I thought that was an X-drive phenomenon?
Are you telling me that my 2:1 gearing on my Mecanum wheels is actually 3.8? Because if so, that’s very bad. I don’t think we’ll be moving well at anything over 3:1.
And as for your problem with cutting metal, you have to. Otherwise nothing is going to work. I refuse to drill, because that means what I built isn’t square, but you do have to cut your metal down to the correct height.
that’s the phenomenon of any drivetrain that relies on cancelling vector forces. So that would affect mecanums, h, and plus as well.
I’ve heard mecanums aren’t exactly angled at 45 degrees, and its more like 60 degrees, which explains why many robots cant strafe as well as X. I can’t confirm it though. If you did ignore it, your drive is probably geared at 1:2.82 speed right now
Personally, i try my best not to cut. the only ways around i’ve found are folding systems
Can someone who is for-sure on this point chime in with the official numbers so I can do the math? And possibly tell me if I’m an idiot? Because if we can’t drive (I haven’t actually hooked a Cortex up to test anything yet), I’d like to know now so I can rebuild.
This year, we invested in a band saw because of the amount of cutting we needed to do. It’s awesome. I was using a hack saw or handheld radial saw last year, but this thing is amazing. I would recommend one to every team. It cuts through aluminum like butter. Steel takes a little longer, but still works fine. That and a grinding wheel have made my life much easier.
You really haven’t cut anything for Toss Up? How are you getting under the barrier? All of my uprights are 10.5" long, cut down from 12".
I am pretty sure that when going forwards, mecanums go at the same speed as regular wheels 
~George
That’s what I had thought I read earlier. And they actually go slower, slightly off-center when strafing.
if it does, then i’m going to be extremely confused. the first thing i did when getting mecanum wheels was roll them and note that they didn’t move straight like normal wheels, they moved diagonally.
You know what, there’s a really simple test I can do.
We have two bases built. One is 2:1 Mecanum, the other is 2:1 Omni. The Mecanum one has my upright on it, the Omni is steel. I think they’re about the same weight.
I could go in tomorrow with a video camera, put it on a tripod, put an autonomous routine on both robots to drive straight, and start them at the same time. Video the entire thing, and then analyze it.
If my Mecanum base accelerates super slowly, Draco is probably right, I’m at 3.8:1 or something stupid, and I have to rebuild. If they move about the same speed, I’m going to assume that I’m right, we have a 2:1 base, and everything is fine.
Does that seem logical?
That’s what I thought, but the team usually uses 1:1 gearing so I may be wrong.
That sounds like what an engineer would do, test a hypothesis by experimentation, I look forward to the results.
not really sure if acceleration is the best bet. especially since i’m trying to grasp calculations for acceleration. if accelerations are nearly identical, i would wait a bit longer to see if top speed is different.
i think i need sleep -_-
There is no reason for mecanum-based drivebases to have any different speed going forward than a normal omni or traction wheel drivebase. The 1:1.41 phenomenon happens with X-drives only because they have 2 sources each of diagonal force acting towards 1 direction. The phenomenon is explained here: Strafing (video games) - Wikipedia Look up straferunning within that page.
Since the mecanum drivebase going forward only has forward force (not diagonal in any way) It will only go at the regular 1:1 or 1:2 speed of a normal drivebase with omnis or regular wheels.
Furthermore, if you examine a mecanum drivebase closely while it is moving forward, you will see that the individual rollers on the wheels are not moving anywhere.
TL;DR mecanum drives do not “straferun” like X-drives do, and run at the same speed a normal drivebase would with omnis or traction wheels.
That sounds like a great idea, and a fantastic resource for teams in the future when learning about mecanums.
If you watch some videos of robots with mecanums and read some reveal threads with specs, you will probably see there is no real difference in the forwards/backwards speed.
Anyway, this is getting rather away from the topic of de-scoring. Could a mod please move this discussion on mecanums to a new thread? 
~George
No. We sat down at the beginning of the year and decided to go for the fastest scoring possible. Our priorities (in order) were:
Able to pick up Buckyballs and place them in the cylinder
Speed, both driving and lift
Able to pick up Large Balls and place them on top of the cylinders
Hanging
Strafing
We decided to ignore throwing game objects and descoring.
