Quote:
|
Originally Posted by CharlieWilken
|
I was curious about this too, and saw many teams doing it successfully in 2000. There is a pretty good white paper powerpoint at ChiefDelphi by ChrisH on different ball lifts that shows how wires/chain could be used for a telescoping lift. So I built one out of Lego in a few hours to learn more about them.
Motorized Lego Telescoping Lift
Photo Contracted
Photo Extended
If you exclude the hook at the top, the structure doubles it length with three segments, going from 6.5" to 13".
The design certainly taught me some lessons about these lifts. Here they are:
1 ) There is a lot of tension in the plastic lego chain, so much that the chain pulls apart before the structure can actually lift itself up the bar, though the motor/gearing has enough torque for that. I would not use chain for a mechanism like this. Instead, I would use flexible steel cable and pulley. At the base, instead of a driving gear from the motor, the motor would turn a spool. The spool would have enough extra wire to move in both directions without running out of wire. Don't forget a tensioner... with cable or a non-continous chain (it can be fixed in place, like the locked gear, on the top segment) tensioning may be pretty easy.
2 ) Side loads. Just like overhanging a wheel can be bad, overhanging the shafts with the chain in the model (the ones you see moving, closest to the camer) is bad. It caused a side torque that jerked twist the assembly toward the side, almost breaking it, when the model reached the top of its travel. To fix this, I would route the chain/wire in the middle of the extension segments, via a hollow extension segment or extension segments on each side of the chain/wire.
3 ) Related to #2: The distance between the bottom-most two gears (the ones on the base) causes a FORWARD torque on the assembly. The longer this distance is, the more twisting force on the linear bearings, and the higher the chance of binding. The solution to this one is keeping that distance tiny, or doing the same solution to #2.
4 ) Adjustable speed. The first time I set this up, it was way too fast. Adjust the gearing from 1:1 to 5:1 helped quite a bit! Two sprockets and some chain could do the same thing. Make sure your motor is running at about half the no-load speed when it's lifting the robot up! This is where DC motors make the most power.
5 ) Overtravel. It breaks when you keep tellling the motors to pull up. The easy solution: limit switches that prevent the motors from moving when the thing's just about to hit a physical stop.
6 ) Your linear bearings MUST be super-smooth. The lift would bind when I had a bunch of 2-stud lego plates against 8-long lego slides. Bad! I replaced the 2-stud plate with 8-stud plates and the binding was less frequent, but still happened. If the plates weren't pressed ALL the way in, a tiny gap would catch.
7 ) Your linear bearings must have sufficient overlap (In my case, 6 out of 16 studs overlap). The greater this overlap, the less each segment can rotate with respect to another. This is an engineering tradeoff between reduced chance for binding and extra extension length.
8 ) I would consider using stretched rubber tubing between segments to reduce the torque the motor is required to produce.
9 ) Exposed chain when the lift is fully extended; may be vulnerable. Left as an exercise to the reader ;-)
The designs that used 80/20 or other aluminum extrusion in 2000 for hanging seemed by far the easiest lifts to construct. It can be done with entirely off-the shelf Al extrusion, linear bearings, pulleys and cable. A preliminary pricing from 80/20 showed a price on the order of $500 for the assembly, which would weigh about 20 lbs.
If anyone listening can chime in about their teams' experience with a lift like this, it'd be appreciated!
Thanks,
Brandon Heller
449 Alum / 931 Mentor