Hi,
I would like to know does 4 393-motors powering on each Mecanum wheels (in H-Drive) with 4:1 gear ratio each wheel creates too less torque when moving the robot forward on the competition field. The Arm of the Robot is designed to be light-weight, Please view the photos.
Our robot first couldn’t travel more than around a meter on the mats, and later on it couldn’t even move from resting(on mats.). When we first test the base on the ground, the robot can travel better in forward and backward then sliding left and right, but then it became the other way around the second day.
There is also two another interesting problems, where the two back wheels couldn’t spin continuously(to right only) when the robot is lifted in the air and these two wheels are powerful enough to rotate the robot to left but hardly to right.
What was going on?
Thank you.
Summary of problems:
Robot cannot travels when placed on the field mats, but it was able to do so for the first time(moved less than a meter).
The back wheels cannot spin continuously to one direction only, and there’s a pattern of timing when they pause. They pause very shortly but multiple times in one cycle, and both wheels act the exact same way.
My Questions:
Are the above problems related with one or multiple factors? (Torque, the constructions, the programming etc.)
How do you know is there enough torque for moving the robot?
Is the a way to keep the 4:1 ratios with Mecanum wheels in a H-drive while improving the driving power? (Things like increasing the contact area of the robot with the mats so the weight spreads out more etc.)
*After reading the responses:
Problem #1 is mostly caused by low torque value. Problem #2 is circuit breaking too often? The pauses have timing patterns and happen in spinning just one certain direction.
other: Mecanum Wheels In H-drive (please ignore the lifting system) - - YouTube
A ratio of 4:1 for speed is far too high for a drive train. I would suggest directly driving your mecanum wheels with your high strength motors in a 1:1 ratio, the speed setting is fine. Additionally you seem to be using your mecanums in the wrong configuration. Mecanums allow all wheels to face the same direction like in a normal tank drive but gives holonomic capabilities due to the angle of the rollers on the wheels. Take care to have the right roller directions on the right corners of the robot.
The mecanum rollers should form an X when viewed from above and form an O when viewed from the bottom.
Is there an easy way to analyze the torque mathematically?
Those wheels seem to work fine when we push the robot on the ground with hands, at least left/right & forward/backward.
Your wheel configuration is interesting but I’m not sure if it’s the best. As for the torque, it sounds like you are hitting the circuit breakers in your microcontroller due to drawing too much current from your drive motors. On a 4:1 ratio, your drive just doesn’t have the torque necessary to move your robot unless it was an extremely light robot (say under 5 pounds or something). I’m not sure exactly how the math works out, maybe more experienced members or mentors can comment on that, but from experience most teams run 1:1 ratios on their drive and vary depending on their design/strategy between internal speed gears or internal torque gears.
Just looking at the wheels in the back - do you want them swapped direction wise? I can’t say for sure not playing with this configuration.
I think the natural driving forces would want to push outward on those wheels in opposite directions versus inward like the front wheels. The forces on mecanums should go along the 45 degree angle. Those forces should still cancel each other out.
Also, would the overall center of rotation be OK in your configuration? Left/right of the front and left/right/back sets should neutralize each other but front set combined with back set may not be ideal.
My drawing is crude but the normal way is to make a square with the wheels oriented all parallel to each other in a square and not in an H. As you get to the center of rotation the torques should cancel too.
However it may be more about the center of rotiation and you have torque fighting you. If you then try and find the real center of rotation on your configuration, you might see it’s not as neutral as you think. (you may need to get a kinematics book out here) Rotational forces may need to be taken into account as you translate the forces about. You need to apply the torque effects too for the angle off the axis. You might be making some forces that fight each other in your configuration when used as a whole which is only hurting you more. The front back center of rotation may be behind the horizontal wheels. So it may just be a matter of placement of the H wheels to get it really neutral.
You can eyeball neutral by following the 45 degree out of the mechanum center and see where they all intersect. Or I may be totally off base. Maybe you’re already there and you’re just over gearing it.
The reason your robot is not functioning correctly is not weight distribution. The problem is that your gear ratio is too fast for the weight of your robot, and the torque of your motor configuration. Adding more motors to your drive could help, but as a rough estimate, I would say it would take all 10 of your motors on your drive with a very light robot (drivetrain only and made out of aluminum) to be able to get your 1:5 gear ratio running on the mechanum wheels.
In short, a 1:5 ratio on your drive with mechanum wheels is just not going to work on a typical VEX competition robot, especially one that is made out of steel. I suggest changing this ratio to a 3:5, changing your 12-tooth metal gears on the wheels to 36-tooth plastic gears.
Since the front wheels are not programmed to spin when the robot is turning/swiping (sorry I didn’t address that at first), the four wheels don’t necessary have to be neutralized.
