Glad I watched until the end. The last 1 minute was the most useful. I like how you showed that a simple design change can greatly affect the amount of backlash in the system.
I was making a mechanism out of Lego that would measure the indentation pattern of a flat surface like wood. It need utmost precision since the change was on a very small scale, and backlash became a significant problem. I simply added a weak rubber band on one of the gears, and this pulled the gear to the extent of its backlash, therefore preventing the backlash entirely.
@@TechnicBrickPower yes, the mechanism worked freely of backlash after I put the rubber band on it. There was basically a spring in a parallelogram design that kept it extended. A stick was attached to the end of the parallelogram, so when the wood indented, the stick went in. In a complex sort of events, the one parallelogram edge was fixed to a gear which spun when the stick was moved up/down by the indent. This gear turned a gear rack back and forth and thus showed the movement of the indentation. I didn't really use the indentation measurements for anything, I just made it for fun :), and I might post a UA-cam video on it.
Interesting video. My 2 cents: backlash is caused by play (between gears and between the axle and the hole) but also based on deflection of the parts. That deflection is based on torque (forces on the gears) and when you line up 30 gears, the input torque will be higher than the output torque (which is zero). Also, using cantilever shafts will cause the center of the gear to move quit a bit out of alignment, increasing the backlash. Backlash is also measured from one direction to the other so double what you calculated :)
Good point, but you forget one thing: the deflection is also dependant on the axles used. the tan pin connectors have a lot of play for example, while an axle supported on two sides has a lot less. the friction pins also have very low deflection, but are detrimental for other obvious reasons
Hi Klaas! Yes there's always more to it than meets the eye. The main point I was trying to illustrate is that the backlash angle can accumulate quickly with a long gear chain. 🙃
You might be able to preload the gear train as well, to eliminate backlash in one direction. To eliminate backlash in the other direction, you could have a duplicate gear train which is preloaded in the other direction. Since you don't need unlimited rotation, you're not as limited in your solutions.
![[UA] NAVI vs MOUZ | BO5 | IEM Rio 2024](http://i.ytimg.com/vi/CMBYk0cbGoA/mqdefault.jpg)
Glad I watched until the end. The last 1 minute was the most useful. I like how you showed that a simple design change can greatly affect the amount of backlash in the system.
Hi Don, thanks for watching to the end. I am redesigning the omni-directional model as we speak! 🙃
I was making a mechanism out of Lego that would measure the indentation pattern of a flat surface like wood. It need utmost precision since the change was on a very small scale, and backlash became a significant problem. I simply added a weak rubber band on one of the gears, and this pulled the gear to the extent of its backlash, therefore preventing the backlash entirely.
Hi Eliot. Wow did your mechanism work? How did you use the measurements? 🙃
@@TechnicBrickPower yes, the mechanism worked freely of backlash after I put the rubber band on it. There was basically a spring in a parallelogram design that kept it extended. A stick was attached to the end of the parallelogram, so when the wood indented, the stick went in. In a complex sort of events, the one parallelogram edge was fixed to a gear which spun when the stick was moved up/down by the indent. This gear turned a gear rack back and forth and thus showed the movement of the indentation. I didn't really use the indentation measurements for anything, I just made it for fun :), and I might post a UA-cam video on it.
Interesting video. My 2 cents: backlash is caused by play (between gears and between the axle and the hole) but also based on deflection of the parts. That deflection is based on torque (forces on the gears) and when you line up 30 gears, the input torque will be higher than the output torque (which is zero).
Also, using cantilever shafts will cause the center of the gear to move quit a bit out of alignment, increasing the backlash.
Backlash is also measured from one direction to the other so double what you calculated :)
Good point, but you forget one thing: the deflection is also dependant on the axles used. the tan pin connectors have a lot of play for example, while an axle supported on two sides has a lot less. the friction pins also have very low deflection, but are detrimental for other obvious reasons
Hi Klaas! Yes there's always more to it than meets the eye. The main point I was trying to illustrate is that the backlash angle can accumulate quickly with a long gear chain. 🙃
You might be able to preload the gear train as well, to eliminate backlash in one direction. To eliminate backlash in the other direction, you could have a duplicate gear train which is preloaded in the other direction. Since you don't need unlimited rotation, you're not as limited in your solutions.
Hi SuperAwac - thanks for that information - yes I will try that sometime! 🍺
Lego parts are suprisingly precise but backlash adds up quick. For long distances I'm fond of the Lego chain or axles and bevels.
It has nothing to do with precision, standard gears have to have backlash or else they can't work.
5000+ gears = more backlash.
Yeah you can create as much of it as you like! 😂
@@TechnicBrickPower thanks for your examples.