That’s the reason for the design we chose. It’s basically a Gateway robot. 2:1 mecanum drive (which may go up to 2.4:1 later), 4-bar arm, side roller intake. It’s fast, it’s efficient, and to be honest not very creative. But it should work for what we want to do, which is to outscore you in the first 30 seconds of the match.
As for close to done, I’m waiting on some parts. I can’t finish my intake at the moment, and we’re actually out of 12" axles. We should have all of the parts by Wednesday or Thursday, which means we should be done building on Friday. With any luck, we could start driving practice early next week. The four or five Autonomous modes we want should be written by the end of the month, and we can have Programming Skills finished by mid-September. Our first competition is in early November, so if all goes well we should be finished in plenty of time.
A big mistake i’ve made multiple times in the past is failing to make my ideas concrete and surviving off of assumptions. Sure you can mess around with theory, dream all you want, but if you don’t step forward and actually figure out ways and writing out steps to succeed, nothing’s going to happen.
First off the challenges of the de-scoring system
-the ability to reach deep enough to descore. The barrier makes this difficult to de-score 2 or 3 balls, let alone 5. Plus due to the weird height of the barrier(12 inches) many teams are forced to cut metal and spend money, or combine structures
-Alignment. Especially with defense, clear goals, packed field, time clicking down, battery fluctuations, instability from metal, and much more, this is extremely hard to do without precise engineering
-getting out of the way of other manipulators. The goal is to fit in, and pull objects out. if this is placed below your intake, it would catch your game pieces. if it’s placed directly above, the goal gets in the way. because of this, it needs to be out.
-dealing with overlapping objects. Objects aren’t perfectly on top of one another. that means the manipulator needs to adapt to changing conditions as it travels down
-getting rid of pieces. as manipulators get more and more passive, and as they become more adaptive to other game objects, it may be difficult to release the very last ball you manipulate
-rulings. the restriction not to pass the perimeter, combined with defense is a minor threat, however it is a threat nevertheless
there’s probably more issues de-scoring systems need to face, but currently it’s still quite difficult to do.
how did you get a perfect 2:1 ratio with the mecanum’s natural speed ratio of 1:1.41?
Done and done. The original thread can be found here:
The reason why i find this hard to believe is because when you take individual mecanum wheels, and roll them, they don’t roll straight, they roll diagonally. That’s a really high indicator that there are 2 sources of diagonal forces acting towards 1 direction
If speed vectors were moving perfectly forwards, or perfectly backwards, it wouldn’t have the ability to strafe like other holonomics because it wouldn’t have any vectors pointing in that direction.
However you mentioned that rollers on the wheels weren’t moving. I’ve never observed my wheels as they drove. This point, it’s hard for me to visualize things. When moving diagonally, 2 rollers would have to spin, or else a lot of friction is created. However, because i know when you roll wheels individually, I would hypothesize some forces are cancelled so the rollers would have to spin.
You may want to read this thread.
https://vexforum.com/t/mecanums-lets-settle-this/21598/1
Not sure we ever did figure out when the rollers were moving.
Alright. Test complete.
The Mecanum move so close to the same speed as Omnis, any difference comes down to weight. The “vector forces” or whatever on an X-Drive don’t apply here.
And, on the bright side, I’ve got a robot that’s balanced we’ll enough it strafes correctly. So, that’s cool.
Only the X-drive has the effect of a 1:1.41 ratio moving forward/backward and strafing.
H-drive and Plus Drive are essentially tank drives with strafe wheels and they will perform as such when moving forward, backward, or strafing. When these kinds of drives move diagonally, at a perfect 45-degree angle, they are essentially X-drives with a different chassis configuration, so the effect does apply when moving diagonally, but not when moving forward, backward, or strafing.
Sorry, I didn’t clarify that part:o That is indeed true
That’s quite unfortunate for me though.
In the past 24 hours, all the robotics physics i thought i knew just got flipped on it’s head. And the robot I have right now isn’t geared as fast as i want it to be. At least it’s clarified now, just kinda sucks that summers over so i wont have as much time to recover.
Are you going to post the video?
I just had an idea to test out what the vector angles are on Mecanum drives when they strafe.
If we built a chassis on which the front left and back right wheels were powered mecanums (with proper roller direction as though they were actually part of a mecanum drive) and the front right and back left wheels were unpowered, free-spinning omni wheels, then the direction that the chassis moves when both mecanum wheels move forward should be the true angle that they apply force in when the rollers are spinning, correct?