Scenario:- 2 Falcon motor driving a shooter with different gear ratios. One geared lower to provide high starting torque to overcome inertia and a 2nd motor geared much higher to reach higher top RPMs.
Both in velocity control mode with profiled velocity ramping so they aren’t fighting against each other.
The lower geared motor would have to be switched to coast mode once it reached its max RPM. Then the second motor would be doing all the work up top just like it was a single motor shooter.
Obviously the lower geared motor would now be spinning above the “max velocity”. Issues of concern are the rating of the components inside the motor to handle the higher RPMs and the higher back EMF voltages being induced in the “free spinning motor”.
Has anyone played with this type of setup before?. I know single direction free wheeling bearings could be employed to release the slower motor or you could add a shifting gearbox, but I’m curious if this really basic setup of different gearing is plausible.
If this concept works, could it then be employed on the drive motors of a drivebase ?
The correct answer is to use two Falcons to spin up the shooter to high speed before any need for torque comes along and let the momentum deal with that, or have a 2-speed gearbox if you’re really paranoid. For a shooter, there ain’t enough inertia to stop the Falcon for long.
As far as using it in a drivebase, I would pretty much say “forget it”. If there’s a motor in the drivetrain that can put power to the floor that isn’t doing so, you’re wasting power.
Now, I will say that similar setups have been used in drivebases before, where different motors were geared differently… but they were also attached to different sized wheels to get about the same speed, and this was back before 3-robot alliances when we didn’t have 4 CIMs.
I take issue with this. Modern FRCFIRST Robotics Competition robots aren’t usually power-limited, they’re either traction-limited or current-limited. In the case of being traction-limited, adding more weight is good! In the case of being current-limited,
This means that as few as 5 motors can continuously deliver the maximum power that a legal FRCFIRST Robotics Competition robot is capable of. So using more motors to spread out the load or generated heat, or operate the motors at more-efficient parts of their power curves (to generate more useful work instead of heat), is entirely reasonable.
I get that you would be “wasting” power an the top end of the speed curve, but couldn’t you argue that the bottom end (where it is needed in a robot pushing match) would be substantially stronger if the second motor was geared a lot lower than the top end motor.
The whole idea of this concept is to boost the pushing power at low speeds for ridiculously high geared drivetrains (like 7+ metre/sec) without having to use a shifting gearbox.
This might work if you take care to put the starter motor in coast mode once you’re at speed. Not sure if it’s a great idea overall though; shifters may still be simpler.
It should be pretty easy to try it out though, and there’s nothing about this setup that poses any real risk to the hardware. Give it a try and write a whitepaper on the results!
The catch is that you’re putting a motor spinning at above its max rated RPM. Depending on the motor and its components and how much above that RPM you go, that could be catastrophic.
As a matter of fact, I wouldn’t test it with a Falcon first. I’d use a brushed motor pairing, because they’re cheaper, then check with the cheapest FRC brushless motor I could get my hands on (sorry, Nidec), then Falcon. (I also wouldn’t use Falcons for this sort of application, because a good working Falcon is a crapshoot right now and I wouldn’t want to risk one.)
The other option here would probably be a mechanical clutch/release to release the low-speed motor from the high-speed motor. I don’t say it’s a simpler option, because it isn’t, but it’s less gambling.
Actually, I can think of an issue. The high-speed motor has to turn both whatever it’s turning and the low-speed motor via its gearbox. This would tend to cause some power loss on the high-speed motor. In a drivetrain this could be mitigated fairly easily (the two motors drive different wheels that aren’t connected) but in a shooter application you may end up running the shooter too slow because you’re dragging the second motor as well as the shooter wheel.
Yes, but this downside is extra rotational inertia in the system, however it is also far less efficient rotational inertia than an overdriven flywheel, so … Yeah… Yeah… I’m with ya here.
I was only thinking of using a different number of teeth on the pinion gears on each motor, with both motors meshing with the same driven gear, so it would only be the lower geared motor that the higher geared motor has to over spin.
Anyway, I was asking if anyone has tried it and cooked their motor controller or flung the winding out from centrifugal force. And it appears it’s never been tried.
Hey SDS, is the MK5 going to have a two speed pneumatic shifter? Looks like I’ve got a good offseason project for the CAD team to work on.
I suspect you will run into issues with this. For example, if you end up running the Falcon at 2x its free speed, you’llLimelight, an integrated vision coprocessor end up generating ~24V on the terminal of the motor. This voltage will back feed to the supply. With a 12V battery and reasonable wiring, it may keep the voltage down (i.e. by charging the battery), but if the battery, or one of the cables on the motor controller comes loose, you will be essentially powering the device, and other devices on the system, with that 24V. There are already known issues with various control system components and high voltages like that. For example, the Falcon’s bulk capacitors are only rated to 16V. It seems like the falcon has some kind of protection against this, but I doubt it was tested this far outside of its operating range.
I get the electronics in the motor controller will have to deal with the higher voltages, but how will the “24V” get back through the H-Bridge to the battery if its in coast mode?
Through the intrinsic diodes in the FET bridge. Voltage applied to the bridge that exceeds the battery voltage will pass through these diodes. This current will show up as a braking effect on the motor shaft.
It doesn’t matter that the FETs are off. The diodes are always there. If the potential between U and W exceed the battery voltage + 2 diode drops, then current will flow through SW1 diode and SW6 diode. This will cause the motor to develop an opposing torque as long as the battery can absorb the current which in most cases it can.
One consideration is the body diodes inherent in the motor controller. At some point, you may get some braking effect, even in coast mode. This results is generating electricity, giving it a path to flow, which means you are doing work, and that power is going to either be dissipated somewhere, stored, or both. Much depends on the specifics…
EDIT: Better presentation of the same answer above.
If you want to see these diodes in action, go over to your Prusa printer and turn it off. Once off, push the bed quickly. The screen lights up and the printer tries to come alive (just did it to my Mini to make sure I wasn’t lying). Same phenomena happens to my X Carve Router and it can happen to FRC robots also.
Thanks @Mark_Wasserman That makes perfect sense now. I forgot about the fly back diodes in the H-Bridges.
And that kills the whole idea proposed at the start of this post as even if you put the lower geared motor controller into coast mode, its not going to coast once the back EMF goes over 13.4v (12+0.7+0.7) and that motor controller will be effectively be braking on the higher geared motor with the battery as the sink.
Looks like you have to physically disengage that motor if you want the two speeds. Glad we didn’t try it.
Only problem here is you can never run your spraque motor in the opposite direction (reversing) because it will just free wheel. Although, with Swerve, you could turn off swerve module optimization (ie the never turn more than 90 degrees feature) and always run your Swerve module in the forward direction and rotate the module by 180 when you want to go in the other direction.
But this could get messy when you are fine adjusting a delivery position because the drive wheel moves slightly as you change the steering angle.
