the basis stick control generated by the CTRE swerve generator and modified it to our team standard gas pedal and slow mode approach. This takes the velocity control off the sticks and puts it on a trigger axis. The benefit of this is for fine maneuvering you don’t have to keep the stick is such a small range. There is also a slow mode button that divides the speed by 3.
We do have the Limelights recognizing tags but but didn’t get to the turn to target code yet. We did notice that, as expected, the detection range on 36h11 was much less that 16h5. Any shooting that takes place from beyond the podium may have to rely more on pose estimation that direct target acquisition. We haven’t set up our pose estimation yet.
Our drivers did get new X-Box elite controllers this year. A huge step up from our Logitech F310s though not as economically replaceable if they get repeatedly dropped . When we first started testing the new controllers on the 50Brick practice bot we did run into nearly immediate emergency stops. We remembered we were using the pre kickoff 2024 game tools and when we upgraded to the new that problem went away.
We have already made a lot of progress, and we are only a week into build season! Our diligent team has completed 95% of our CAD for our Gene Eric Robot. The remaining CAD work is mostly tweaking details, such as the shooter and intake dimensions.
This year, we were able to buy a drill mill with DRO and a lathe with DRO. Before build season started, one of the mechanical mentors trained two of our mechanical leads on how to efficiently and safely use these machines. We have manufactured over 50% of our robot in the last few days. Last year, we used â…› inch polycarb, and discovered that it was too flimsy, so this year, our CNC team made intake plates out of ÂĽ inch polycarb.
Our practice robot 50brick is up and running, with two working limelights, a limelight 2 and 3. We used custom 3D-printed mounts, so programming can adjust the angle of the limelights as needed. Once we get the angle where we want them, we will 3D-print new mounts that are not adjustable.
This year, we have a specific electrical pan, and we are putting electrical on the top and bottom of this pan. This added a complication because we had to decide which items would go on top of the pan and which ones would go underneath it. For example, at first, the pigeon was under the pan, but we feared that it would become damaged if it fell. So we moved the pigeon on top of the pan, which provides greater stability.
After a lot of deliberation, we finalized the placement of our electrical panels. We mocked up several orientations that would provide enough space for the wiring and the battery in the laid-down position. We found an orientation that we thought would work. Once we added this design in CAD, however, we realized that the shooter bolts were in the way of the RIO. We had to move the RIO down, which meant we had to move the PDH down. This did not allow enough room for the battery. Finally, we set the battery in an upright position, which allowed all the components to fit. Another complication was fitting the electrical components around the arms of the climber. The climber took up more room than we anticipated. Finding enough space proved to be a challenge. We overcame this challenge by making a vertical electrical panel that attaches to our base electrical panel. This design provides more room for the necessary components.
The last week has been spent manufacturing parts for the robot non-stop, which means that we’re ready to start assembly on Saturday!
CNC
We made some improvements to our CNC, including the Shapeoko 3 HDZ which we installed in December, and some new .25in and .125in endmills that arrived yesterday. As Claire mentioned above, we cut some intake plates out of ¼ inch clear polycarbonate, but once the design is finalized we’ll cut them again out of ¼ inch smoked polycarbonate.
After that we attempted to cut the wood for our bumpers, but we had a couple problems so we had to put that on pause. Our goal is to get that done during tonight’s extended meeting so that bumpers can be assembled over the weekend.
Yesterday we cut out the plates for our shooter out of ÂĽ inch smoked polycarbonate, which was a nice test for our new endmills and also gave @JoeyK the chance to teach me how to do CAM.
Sponsor Manufacturing
This year we had access to a team member’s uncle who runs a metalworks company and offered to cut some parts for us. We had planned for a horizontal and vertical electrical panel to be cut out of 6061 aluminum, but they ran into issues with the sheet warping on the laser, so they upgraded us to whatever they had on hand (7075 or 7005 I believe) and cut it on the waterjet instead. We also had a ballast pan cut out of .25in A36 Steel, which was cut on the fiber laser. We’re extremely grateful to them for cutting these parts for us!
Powder Coating
We’ve been trying to find a powder coating solution since last year, but weren’t able to figure anything out until now. We noticed that the engineering teacher had a desktop lab oven that she didn’t use, and after asking she let us use it! After our powder coating supplies came in, we got the chance to test it out last night and the parts came out looking pretty nice. We hope to get all of the parts that we can powder coated during tonight’s extended meeting. Any parts that are too big to fit in the oven are going to be spray painted.
Our programming team have spent the last couple of days hard at work implementing our TensorFlow note recognition model that we created last week into a drive to note/Limelight target command on our practice bot. This command uses a linear interpolator based on the “ty” value that we get from our Limelight to determine the straight line “distance” value that our robot is away from the note, and uses the horizontal angle offset from our Limelight to the note as the angle we need to turn. Here is a link to our models that can be used by the Limelight.
