Hi all, I am just getting started with quad and hex robots and I just dont understand a few things with respect to servo’s and torque ratings.
Using a quad 2DOF as an example how do you determine the needed torque of each servo? In other words, there are two servo’s on each leg…do they both need to be able to support the total robot weight? If not why and how do I calculate these values? Also, is there a simple formula to determine this that remains standard throughout various size, weight and leg configs?
There are a lot of simple physics formulas to determine such things.
But, to get what you desire, you’d need to combine many of those simple formulas into one big yucky one.
I’d say… just skip ahead to the rules of thumb and to the suggestions of those that have already done it.
It is vector math that your looking for. I started looking for it and been a lot of answers to what the weight should handle … still looking for the formula. I know they teach it in schools as I been exposed to it… but I have not been in a school since 1990.
What are the rules of thumb if I wanted a standard Quad or Hex to carry a much larger payload. Say something like 50lbs.
Or if I redesigned the body or legs, either overall size or formation, to attempt to better handle this weight, how would I begin to figure out the needed servo size?
Again, I am really in the hypothetical phase but I just want to have a better understanding of how things work before I pour random amounts of money and time into things.
Really? There are no servos that can handle that weight?
See, thats where my confusion comes in. I dont know how to figure out what each servo needs to move in a 2DOF system. Surely they would not all need to be able to support the entire 50lbs.
I am looking into hydraulics, pnumatics and other areas as well but they all add so many extra elements that I am not familar with.
If you want a really powerful servo get the SSPS105 Servo from Vantec vantec.com/ssps105.htm
It has a stall torque of 27 FOOT-POUNDS. Yup, thats right FOOT-POUNDS.
In other words, these servos have 5184 oz-in of torque.
This is over 12 times the torque of the HSR-5995 servos at only 5 times the price. Its a pretty good deal for the torque per dollar ratio for any servo.
Some people make their owm large servos like below from geared motors with external pots to determine position. Attaching the pot would probably be the difficult part.
Just try to keep terms of scale in mind.
Pete and zoom link to very powerful servos, but it’s important to keep in mind the cons of those monsters.
The biggest con is probably their current draw.
That vantec servo had a max draw of 9A!
If we shoot low and assume that servo draws 1A under normal strain conditions, then that’ll be 18A of current draw in a 3DOF hexapod or 12A in a 3DOF quadrepod.
That’s a LOT of current to draw from a battery.
This’ll mean larger or more expensive batteries.
In the long run, I’m thinking that Jim’s suggestion of hydraulics, or even just pneumatics is a more feasable option.
If your running a 50lb walker, you need to expect a higher current draw or a much much higher voltage. If your running hydraulics or pneumatics, you need an energy source to develop the fluid pressure. That in itself will require a lot of power. If you run through realistic energy calculations, to move a 50 pound walker at any reasonable speed, you will find that 18A at 12V (or 216W) isn’t a lot of power for this beast. There isn’t too many hydraulic systems or pneumatic systems that operate with than 250W.
The only reason I mentioned these servos is because an earlier post said there wasn’t anything that big out there.
I appreciate everyones additions here. I’m slowly starting to see the relationships involved.
Anyone have some links for hydrualic or pnumatic systems?
I still need to find the details on calculating servo/weight issues as they relate to the 'bots. I dont want to buy a bunch of expensive servos and throw them on something with a lot of weight only to find it wont move and I need to buy bigger servos. That being said I can see 50lbs is not realistic but I still would like to carry a much higher payload than the standard 5-10lbs.
True enough.
If one goes with hydraulics/pneumatics, a rolling pump of sufficient power will be necessary.
I suppose, though, that a hydraulic system would have much less trouble carrying it’s necessary weight of components in addition to the payload than a completely electrical system would.
Probably the best thing to do is some research on the net for larger leg types of robots and how they are designed and powered. Universitys and such have been testing these for a long time, so they may have some quick info.
Calculating loads for joint motors is beyond what could be explained here.
The first thing that needs to be done is a simple kinematic design of the leg. Then get an estimate of the total weight of the robot (including potential payloads). Then start looking at the worse case situation for the minimum number of legs that is in contact with the ground, because the minimum number of legs have to support the entire weight of the robot. Then divide that by 2 and round down to the lower interger (but never less than 1). Then divide the weight by the number of legs here.
This will tell you how much weight each leg must support. Now that was the easy part.
Next you need to calculate the load distribution through all the linkages in the leg. To do this you will need to get some text books on Statics and Mechanics. These books will teach you how to calculate load distributions and torques.
From this you can then properly specify the servo sizes for the robot.
Yes, I know, its a lot of work.
But then again, you can always do this the old school way. Pick motors that are much bigger (stronger) then you think you would ever need, and build the robot around them.
Would you, or anyone else, be able to reccommend any modeling or simulation software that could help with design and calculations for weight, power, etc. ?
I am aware of standard CAD programs and have heard of Solidworks but dont know anything about it.
