#robotics Logs

Jan 18 2020

05:08 AM mun: anyone know why delta CNC machines aren't so popular?
05:13 AM mun: all CNC machines i've seen are cartesian. are delta ones more expensive to make?
05:51 AM ace4016: mun, they lack good torque at the bits iirc
05:52 AM mun: ace4016 right. but why would it lack torque?
05:52 AM mun: because of the ball joints?
05:58 AM ace4016: nay, i believe it has to do with the indirect application of force from the linker arms; delta robots are fast, but for the same amount of force applied at the actuator, you get less force out of it. i'm going off my memory from quite some time ago and it may be incorrect
05:59 AM ace4016: also, delta robots are more expensive, more constrained in their work area, and a few other disadvantages that lead to using the cartesian types more so
06:03 AM ace4016: i'm starting to remember some details about precision and the work you need to get it up to a simpler design like a gantry/cartesian robot
08:14 AM mun: ace4016 right
08:14 AM mun: but do delta arms guarantee the load to be parallel to the ground plane?
08:15 AM mun: is that a characteristic of delta arms?
08:15 AM ace4016: yes
08:18 AM ace4016: i think if you go outside of their control bounds that doesn't hold true, but you're also about to break the machine at that point :P
08:19 AM mun: hmm i wonder why the base is always parallel to the ground plane
08:20 AM ace4016: it's the property of the parallelogram (sp?)
08:22 AM mun: but, is it the property of the parallelogram or property of 3 parallelograms arranged in a triangle?
08:23 AM mun: suppose one pair of the arms were missing, would the base still be parallel?
08:32 AM mun: yeah i suppose
09:22 AM ace4016: technically there would be nothing constraining the platform, so probably not; though maybe two linkages are sufficient (though you won't have full movement/control without the third)
09:23 AM ace4016: it's a geometric property (like most kinematics)
02:41 PM SpeedEvil is now known as Guest84965
02:43 PM BitEvil_ is now known as SpeedEvil
05:16 PM chisight: ace4016: couldn't you just adjust the gearing ratio of the delta to increase torque and decrease speed? low torque and high speed don't seem to be unique to deltas. you could also do the opposite, increase the gearing on a cartesian to make it faster and lower torque.
05:19 PM torque: wouldn't delta also be more susceptible to vibrations at the cutting bit because of the use of linkages
05:56 PM rue_: if you can build two machines, one with 800 parts at a cost of $400, and one with 20 parts at a cost of $40, which would you choose?
05:57 PM rue_: hint, each part is a point of wear, service, and failure
06:02 PM veverak: second one
06:28 PM ace4016: chisight, probably. once you have the IK solver figured out, you can figure out the various things the machine can do and the trade-space you can play with. i never successfully built one so don't have fancy calculators like i normally would to play with those numbers :P
06:31 PM chisight: hmm, it makes sense there is a way to calculate the speed and torque and i know that for at least some non cartesian the speeds and torques vary based on where the head is but i've not played with any sw that could do something like this.
06:49 PM rue_: http://www.kiblerelectronics.com/bob/app_notes/note11/images/schematic1_zvs_driver.png
11:41 PM rue_ is now known as rue_mohr