Spectrum 3847 - Build Blog 2020

Blog.Spectrum3847.org is host of our full blog. I’ll try to post of the content here as the season goes on.

2020 Day 1: Rules and Inspiration

Humans Play Infinite Recharge

We often have our team play the game out as if they were robots. It’s mostly just for fun and it reinforces some of the field locations and names of the field objects. We apologize for the audio, filmed quickly on a phone.

Robots that inspired us today

2006 Robots - 217

2012 Robots - 971, 341, 254, 33, 118

2016 Robots - 3476, 1241, 33 (hood)

2018 Robots - 118(forks), 148 (wrangler)

We were looking at various ball paths and intakes from 2012 & 2016 robots that we may be able to use ideas from this season. We were also looking at various ways to possibly lift a partner off the ground with our climb as well.

Basic MCC Discussion

Our initial MCC (Minimum Competitive Concept) thoughts,

  • low goal, human feed and simple floor intake through a bumper gap (over bumper is harder)

  • hook flip out style climb that deploys and then you can winch up. Multiple robots in 2018 did this.

  • no color wheel manipulation, it takes too many balls scored before this becomes valuable.



Day 2

Climbing is incredibly valuable

If you are thinking about which tasks on the field to do, climbing is the most beneficial after only being able to drive successfully around the field and over the small barriers.

  • At 25 points (20+ over just parking), it is worth ~7 balls in the inner goal or 20 balls in the low goal.
  • It can happen very quickly at the end of the match, leaving you time to do other teams like score balls, feed balls, & play defense.
  • It is points that only your robot can get. Your partners can pick up and score balls that you don’t, they might be able to use the control panel to score points, but the 20+ points for you being off the ground requires you to be there.
  • If you and one other person on, your alliance can climb and balance your alliance receives a ranking point.

The climb does not have to be difficult. This is the first time in some time where the only requirement for climb points is that your robot must be off the ground. (2013 was the last). If you are willing to be a tall robot, you can fold out an arm with a short (<8") pneumatic cylinder and use that to lift you 4"+ off the floor. There are also ways that you can climb be getting a hook on to the bar and using a winch to pull yourself. There are countless other options, but doing one of them successfully will make you much more likely to be picked for eliminations.

Control Panel Tasks are less valuable

In contrast to climbing, the control panel tasks are less valuable, and your team should strongly consider limiting the number of resources you devote to this task.

  • While there are a lot of points (35) associated with these tasks. Your alliance is required to score a lot of balls before these tasks ever become an option for you to get points. (29+ and 49+ balls) These ball counts will be rare in qualifications matches (and probably playoff matches as well) at many events around the world. You can be the best control panel robot in the world, but if you can’t score enough balls to charge the 2nd and 3rd phase, you’ll never be able to use your skill.
  • Even if your alliance is able to score enough balls, only one member of each alliance is needed to complete the tasks. Strongly consider where that robot should be your team, or if this is a task, you can leave to another member of your alliance.

Building Field Elements

We have started building our field elements. Working to make sure we can interact with our team elements in similar ways that we will be able to interact with the real field.

Image may contain: 1 person


Robots that inspired us today

2006 Robots - 111 (Rack & Pinion Hood)
2012 Robots - 33 (linkage Hood, CD7 Intake, and J shooter path), 118 (intake only has grip in the middle)

2013 Robots - 1986 (Climber), 254 (climber)

2016 Robots - 195(scissor lift), 2056(climber)

2017 Robots - 971(rack hood), 33(accelerator rollers), 5803 (ball path)

2018 Robots - 33 (double reverse 4-bar), 125 & 1323(buddy climbs), 2056 (forks)

2019 Robots - 3940 (floating roller on intake)

If you’re looking for some inspiration, here is a good place to see photos of some Aim High 2006 robots.
Pictures - http://www.firstrobotpics.com/?page_id=49

Sketch of the Day

A long robot (32-34") can store 5 balls in a line before entering a shooter.



FYI, a few of the images aren’t showing up. It appears that they are images copied from Google Photos, Google Drive, or Gmail (likely not publicly shared). This is happening on the Chief post as well as the blog on your website itself.

Here’s a screenshot of what I see on Chief, website:

1 Like

Hmm, thanks, google photos does that some times, I’ll get it fixed.

1 Like

Just out of curiosity… can the robot outlined in the sketch fit under the color wheel on the field? Did u think this would be a priority? Why or why not?

That robot does not fit under the trench (the dashed line at the top is 45in). If we had to start building a robot tomorrow Spectrum’s likely doesn’t go under the trench, that’s not to say in a few days we won’t change our minds. I’m sure I’ll miss some arguments but here are some off the top of my head.

The trench is narrow, only one robot at a time can go through, if you and a partern want that path at the same time in opposite directions its not gonna work

Climbing is easier if you don’t go under the trench since you don’t have to reach as far.

