View Full Version : Servo Catapult Accuracy

02-21-2012, 07:14 AM
I was so impressed by the "Positive Pressure Gripper" thread & video at http://forums.parallax.com/showthread.php?137966-Robotic-arm-uses-quot-positive-pressure-quot-to-work-its-magic&p=1074926#post1074926 that I wanted to see if my simple robot arm could shoot any baskets. But after rethinking the G-forces and resulting abuse involved, I left my 6-servo arm alone and made a single-servo catapult-style shooting arm test jig to see how repeatable a servo "sling" could be. Pretty dang repeatable. Futaba S3004 analog servo running on a 7.4V lithium battery and my fave, a BS2 HW board. The arm is 14" long, lightweight aluminum tube.


PJ Allen
02-21-2012, 12:19 PM
Good fun.
Next step: Auto-loader.

02-21-2012, 12:20 PM
whoua !!

I build bots wit a group of kids , they will like a catapult.

Please some details and close pictures .

7,4 volts for a Futaba S3004 is it not a bit high ?

Jean Paul

02-21-2012, 01:20 PM
A while back, there was a thread asking "What's an engineer?" If it has to do with the practical application of theory and technology to a problem, then, sir, you are an Engineer!! What a great quick build teaching tool!!

Multiple servos, multiple arms, multiple auto-loaders.......

Duane Degn
02-21-2012, 01:52 PM

Very Cool! That's impressive.

Servos sure make easy for a uC to interact with the physical world.

Good fun.
Next step: Auto-loader.

Maybe this is where your (erco's) arm could help out?

02-21-2012, 02:45 PM
As Charlie Sheen would say Winning!! Great project.

I agree, an auto-loader would be cool. So too would a turntable with two or three cups, and the chance that a ball won't make it in.

-- Gordon

02-21-2012, 03:28 PM
Great feedback, lads. I envision a second servo as a turntable to steer the arm to hit a variety of targets at diffferent distances, requiring different trajectories. This simple design puts a lot of stress on the servo. The "launch" sequence is simply slamming from one pulsout position to the next. The servo speed is limited, that's why I ended up with 14" arm and operating at 7.4 volts. It would be more efficient to use the servo to cock a rubber band or spring catapult and release it, but that would complicate things a lot and probably require 3 servos (cock, release, and change endpoints) to do what this one is doing now.

The consistency is surprisingly good, but the servo power will probably degrade over time as the motor brushes & commutator wear. Such is life in the big city, but it's fun for now. Too bad the scale is so big. I'd like to nail everything down to a board. Like a 4x8 board!

02-21-2012, 04:25 PM
Surely I didn't need to mention that a laser sight is in the offing!

02-21-2012, 06:02 PM
Maybe something like the old "Grand Prize Game" on the Bozo the Clown show. It was a series of buckets at increasing distance and the contestant tossed a ping pong ball into each bucket for better and better prizes. Top prize was a bicycle or something like that. There are clips of it on YouTube. Automate it with range finders, vision and who knows what else.

02-21-2012, 06:13 PM
This guy still sets the standard.


02-21-2012, 08:18 PM
Nice! Just one thing missing... explosives!!!!!! =:-} (Tho' you might want a bit more range.)


PJ Allen
02-21-2012, 10:01 PM
How about those little wads of paper thingys, with the gun-powder, or whatever; they go POW! when they hit their target (floor, door)? What're they called again?
I have some. Perhaps it's my turn.

02-21-2012, 10:22 PM
I've heard them called "cracker balls". Fun things. Used to fling them with a slingshot.


PJ Allen
02-21-2012, 10:44 PM
From a slingshot? They snap if you look at them cross-eyed.

Well, anyway,... all I have is Parallax's standard servo, and it's not transiting anywhere near fast enough. Dang.
Major depression sets in.

Wish I hadn't torn down my SpinStamp stepper circuit.

02-21-2012, 10:50 PM
According to ServoCity specs, a cheap Hitec turns faster (HS-311=.15 sec/60 degrees) than my cheap Futaba (S3003/3004=.19 sec/60 degrees). That's pretty significant in this application. Looks like I'm off to Pacific Coast Hobbies again for another HS-311 or two.

@PJ: As I said, I had to use a long arm and 7.4 volts to the servo to achieve that moderate toss. I'm not above trying 8 or 9 volts either. Tune for minimum magic smoke...

PJ Allen
02-21-2012, 11:02 PM
I was trying to get by on the cheap.

I still have one unopened box of "snappers".

From the box:
"D.O.T. Class C Explosive" - Talk to me.
"A novel trick item." - Whatever
"Recommended for outdoor fun." - Oh, stop already.
"Please do not put in mouth." - I'm way ahead of you.

02-22-2012, 12:47 AM
Cool project, but why do I suspect we'll see a flock of robots hurling snappers in the near future.

02-22-2012, 02:58 AM
Might as well make it a UPEW contest! :)

02-22-2012, 03:16 AM
At UPENE, OBC(Jeff) was hunted by Microcontrolled and his Nerf shooting Stringray.......the way this thread is going, 2012 might be a bad year for Jeff!

02-22-2012, 03:36 AM
Oh great! even more obstacles for the line following contest...
now how can we incorporate these into the ELEV-8 air show?
Catapults tossing snappers could add a whole new sense of urgency for the pilots in the baseball dropping contest...


02-22-2012, 11:06 AM
...and so it starts....the ground pounders increase their air defenses...then the Airedales come up with ways to counter the new defenses (dropping nets to tangle the servo arms...then a new projectile system on the ground....oh the waste...the electro-mechanical carnage!!

02-22-2012, 02:40 PM
From a slingshot? They snap if you look at them cross-eyed.

Well, that was in the "good old days" when they made things more substantial.


02-22-2012, 03:06 PM
Servos sure make easy for a uC to interact with the physical world.

From this ME's POV, uC's make it easy for servos to interact with the physical world!

Nick Ernst
02-22-2012, 10:48 PM
That lego launcher is pretty cool, but can it hold up against an air attack from an ELEV-8 armed with Estes rockets on it?....I think not!!!

04-10-2012, 10:52 AM

Well, PJ anyhoo... just roughed out for Expo. Auto loader, IR remote, multiple target zones, etc. Might not have laser & sonar seeking yet for Expo, depends on timing...


Loopy Byteloose
05-08-2013, 06:19 PM
That catapult is a great idea, but lots and lots of preparation.

05-08-2013, 09:20 PM
About 4 hours total to build & program. Lots of superglue and "that looks about right"! Now I could do it in two. :)

Hal Albach
05-08-2013, 10:26 PM
Perhaps you may want to look into a rotary solenoid to replace the servo. They can provide tremendous torque through a range of arcs (determined mechanically) and probably could hurl a missile quite some distance. Activation couldn't be simpler, a high current mosfet and a healthy 20-24 volt power source. By shortening the power pulse shorter distances could be realized. Since they are basically a very simple mechanical device with a coil they can take a lot of abuse.

05-09-2013, 12:39 AM
Activation couldn't be simpler, a high current mosfet and a healthy 20-24 volt power source.

For the ultimate in simplicity, how about nothing, which the current servo requires? :)

Additionally, the servo's trajectory (range of arcs) is all programmable using pulsouts.

It works amazingly well over a variety of trajectories. The only change I'd consider making would be to increase the range by using a torsion spring. The servo would preload the torsion spring to a programmable amount, and then the lightweight catapult arm would be unlatched by a tiny release solenoid.