I have a foam pumpkin that I've carved with a weeping angel from the Dr. Who Blink episode, with her face covered on one side and attacking on the other.
I want to use a servo and arduino to spin the pumpkin 180 degrees as fast a possible, so that people will be looking at the pumpkin then I can hit a button to switch it when they blink/aren't looking.
You can see the pumpkin on the ground on the far right in the attached picture from my display last year.
Is there a servo that isn't too expensive that would be able to turn the pumpkin fast 180 degrees & back?
The pumpkin weights a couple pounds with the light and I have taken a wood turntable with ball bearings and manually turned it with a rope in the past and could use that if a servo would be able to spin it.
Is there something other than a servo that I should consider using that could spin it fast with enough strength but cheaper than a servo?
Numbers numbers. You haven't said what you consider 'fast' - 1s? 0.1s? 0.1ms?
Lets assume this is about a foot in diameter, assume all the mass is thus 0.15m from the axle
and the mass is 1kg. That gives a MoI of about 0.02 kg m^2, and say we want to accelerate
to 90 degrees in 0.1s, decelerate the next 90 degrees in 0.1s, that's 0.2s to spin 180 degrees.
That's an angular acceleration of 314 rad/s/s, so a torque of about 7Nm is needed (top
angular velocity is 31.4 rad/s or about 300rpm.
That's upto 220W load, and the same for dissipating the rotational energy in the second 0.1s.
[ thats no hobby servo!! ]
Doubling the time allowed will reduce power requirements by a factor of 4, note.
Thanks for the great reply, a lot of information and good to see the math.
I'm home so I was able to weigh the pumpkin and light and they are 11 ounces (or about 312 ounces if it makes the math easier) about 2 ounces of which is the light.
The diameter at the widest area is about 10 inches and a good amount of the pumpkin's weight (25%?) is towards the middle with the stem and everything else.
Also it looks like a blink takes about 300 milliseconds so if possible to get a servo that would still be a "hobby servo" and be within that time that would be great. But if not around a half a second to one second would be great as well.
Hopefully this would improve the math to make it a little more reasonable to accomplish without having to wire in a 240V line and pay thousands for a commercial servo
You could look into using a solenoid of modest power attached to the bottom of the turntable. The solenoid would be a rapid linear motion, with the appropriate joint to allow for the short arc path (think train wheels).
Using a good low friction pivot solution (ball bearings, slick washers, grease, etc), you don't need much force at all.
A solenoid is much better than a servo because getting it going is one thing, but stopping it from turning at the ends is a lot of moving mass, and a solenoid is simpler to incorporate travel limits than likely chewing up the gears of a servo.
You don't get it, we are talking 100's of watts to perform this motion. A solenoid that powerful would be
really massive. Solenoids are also very inefficient compared to motors.
(And a beefy motor controller and an encoder to determine position and set up a feedback loop)
In fact that sort of motor comes ready to drive a chain which is probably going to give an easy way
to gear down. Its also very cheap for its specification, being a commodity item.
So this this something doable with a 5kg 120v solenoid which has about a 1/2 inch movement?
I think I'll try to mock something up to try with cardboard, but is there any online simulators that I could put in the lengths and different pivot points to see the force needed to move it?
INTP:
100's of watts to push something on a low friction surface that's <1lb? Or did you miss the part about it not being a real pumpkin?
Have you read the thread before jumping in with a comment that rather misses the point?
What has friction got to do with accelerating a mass?
Yes, 200W was the back-of-the-envelop calculation to rotate the 1kg pumpkin in 0.2 seconds. There
are high accelerations involved, as a = 2s/t^2 (or for rotation angular accel = 2 theta / t^2). That
exponent of -2 for the time is the killer. And MoI goes up as radius^2 as well, which is also a killer.
The mass seems to have changed after I posted from "a couple of pounds" to "11 oz", and the mass
has been determined to be sightly more on-axis than standard pumpkin geometry would imply.
So its going to be rather less than 200W as a result, but still beyond the reach of a hobby servo.
