Hello all! My team and I were planning our robot design for the Spin Up season. In our though process, we thought that strafing might be a convenience for this game in picking up discs. We saw the Ri24Hr reveal and saw a mecanum drive. We also then saw some CAD leeks and those had some X-drives. Can you help us decide which to build and provide the pros/cons of each of them? Struggling to decide which to build.
1 Like
In general, mecanum drives are much easier to build with.
But when strafing, mecanum drives can experience more friction.
Honestly, just stick with whatever’s easier for you
Gotta warn you, strafing kills your motors. My team experimented with mecanum wheels last season for Tipping Point (We figured it would help by strafing over for acurate positioning to set a goal on a platform) and we ultimately switched before worlds as we couldn’t make it through many matches without our drive motors stalling out due to high temperatures. Strafing is good when used in short term, but, at least from my experience with it, it doesn’t work well when used a ton.
Are you sure that it wasn’t an issue with your design? I’ve teams use mecanum drives successfully, not to mention ftc teams and industrial applications that use them without issue such as certain fork lifts. (Granted vex mecanum wheels use a different design than industrial ones, and that design is not very good in comparison, and most teams with mecanum drives do seem to strafe slower than x drives but it shouldnt be a robot breaking issue.)
4 Likes
Could have been. To be honest our base was not built the best so we could have had an issue there.
I will say though x drives tend to work better due to the omni omni wheels being lower on weight and I beleive the smaller rollers tend to work better. Or at least this is from my experience with an x drive and observing mecanum drives.
I’ve build robots with X-Drives and with mecanum wheels. Both can be very good if built well. However, the much more critical thing to get right when using any kind of holonomic drive is to have even weight distribution. If you don’t have that, then the robot might move sporadically when accelerating, and be very inconsistent in autonomous.
That being said, X drives are generally better on a function sense, although they take up more space inside the robot, are harder to build, and even harder to build well. In my experience good alignment between all four wheel is key to a consistent autonomous. Also mecanum wheels tend to strafe sideways at about 87% the forward speed due to friction on other losses. X-Drives on the other hand handle strafing much more smoothly and there are much fewer losses due to friction.
If you do build an X-Drive, I would highly recommend making polycarbonate gussets. It’s easiest to cut them on a CNC router I’d you have access to one, but if not they can be cut out of hand with a printed template. If you cut them by hand you should use a center punch to ensure all the holes are perfectly aligned. Using robosource shoulder screws and drill bits as close to the shoulder size as possible will ensure good alignment. I built our X-Drive without shoulder screws and found that the play in the screws was sufficient to skew the drive enough to negatively effect the performance. One other note for CNC cutting polycarbonate, it works very well to use carpet tape or other strong double-sided tape to pulled down the polycarbonate to a sacrificial sheet of wood to hold the parts down when cutting.
(Please ignore the keps nuts and poor build quality)
16 Likes
What size drill bit do you recommend?
I believe 11/64 is the closest you can get with common drill bit sizes, if you go one size smaller you can make the hole self-tapping but it won’t work so well with shoulder screws.
3 Likes
If using shoulder screws then a #15 drill bit is the exact size of the shoulders at 0.180"
For non-shoulder screws, a #20 drill bit fits perfectly at 0.161" (I used this on a drivetrain this year on the bot that sadly wasn’t ready for worlds where I needed screw joints on 2.5" screws and it kept them perfectly aligned)
6 Likes
Not make-or-break argument, but mec wheels have a theoretical sideways speed of 70.7% of the forward speed (sqrt2/2) since the rollers are angled at 45 degrees. This comes into play if you want a motion planning system that spends more of the route time going forward than sideways.
I assumed that the decrease in speed was the same as the decrease in distance traveled, which when I had measured it on my tower takeover bot was ~87%
Edit: Just checked my old code and it was ~88.85%
4 Likes
I get it moves around 1/sqrt2 the speed diagonally just like an X drive does, but I don’t see why it should go slower straight sideways besides friction, and that goes against my personal experience using them.
1 Like
I’m dumb, I remember like 2 years ago I had this discussion and had my mind changed. Yeah, it theoretically goes the same speed sideways as forwards. Going diagonally actually has me stumped, because I think it should go at a velocity of 1i+1j, where driving forward has a velocity of 1i.
Why would it go faster diagonally? it goes full speed when all four wheels are moving, but only uses two of them when moving diagonally. An X-drive and mecanum drive are programmed exactly the same and other than friction I believe they are fundamentally the same concept. This website has come cool interactive diagrams to visualize the speed. It even lets you switch between an X-drive and a mecanum drive and you can see they move identically.
8 Likes
What size wheels did you use?, one concern I have is my schools robotics class only started last year so we only have 4” wheels so far, with such big wheels on the sides, to intake the frisbees, have to be dead on or would just push away because we do not have expansion in most of the Match so pop out funnels aren’t a opinion
Thank you so much for this advice, I’ll definitely keep it in mind. Can you elaborate more on how to design and print the custom gussets? What advantages do they have over the metal VEX ones?
Also, I have heard that X-drives increase the robot’s forward drive velocity by a factor of sqrt(2). Given this, can y’all share your experience on gearing up X-drives? Is this necessary, or is it fast enough without gearing due to the increase? For example, if I were to gear to 280 rpm, would it be too fast (aka too less torque) because of the sqrt(2) speed advantage?
I have never built an x drive, but yes, you would definitely want to factor in the sqrt(2) when selecting a gear ratio, since the increase in speed by 1.41 is accompanied by an inversely proportional loss in torque.
So something like 280 rpm on 4" wheels for an x drive would really be like almost 400 rpm on 4" wheel in a standard tank setup. Which is way too fast and weak, especially if you’re using a standard x drive with only 4 motors. With 4" wheels, just having them run at 200 rpm on an x drive will actually get you a speed roughly equivalent to 280 rpm on a tank drive, which is a pretty good speed for a light robot. So for 4" wheels, it seems to me best to just run 200 rpm.
For 3.25" wheels, running 200 rpm will only get you the equivalent of 230 rpm on 4" wheels, on a tank drive. This might not be too slow depending on your preferences, but you could certainly push it to something like a 5:4 speed ratio, which would work out to around 280 4" tank drive equivalent. But the gearing on that would be somewhat annoying and bulky, it would be more space efficient at that point to just run 200 rpm on 4", and while I’m normally a fan of 3.25" wheels, I don’t see much benefit to using them in this case.
Now, for 2.75" wheels, direct driving them at 200 rpm will be far too slow, and direct driving them at 600 far too fast. easiest gearing for these might be to do a 2:4 ratio gearing down 600 to 300. This could be made very compact, and would get you around the equivalent of 290 rpm on 4" tank drive.
So there’s a lot of options and choices to make, but direct driving 4" wheels at 200 rpm definitely seems like a perfectly good, simple option.
11 Likes
Thanks @Xenon27 . What about 257 RPM (Blue, 36:84), 4" Wheel X-Drive? Do you think that would be too less torque? What do you think is an ideal ft/s for Spin Up?