Is anyone aware of what the maximum load to a servo is when such load is perpendicular to the shaft axis? Specifically, lets say my robot weighs 4 kg. and I have 6 wheels, and a servo on each wheel. I would like to know if the servo bearings will be worked within a non destructive loading.
Is there a rule of thumb, or better yet a chart of this information for some or most of the common servos?
I ask this for situations where the wheel is connected directly to the servo, and the servo body becomes the mount point to the body. In other words, where using a separate bearing for the wheel shaft is not an option.
Cheap servos typically have plastic on plastic bearings. More expensive servos have metal on metal bearings. While expensive servos have ball bearings and thus are the most capable to handling shaft loading. Each of these would have dramically different shaft loading tolerances.
I’d look for metal output shaft with ball bearings to find the most capable load bearing solution. However, servos for RC are not typically designed for the shaft loading you are asking about. I have never seen a chart of this information for small hobby varios servos. Larger servos outside the hobby arena are designed for much greater shaft loading (at a premium).
The typical hobby servo - even the “heavy duty, dual-ball-bearing” types - have those dual ball bearings mounted to a stubby little shaft, in such a way that their outer edges are about a quarter-inch apart. This is hardly ideal for supporting side-loading, even if the load is to be applied directly in line with the bearings, and especially if, as is the case with most of such applications, the load is to be applied off-center from those bearings.
The “heavy duty” aspect of a servo assumes that it will be used in a servo-type application, pushing and pulling on a control rod that is mounted no more than an inch or two from the drive shaft, and a fraction of an inch from the support point of the final output bearing. Attaching a wheel to it that is several inches in diameter and significantly offset from its points of support is so far out of the design spec that it’s completely off the radar as compared to what it’s actually rated for.
For the most part, I won’t even begin to consider servos for any application that has their output shafts in a load-bearing configuration for anything over a few ounces. They just aren’t designed for that sort of thing. Yes, people design mobility bases with servos supporting a lot of weight every day; popularity doesn’t make it a good design ethic.
The general rule of thumb that I use when determining whether to use direct-drive servos or some other configuration is this: If I have to wonder if servos can support the side-loading and/or weight, then I assume that they can’t, and go with something more robust.
Thank you for the good answers. They really make sense. As I write this I’m thinking of ways to design the drive of a rover without the radial loading situation. Oh well…
