pic: Shooter damaged a Boulder

[electronicsdude]

Originally Posted by MrForbes

After playing with the ball for a few minutes, I have very little interest in trying to shoot it using a spinning wheel. It's not that type of ball.

However, building a medieval siege engine (catapult) looks like a neat plan!

I will testify that the catapult works, I will also testify that it creates one hell of an accuracy problem.

[Ether]

Originally Posted by electronicsdude

I will testify that the catapult works, I will also testify that it creates one hell of an accuracy problem.

Please tell us more about the design of your catapult that has an accuracy problem.

[Tom Line]

Originally Posted by electronicsdude

I will testify that the catapult works, I will also testify that it creates one hell of an accuracy problem.

I suggest you take a look at the end of the catapult. Do some slow motion video on it - even the slow motion that a Samsung phone can shoot. We did quite a bit of testing in 2014 and found the biggest contributor to inaccuracy from a catapult is when the ball 'rolls' off the ends of the catapult as it is shooting. As a result, we put wedges on the end of the catapult to keep the ball on until the end of motion.

In addition, keep close mind of how you are powering the catapult. If you are using motors, it will be inaccurate unless you are accounting for battery voltages, motor speeds, etc. If you are using some type of spring, get a fish scale so you can measure the spring tension and make sure it is consistent. Also insure you are shooting from a hard stop to another hard stop on the motion so that the catapult follows a proscribed motion each time.

People with wheels need to worry about slip, moment of inertia, and wheel velocity and how the ball is introduced to the shooter. Catapults just offer different challenges.

[billbo911]

We just received two Rhinoskin balls.
Both measured the same diameter as our KOP ball +about 1/16th".


In addition to measuring their diameter, we compared their compress-ability to both a new KOP ball and the well used one in the image.

We placed a 13.875 LB battery on the ball and measured it's compressed height.

Used KOP boulder = 8.25"
New RhinoSkin = 8.5"
New KOP boulder = 8.675"

So, all in all, they appear to be a good match. That said. actually shooting them will be the real test!

[bobl]

We put a small tear in the skin not by the spinning wheels but by hitting a corner inside the tower. We have since put a stop (piece of lexan) in the tower to stop the boulder and roll it back.

We are using 4" plaction wheels spinning at 2950RPM and have only minor scuffing on the skin. We are compressing the boulder 2" and are able to hit the high goal consistently (19 for 20) sitting on the outer works.

[Ether]

Originally Posted by bobl

We are using 4" plaction wheels spinning at 2950RPM

How are you controlling the 2950RPM?

[bobl]

We aren't really controlling it. Using a Banebot RS775 with a 3:1 transmission. Running at full power our tach shows a constant 2950 +- 15RPM. No encoders installed. The shooter is a prototype but it shows a lot of promise.

[Ether]

Originally Posted by bobl

We aren't really controlling it. Using a Banebot RS775 with a 3:1 transmission. Running at full power our tach shows a constant 2950 +- 15RPM.

The Banebots M7-RS775-18 motor has a 12v spec free speed of 13,000 rpm.

But you say the motor is spinning at only 2950*3 = 8850 rpm with a fully charged battery.

8850 rpm is only 68% of free speed. Of course the motor speed will be less than its free speed, but 68% is quite a bit less. Some rough calculations indicate the motor would be drawing in the neighborhood of 30 amps. Can you measure the amps the motor is drawing to confirm this?

Originally Posted by bobl

We are using 4" plaction wheels spinning at 2950RPM and have only minor scuffing on the skin. We are compressing the boulder 2" and are able to hit the high goal consistently (19 for 20) sitting on the outer works.

Try reducing the motor voltage by half a volt and see what happens to your RPM and whether you still hit the goal.

[Type]

We are running 4" stealth wheels that are connected 1:1 onto mini cims. The only time we damaged the ball was when we were testing, and somebody forgot to reverse the left motor. This made both motors spin the same direction. It made two small gashes in the ball but it's still usable. We have another ball we bought, but we don't want to use it until we get the final version of our shooter made. Has anybody noticed a difference in performance in a wheeled shooter from a new ball compared to an old vall? And does anybody know how often they will switch balls out in the game?

[Ether]

Originally Posted by Type

We are running 4" stealth wheels that are connected 1:1 onto mini cims

What speed are you running it at?

[Type]

Originally Posted by Ether

What speed are you running it at?

To be honest, I don't really know. We have all of our electrical on a practice robot so we just plugged the motors into an old power distributor, with no motor controllers. It shoots more than far enough, just need the correct angle.

[Ether]

Originally Posted by Type

To be honest, I don't really know. We have all of our electrical on a practice robot so we just plugged the motors into an old power distributor, with no motor controllers. It shoots more than far enough, just need the correct angle.

The correct angle when you are running it straight off a fully-charged battery may not be the correct angle toward the end of an actual match when your battery is getting tired.

That's why it's important to test your prototype to see how much the correct angle is affected by voltage. If the effect is enough to be a concern, you will want to design your shooter to operate at a lower voltage (say 9 volts) and then used a closed-loop controller (PID, bang-bang, Take-Back-Half, etc) to hold the wheel speed constant as the battery gets weaker.

[Type]

Originally Posted by Ether

The correct angle when you are running it straight off a fully-charged battery may not be the correct angle toward the end of an actual match when your battery is getting tired.

That's why it's important to test your prototype to see how much the correct angle is affected by voltage. If the effect is enough to be a concern, you will want to design your shooter to operate at a lower voltage (say 9 volts) and then used a closed-loop controller (PID, bang-bang, Take-Back-Half, etc) to hold the wheel speed constant as the battery gets weaker.

We were planning on getting a gear box to run at 2:1 so we could decrease the speed of the motor but still shoot the same. The mini cims have no issues with shooting the ball through a 7" opening, the motors don't really even make a noise as the ball goes through. We just need to figure if we can afford to run all the motors we want on our robot (11 total).

[Chak]

Originally Posted by Type

Has anybody noticed a difference in performance in a wheeled shooter from a new ball compared to an old ball?

We made a second iteration of our shooter that uses 4'' colsons and no longer rips up boulders. We're running that shooter off 2 775pros. We've noticed a significant and problematic difference between a fairly new boulder and the boulder pictured here, which is now even more torn up than before. I think that this is caused by the torn up boulder losing air faster and not "pushing back" as much on the shooter wheels. I guess catapults have the edge on this problem, although they might have their own problems.