Better late than never? Steamworks has proven to be an exceedingly difficult game, but progress is progressing. Very proud of the students.
Killer Bees - Auto Mode Testing
33 has/is competing at:
If anyone has questions about the robot this year, we’ll do our best to answer them.
Cheers, Bryan
Do you have any close-up video or pictures? I’d like to know how you get balls to your shooter so fast.
Sure!
Here’s a Dan Ernst image with the shooter assembly partially removed. The fuel processor is designed such that it can be flipped to shoot out of either side of the robot depending on if we are red or blue. Bottom feeding belts convey the fuel into the ball lift which consist of three rollers, these lift the balls into the flywheel getting them fully up to speed and going the right direction. By maintaining a barely two ball wide (11") aperture, we get the benefit of not having to single file sort balls (which is what is really hard to do fast) while avoiding most of the multi-ball wide shooter accuracy trouble - the balls usually come out in a Z pattern rather than at the same time. There are some other small details (for example we found the loose floor belts convey faster than hard rollers) but mostly we made everything spin really fast. :rolleyes:
,Bryan
neat.
What are you doing for vision?
We’re using a pixy cam connected to an arduino which streams the center of the goal.
On that note we also have a LIDAR lite v3 connected to the arduino that tells us the distance from the goal, which controls the RPM of the flywheel.
I would love to see a close-up slo-mo of the shooter if you all have that!
Have you guys had any issues with concentricty of the shooter wheels? We did some tests under a strobe light with those wheels and found that they expand almost into a cam shape at high speeds, causing a lot of vibration.
Nice, congrats! I know it was a ton, but do you have a count of kpa for each auto?
We can take some video this weekend before we put it back in the bag. Are you interested in close up of the balls passing through the ball lift/flywheel or the feeding?
Your results are very interesting, can you explain what you mean by “cam shaped”? We have found that the wheels are very concentric and have very minimal vibration. We are only spinning them at about 3000 RPM nominally though, so not that fast. In 2016 when testing these wheels we found there was a ‘rough spot’ around 5000 RPM where the rubber expanded and had to retransition to concentricity. If you were trying to maintain speed in that range, it could be ugly. These wheels also require making your own hubs, so concentricity there is paramount.
We were counting scores at the time but I don’t remember which was which. The top right looks like it was one of the tests without the lidar so it’s shooting slightly short. On our practice field we’re making mid-30s occasionally straying into the 40s, that’s a separate matter from what the field counts during auto though. Of course, we’ll probably need to adjust some stuff once on a real field.
Edit: The wheel began to break apart- a slice developed allowing part of the wheel to fly out more under the forces of spinning
Basically, the wheel appears lopsided when viewed with a strobe at high speeds and vibrates a ton. This seems to be similar to what you describe. The “bad” speed is most likely well above our operating speed, so not much of an issue.
When we did the test under the strobe, we also had some AndyMark compliant wheels on the same shaft. They stayed pretty round.
What every team wishes their autonomous was like:
What their autonomous is actually like:
http://i.imgur.com/EuPrGDZ.gif](http://i.imgur.com/EuPrGDZ.gif)
Both if possible please!
How much compression are you running through your shooter? And what ratio do you have from the motors? Because whatever you are using, it’s working
Do you have any issue of fuel spinning in place rather than moving to the back of the processor? If so how do you get around that?
What are you guys targeting with the LIDAR? The boiler is a strange shape so we were trying to figure out how to properly adjust it.
I know this is probably the most boring part of you robot, but any chance you could comment on your floor gear pickup?
Sorry for not getting back to all of you sooner!
So, this wasn’t happening on the field. To the best of my recollection the most fuel an alliance we were on made this weekend is auto was 31. Two weeks to improve for MSC.
Gotcha, our wheels have not developed any cracks. Its possible that you were testing a defective wheel.
I’m so sorry, I totally forgot to get video for you. :o
The compression is adjustable. I “think” it is around 1/8" - 3/16" right now.
No issue, when the floor is acting on the belt, it does spin the fuel, but also translates the fuel along the belt.
We are targeting the cylinder part of the goal (around where the retro-reflective tape is). This was not working as well as we wanted this weekend so we disabled it early in the competition.
The floor gear pickup is three parts. The pivot (the wrist), the roller, and the reaction surface (the tongue). The pivot is actuated by a bag motor and articulates the whole assembly 90 deg in and out of the robot. The roller is powered by a 775pro through a versaplanetary then a 5mm HTD belt. The tongue is a flat sheet that is separately actuated via an overcenter linkage that locks it closed.
It’s a pretty good collector. We like full width collectors so no complaints there. If I was to do it again, I would probably eliminate the tongue in favor of a pneumatically actuated claw (keeping the overhead roller.) The leading edge of the tongue works fine, but is a constant maintenance point. Centering the gear in the robot would be a nice bonus.
Cheers, Bryan
Very impressive! Good job
I noticed that you wait a few seconds before shooting after triggering the hoppers. Why is that?