The robot is designed to swipe when rotate; The pivot is at the front instead of at the center for reducing the risk of the robot swiping away the objects during the intake since the intake system is at the front of the robot. The rotating of the robot is fine, but in one direction only (eg. able to swipe left but too week to swipe right) which is wired. So the torque is large enough during the spinning/swiping, what about why the robot can only spin/swipe in one direction?
Is there a mathematical way to confirm? Because each wheels are powered by a 393-motors which I believe is quite powerful even though with the speed favored gear ratio on. Is it possible to run this 'bot with the 4:1 ratio?
The robot did travel around a meter on the mats for the fist time though, with 4:1 ratio and Mecanum wheels.
Thank you for the tutorial.
The equation you’ve provided: F=UMG has the “U” represents the static friction, but why when the value that plug into “U” is “Tan 30 degree” ?
What’s happening when in theory it’s 12:1 gear ratio but has to reduce 40% and then becomes 7:1 in the end?
The image you’re showing has omni wheels in the diagram, not mecanum wheels. Omni wheels are commonly used with H-drives to allow 90 degree translation movements, or strafing, of your robot.
I agree with others posting here that the way you are using mecanum wheels may give you some strange motion results. Mecanum wheels are intended to be set up in a specific orientation, see this video to see an example: http://www.youtube.com/watch?v=o-j9TReI1aQ
This diagram that VEX provides is helpful as well for seeing what the orientation of your wheels should be - Mecanum Wheel Diagram
I don’t know how to figure it out mathematically, but I know from experience that you will never have enough torque on 4 393 motors to drive a competition robot at a 5:1 ratio. The highest I have ever tried is 2:1, and even that faced issues of a lack of torque leading to overheating sometimes. On a 5:1 ratio, assuming the motors span at 100rpm, if the motors had enough torque the robot would travel ~150 metres (~170 feet) in a minute. That is far more speed than is necessary or possible in a competition robot. The only feasible solution to your problem is to change the drive to a more reasonable ratio like 1:1, 1.4:1, 1.6:1 or maybe 2:1.
I feel excited to dig in more to Mecanum wheels. Here is what I think: As long as those wheels are place parallel or diagonal to each other, they will work when moving froward, backward, left, right and 45 degree diagonally, and even rotating/turning the robot, the key is to run and cease certain wheels during the motion(programming). But the turning pivot point may vary depends on how the wheels are placed.**update: it’s not “parallel or diagonal”, it’s “parallel or perpendicular”. (Thanks jgraber)](https://docs.google.com/document/d/1eEtPqp07xwzOsHPBJJat8pAIPO7ao19jTUIE8eqctBY/edit?hl=en_US), the red zone indicates the torque is too high?If so, how can this be defined without knowing the weight of the robot?)
Yep, 81A did end up doing this but they put normal holonomic wheels on the front too. You can visit them at Worlds to see it in action. The advantage they have is they can “wiggle their backside” around those goal elements to get a good read on it without having to back up and reposition.
But like Ryan pointed out, you’re using mecanums, so the forces end up different on the 45 degree offset. You would have to angle your front wheels on the 45’s to be the same exact driving forces as 81A’s. Yours are pointing in on the front. So it’s slightly different. it might still work but I think you need to be careful on the center of rotation.
To figure out the force and velocity directions on the mecanums I generally use the right hand rule of the center wheel spin and then offset the force to the 45 degree of the outer mecanum wheels.
Yes, Red means the torque requirements are too high. It can be determined experimentally with the lightest possible robot consisting of just battery, cortex, motors, wheels, and minimal frame. If the red was defined by a squarebot with only one motor per side, maybe the boundary zone will work with more motors. 5:1 speedup is just crazy though, as your evidence shows. Also remember that the speed chart is defined for simple wheels pointing in same direction. Omni/holo/mechanum drive require more torque.
Excitement is good motivation. If these are your first mecanum wheels, then start with the simplest proven method and learn how to program that, so that it gives you an experimental control. Then you can make a change (like H drive) and see if it works better or worse.
Do you think that mechanum wheels are just like omniwheels mounted at 45 degree angle? If so, then you can redo your drawing with omniwheel swap and see if it makes sense.
Note that a circle of omniwheels (+ or x config) has axles perpendicular, and can translate and rotate, but a star of omniwheels also has axles perpendicular and can translate, but NOT rotate. So your “parallel or diagonal” is likely to be insufficiently constrained.
Indeed, x config of omniwheels cannot rotate, but this is because they are placed in holomonic drive. The drive we have is an H-drive but equip with Mecanum wheels.
By center of rotation, you mean the pivot point right? We have our pivot point at the front of the robot (front wheels).
The problem we have is that the base is powerful enough to swipe the back of the robot in one direction while not powerful enough to so for the opposite direction. Both direction is possible to swipe when pushing with hands.