biorobots usually have fine positioning feedback via a route which is largely disconnected from the actuator circuit. the position encoding is applied to the effector area of the mechanism rather than close to any pivoting points. this allows for an overall flexibilty of the mechanism without sacrificing output resolution.
if two diamterically opposed rollers arent stiff enough, why not use four, or eight, or sixteen, or..? they could be of various sizes, and form an ellipsoid, or a triangular drive form, or a square wave generator;) --- On Tue, 5/15/12, Viesturs Lācis <[email protected]> wrote: > From: Viesturs Lācis <[email protected]> > Subject: Re: [Emc-users] Some advice needed > To: "Enhanced Machine Controller (EMC)" <[email protected]> > Date: Tuesday, May 15, 2012, 5:28 AM > 2012/5/15 charles green <[email protected]>: > > i dont see why to care what any part aside from the > engagement area is doing, except maybe at some resonance > frequencies (grease it up good) and/or high speeds (why use > a reducer for high speed application?). > > > > With 200:1 reduction ratio the input _has_ to go fast to get > more than > few RPM on output. And robot arms need more than 2-3 RPM on > output to > move a joint by 90 degrees in a second or two. > > If only engagement area is controlled, then the rest of the > profile is: > 1) uncontrolled curve from > 2) thin and > 3) flexible material. > Every of these three factors contributes to unwanted warping > of the > flexible gear and here they all are together. > Under load the flexible gear is not elipse anymore, but some > kind of > strange geometric figure, becoming close to circle with > segments on > opposite side cut off (and displacement of output flange > relative to > input flange). > In text books they distinguish the wave gears: > 1) free form wave gears: > wave generator is diametrically positioned rollers; > 2) strained wave gears: > wave generator is either eccentric discs or elipse. > > Please, do not tell me to reinvent the wheel - it is proven > long time > ago - free form wave gears are not meant for pretty quick > applications, where stiffness is needed. > > BTW "strain wave reducer" name is the name to describe whole > wave gear > principle. And that is not a coincidence - free wave > reducers simply > are not up to the task. There is a flex in the reducer, > causing a > displacement, so encoder on the motor is useless for > deriving the > actual joint position. > > As I said and I do not want to repeat it again - forget > about free > form wave gears in robotic application!!! > > > 2012/5/15 andy pugh <[email protected]>: > > On 15 May 2012 07:29, Viesturs Lācis <[email protected]> > wrote: > > > >> If I cannot get the flexible bearings, then there > is option of > >> eccentric discs as a wave generator. > > > > You could probably dismantle a normal bearing, machine > down the races > > to make them much thinnner, and then press them onto an > oval former. > > No, those flexible bearings actually are _required_ to be > flexed for > normal operation . They even specify a range of min and max > ovality - > (D-d)/2 = 1,2...1.6mm (D and d - large and small diameters > of elipse) > for flexible bearing with inner diameter 90 mm and outer > diameter = > 120 mm.. They are not meant to be used "as they are" - as a > round > circle. > The thing is that there is extra play in the bearing, which > decreases > as the bearing is flexed, so even reducing the rings of > normal bearing > will not give a good flexible ring - the existing play in > bearing will > be not sufficient for it to flex to the extent I need. > > Viesturs > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's > security and > threat landscape has changed and how IT managers can > respond. Discussions > will include endpoint security, mobile security and the > latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Emc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-users > ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