If you are short enough to go under the trench, the trajectory of your shots into the goal are less likely to go into the inner goal, assuming you are shooting at a high velocity. There are shots from lower angles that can go in, it will be easier if the ball is released from higher off the ground. This includes shots from the starting location without moving.

It’s harder for your shots to be blocked if the ball is released higher. You can make your shot harder to be blocked in other ways, as well this is just one of them.

Its harder for your camera or aiming flashlight to be blocked if they are mounted higher.

If you need to play defense it’s better to be tall as you have a possibility of blocking a shot or making it harder for a robot to aim.

There will be many really good robots that go under the trench, very possibly the best robot in the world is one of them but I’m pretty positive that the best alliance in the world doesn’t have only robots that can go under the tench.


Correct me if I’m wrong, but won’t a conveyer system like this be prone to jamming if the balls get too close to each other?

This is just a sketch, our goal here was to see if we could store them in a line.

This sketch however does have belts on both sides of the balls, so if that is the case I don’t believe jamming will be as much of an issue as the balls are not rolling but instead progressing upwards together.


Thanks so much for this detailed response!

Any conversations about drive wheel size? 4 vs 6 vs 8 inch wheel diameter, and 6 vs 8 wheel drive?

Plan is to base our drive base off this prototype wheel configuration we built in the winter.

6" wheels, 4 pneumatic + 2 omni.

Nothing set in stone, we could still switch it up.


I think if you use omni and slam in to the 1.5 inch bars in the center they will shatter

Teams successfully used them in 2016 without issue. We will have some drive test videos in tonight’s post going over bumps.

The vexpro omnis are more durable than a lot of people think, it is possible to break rollers but its not easy.


What is the desired effect from not spacing your wheels evenly? I don’t think I’ve seen that before.

It is a little uncommon. The goal is to get high traction while eliminating scrub. It’s a similar concept to a 2+2 (Winnovation used this very effectively before). We ran a 2+2 in 17 and liked it but it’s hard to keep it tracking straight. Adding the 2nd set of traction wheels helps increase scrub a little but still give you a small wheel base for steering. The omnis clearly don’t scrub at all. This all comes at the disadvantage of having your center of rotation, not the center of your robot.

2 examples I know of teams have used a 4+2 setup are 148 and 2168 in 2018. Other teams have used a 4+4 setup that should drive similarly, 1684 and 3542 in 2019.

From the bump testing we did tonight it seems nicer because you don’t have a center wheel to lift your robot straight up when you drive over a bump.


Day 3

Today’s Tests

Bump crossing tests - CG is important and braking hard is bad

Low CG and CG at the front of you robot handles the bumps better. Also this chassis has pneumatic tires.

Intake Test
Protopipe has proved very useful for quickly making intake tests.

Drawer Slide Elevator Test

More videos and photos of these tests can be found on our photo gallery Photo.Spectrum3847.org - 2020 Build Season Gallery

Robots that inspired us today

2009 - 67 (Multi ball shooter)

2012 - 48, 341 (Photo Gallery), 548 & 330 (shooter on a pivot)

2013 - 33(Frisbee path), 987 (gas springs on climber), 1986 (climber)

2014 - 2590 (ratchet)

2017 - 319 (ratchet)

2019 - 148 (elevator bearing blocks)

Sketch of the Day

Goals have the intake behind the shooter and enough room to hold 5 balls. Intake would be able to bring two balls up on to our robot at a time, those two balls would then funnel into a single file line.



With this sort of design, wouldn’t you have to make sure that you pick your balls up at the right interval? Since what I see is that if you have one ball in the feeder and you go to intake another ball, you may need to move the ball that’s already in the feeder further up to make room for the new ball. If your timing was off then you wouldn’t get a full 5 balls in, correct? The gaps between the balls would grow.

I assume you are talking about the sketch. You don’t have to worry about timing you have to worry about making sure your belts can slip against a ball if needed and that something is stopping your final ball from going into your shooter until you are ready. While you are intaking something is at the top of the path preventing that ball from moving and the belts slip nicely against it (test to make sure your belts don’t cut the ball, etc) and balls load in behind it. When ready to fire remove the obstacle at the top and all the ball slide smoothly into your shooter.

A good example is here with 118’s 2012 robot. They intake two balls to the top of their conveyor and are still able to bring in the third ball.

There are more things you can do to sequence the balls nicely with sensors to detect ball locations and reversing your belts during parts of the intaking process.


So a simple cylinder with a hood or latch sort of mechanism would do the job?

I don’t know, this is just a quick sketch, you’ll have to prototype and find out. Maybe this whole concept fails with these balls, there is only one way to know for sure and that’s to build a prototype and find out.