We encountered some “skipping” of our drivetrain during the testing of this command, so we decided to reduce our electric current usage and change our PIDController to a Trapezoidal ProfiledPIDController. This approach ended up working out for us, and this command now runs much smoother and more efficiently.
Both of these videos of the robot are with the driver only holding the X button on their controller, which allows our drive team to focus on the game more than the intake process. We plan to add the automatic intake into LED flash parts of this command when our real 2024 robot is completed.
We have officially started week 3! This week we had a bunch of rather productive meetings, with a couple struggles. Our Mechanical team finished manufacturing their last parts on our Drill/Mill and Lathe, and snapped a handful of taps trying to make spacers. We began assembly of our drivetrain with our black powder coated tubes, MK4I Swerve, and a 7071 aluminum electrical panel.
We had no hiccups on mounting electrical to the panel, so we moved on to mounting our ThriftyBot telescoping arm kit which turned out to be quite the challenge. We still are not used to packaging this small of a robot since this is only our second year doing a small, custom drivetrain. We got a couple students to take a file to the climber arm slots, since the powder coat made the already hard to fit tubes, way harder to put in and take out. We then had our previous members of Pit Team, student and alumni, to see how easy we want it to come out in the event of arms being destroyed. The arms themselves were a challenge to be assembled, we had a couple holes be slightly off from what our bearing block heat set inserts were, and then as the stages extended and retracted they were scraping from aluminum shavings making a very unpleasant sound. To fix that, we disassembled and dry swabbed everything, which made it a thousand times better. After that, we began mounting them, and attach the gearboxes, that added another challenge. That challenge being the gearboxes assembled a couple hours earlier, not moving because the mounting heads screwed inside of the motor housing. It was caused by us once again forgetting to buy gearbox spacers, so we had to cut down some screws to solve the problem.
Our Shooter went smoother than our climber, but not perfect. It was being assembled at the same time as the entire robot, but took a lot of editing of parts on the lathe. Whenever we were designing the shooter, we didn’t accommodate for motor mount fasteners. We also had to remake a hex shaft, we cut it a little short and wouldn’t fit. One of our mentors decided to change our c-clip geometry, he didn’t like the fit that the feature script was auto-generating. As we were assembling the shooter, using Colsons instead of our Green Compliant Wheels, the shooter became a lot heavier quickly. We later weighed it at 17.8 lbs.
Although we could not get to properly mounting the shooter, but it is technically on the robot and that counts. Regardless, we still made it surprisingly far considering the amount of setbacks we faced during the 13 hours we had been working.
Lots of big updates this week! We have completed all of the field elements and have been running autos on 50Brick (our programming robot). We finished manufacturing last week and have been diligently working on our comp bot. Since our last update about mechanical, we have been working on assembling the intake and tweaking the spacing on the shooter.
After we started working on the intake, we realized that our back plate, which is designed to hold the note, was not designed properly in CAD. In CAD, the back plate was just floating in space. In the physical world, we used epoxy to secure it in place. We updated the design in CAD to add brackets. The epoxy is a temporary solution. We will cut new intake plates and add brackets soon.
Because we have a quarter-inch belly pan and an electrical pan, separated by a two-inch space, we wanted to give the electrical pan some stability. For the past two weeks, some of the students have used lathes to cut 26 two-inch spacers. All of these spacers had to be hand-tapped on both sides. Four broken taps later, we finished tapping the spacers on Saturday and added them to the electrical pan. We had to weave around these spacers while wiring the robot.
This year, we have been prioritizing the electrical wiring for the robot. We decided, instead of wiring little-by-little, to wire the entire robot all at once in order to more systematically organize the wires. We took a lot of our electrical inspiration from the Holy Cows. We spent about 18 hours wiring the robot, which is crazy. The hard work paid off in the end because all of the wires are zip-tied and secured to the belly-pan, so they would not move. We gave all of the wire enough slack for moving parts. The main downside is that the Kracken motors haven’t arrived yet (they will be arriving on Tuesday), so we have not been able to wire the drive train fully. But besides that minor inconvenience, the robot is fully operational!
We have made a lot of improvements to how we build our robot and put together the wiring. For example, we are using XT30, XT60, and XT90. I have been soldering all of the XT connections. In the past, we used Wagos and Andersons, but now we are using XTs because they are a stronger connection and more reliable. XT connections allow us to reuse and sustain electrical systems for future use. We did learn that using Flux makes a world of difference when soldering the connections. Another improvement we made was changing our battery connectors from SB 50 to SB 120, which is a far bigger connection. I made a mount for our SB 120 directly connected to the robot to make it easier to change the battery.
FRC Championship 
though not as economically replaceable if they get repeatedly dropped 
. When we first started testing the new controllers on the 50Brick practice bot we did run into nearly immediate emergency stops. We remembered we were using the pre kickoff 2024 game tools and when we upgraded to the new that problem went away.
