I was browsing through “What can I build?” section of the SES on the .com, and I noticed Sparky…
What are the advantages of using the ball stud and link for the 3DOF legs versus the tranditional 3DOF design such as the ones found on the *H3-Rs and Phoenixs???
The advantage is lower DOF 2DOF instead of 3 DOF which means fewer servos to buy and lighter weight. The movement is more static however.
Still, it’s a clever build!
I’m wondering about all the weight on the tail! Probably ok IF there’s someting in the grip!
Alan KM6VV
Some linkage actuated legs are setup such that when the leg is in the resting load bearing position, the linkage is in line with the servo spline. This takes the standing rotational load off of the servo. Some what like the human leg being locked at the knee while standing to take the load off of the leg muscles.
I’m thinking the linkage is more for providing pantograph-like motion, you get two leg section moves with one servo!
EDIT:
I think I found the legs at:
lynxmotion.com/images/html/build084.htm
Indirect-Drive Parallelogram 2DOF Leg
Alan KM6VV
The linkage on the back side provides for the parallel operation. Check what the position of the servo linkage would be with the leg full down and probably supporting the most weight.
Yeah, the leg tip travels vertically. Clever.
Leg full down, The bell crank of the servo is parallel to the ground, and thus would be at max torque (max weight on leg). As the leg is raised, the angle on the servo horn to the link approaches 0, and thus the toque drops off. Max torque is needed when supporting the 'bot, less when the leg is off the ground. Nice match!
Alan KM6VV
I’m not sure if the servo horn stops at parallel to the ground or not but if the position of the horn is parallel to the servo case, then the load could easily be supported. It looks like in the image above, the servo horn is perpendicular to the ground?
In any case, the design is very good for a low DOF quad. 8) 
That’s true, we don’t know where the servo horn stops. It might even operate in two positions, both at 45 degrees to the body!
This photo looks like the more “up” position, but as it’s probably not being driven and is obviously not on a 'bot, we can’t really tell.
You think it toggles between “horn up” and “horn down” positions?
Please tell us Jim!
Alan KM6VV
This is the 2DOF indirect drive (mechanical advantage) leg design. It’s usually optimized for throw in one direction or the other. The fun thing is there is a nonlinear response in tibia motion as it relates to the servo motion. When in the middle the servos under the most load, but at the two extremes the servo is at minimal load. Most bots using it can operate anywhere in the range of motion. But heavier bots can keep the legs in either up or down position and carry a lot of payload doing it.
The way 2DOF indirect drive legs operate (generally) is the servo horn is connected to the femur via a connecting rod. In the image the leg is a little raised from mid position. You can see all of the load on the servos output shaft is rotational. However when the servos output is rotated about 45° CCW the femur lowers the leg, and the connecting rod and servo horn line up. At this point the load on the servos output shaft is no longer rotational. This means the servo isn’t doing any work to hold the leg in the down position.
Technically this will happen at both extremes, but at one extreme a lockout situation may occur. This happens when the vertical pivot point and the femurs connecting rod connection point and the end of the servo horn all three line up. If this happens the servo can’t force the femur to budge. For this reason it may be necessary to prevent the leg from going past a certain position.
This is a very strong and stable leg design, although not cheap due to the large number of components needed to build it.
I also did a lot of experimentation with a similar concept… This leg also has the parallelogram for vertical tibia motion. The leg is in the full down position. It’s just as strong as the first method only more streamlined and elegant.
Thanks for that explanation.
I love this design:
What is the mechanical advantage? the ratio of the servo horn bell crank length to that of the attachment point on the C bracket to it’s pivot? Or is there more to it then that?
I can see that the torque would probably follow a sine function, That is to say 0 torque at 0 degrees, and max torque at 90 degrees. so 45 degrees (midway between right angle and lockup) would be roughly 70 percent.
I suspected you might have a lockup at the extremes! But that can be prevented easily enough.
I wonder if the mechanism could be taken one step (pun intended) further, and intentional detents allowed at the two extremes of motion, such that the leg would “lock up” in those positions, and no further torque would be required to maintain either position. In use, the torque would be applied to a joint in the connecting arm, rather onto one end of arm of the arm its self, and thus be able to get the link into and out of either of the detent positions. (too much thinking).
The second example is indeed streamlined, however I can’t see enough detail to discover the linkage arrangement.
They would both be fun to play with!
Thanks for the detailed description.
Alan KM6VV
