This is just a proposal but I have seen dual motor differential swerve from a few teams this year and high tide had a fantastic 2 drive 1 azimuth swerve setup which worked incredibly and I’m wondering what the mechanical feasibility of a tri-motor differential swerve would be, coming from a programming stand point I don’t see any issue with it and electrically just have to be careful for brownouts, so I was wondering if this idea is doable mechanically.
What would the point of this be? Diffy swerve with 2 motors is already questionable as it is…
2 Likes
Push ever more power into the carpet for pushing, mostly to better support playing defence as a swerve bot because even two standard 2 motor coaxial swerve bots struggle to push a tank bot facing sideways at all.
Also you know, a challenge to find the limits.
This is simply not how traction works.
10 Likes
2 drive motors per module is already taxing enough on the battery. 3 would be just begging for brownouts
1 Like
I’m aware, High tide managed it with some current constraints and using Talon FX controllers you can do some very low level voltage and current constraints to avoid brown outs. It’s not a worry but 3 motors per module clearly provides a benefit competitively.
1 Like
Mechanically or design wise, I’m struggling to see how three motors could achieve a differential swerve aside from just packaging two together for more power.
8 Likes
That’s what I was thinking about but I know it’s not my area of expertise but I will say having two together sounds like a terrible idea at least programmatically because they’d have a much easier time spinning faster causing the module to rotate or being capped at half power defeating the purpose.
1 Like
The extra power means nothing if you can’t use it. Burning holes in the carpet will not go over well with the FTAs
Also I am curious if the OP understands how differential swerve works and the mechanics that make it possible to control (chiefly mechanical symmetry)
6 Likes
That would be my thought on why differential swerve with three motors would be very difficult, if not impossible. On the other hand, my understanding of differential swerve is that when going straight you effectively have the power of both motors driving so I’m thinking you’d probably already be traction limited.
The more interesting use of those 4 extra motors if you really want increased pushing power might be 6 swerve modules. This also opens up the opportunity to really mess with people by mounting them to a KOP drive frame at the wheel positions of a KOP drive.
2 Likes
I’m looking to design a dual motor diffy swerve module. Would you mind telling me what teams you saw running diffy this year?
At the beginning of every match, you can call thirdFalcon.set(ControlMode.Mitosis, 0) to end up with an even 4.
4 Likes
A three-motor differential swerve module is going to have the same mechanical design roadblocks as a three-motor differential drive. How many of those do you see?
Internet points to anyone who draws even an abstract linkage that does this sensibly.
2 Likes
whoops
Could it be done? Sure, for varying definitions of “sensible” and “works”.
Should it be done?
Not in FRCFIRST Robotics Competition any time in the near future.
@OP the math you should do, which you can use to answer your own mechanical question:
- For a 150lbs robot, what’s the maximum force that wheel/carpet static friction can sustain without slipping
- Assume SKS Mk4i L3 swerve drive modules, with infinite traction. What’s the maximum force it will apply to the ground with wheel-driving 1 motor? 2 motors? 3 motors?
- What’s the motor-count crossover point where adding more motors doesn’t actually add more pushing force (forces become kinetic friction, not static)?
- How many amps does that quantity of motors consume at stall? How does that compare to the MK Battery’s discharge curve? Where would you have to set the current limits to achieve peak power (W = V*A, but battery voltage decreases with increased current)?
What I think you’ll find when you start running the numbers - with current FRCFIRST Robotics Competition technology - more than two motors comes with lots of caveats for marginal performance improvements.
25 Likes
technically yes, but to bring to an actual thing is almost impossible for a couple of reasons.
as a person who tried and keeps trying to develop a differential swerve with the controlling aspect of it is a nightmare you either end up with 64 different values to calibrate or with a state space module which most people don’t know what that even means.
now out of the differential part of it, giving enough power for it is one big challenge even with PDH having more slots you still have a big problem if you have a complicated season where you need enough slots in the PDH and the power demand is out of this world
1 Like
This might be fun project. What are you trying to solve with this tri motor diffy swerve? what are your spec expectation?
We can’t even buy Falcons right now and you are proposing insanity!! Preposterous!
I love it.
If only we had access to motors.
please don’t refer me to neos…sparks are a nightmare
3 Likes
5 Likes
Oh I’m well aware of the tuning challenges and I felt like taking on something new after building H-Drive systems from the ground up and trying coaxial swerve simulations with basic physics, I’ve already got an entire tool suite for tuning just for it, playing with current limits, spike limits, and voltage limits using the low level interfaces on the falcons should make it doable just a bit difficult. I’m really only concerned about how mechanically out of hand it could get to fit everything together, and I feel a little bad for all the batteries.