I’m a FIRST alum and am now part of the robotics club I started at my university. We just started a multi-year project that will hopefully result (many years from now) in making a full-sized vehicle operate autonomously (think DARPA grand challenge).
This semester, we’re building a very basic small vehicle to use as our first-phase development platform. The chassis is going to be about 24" x 16" and the total weight of the vehicle/robot should be around 30lbs. We’re wanting the vehicle to be able to drive up to about 6 mi/hr. Here’s a very (very, very) rough concept design.
We’ve been looking for motors and gearboxes for our vehicle and it occurred to me that the ones FIRST gives out in the yearly KOP might work really well. What do you think?
If you think that’s a good idea (or if you have other suggestions), any advice on where to purchase from?
That’s a good start. However, you may like looking at either swerve or holonomic drive because that will give a much greater maneuverability. I’ll post an image of my VEX Holonomic robot tomorrow.
Also, what is the gear ratio and what motors are you planning on using? Have you planned that out?
What control system are you planning? A cRIO system may take up all the thirty pounds if not designed correctly. There are many parts, a lot of which are quite heavy!
I agree it would get a little complicated, but if they like the idea, it isn’t even close to impossible!
Nathan,
If you can give me a tad more information, like:
Motors you are interested in
Wheels you are interested in
Battery voltage
and a few more details, we can help you design your drivetrain, and calculate gear ratios and find the correct gearbox
They are scaling up to a CAR–like the DARPA Grand Challenge or a Google whatever-they-call-it. No car on the market today has a holonomic or swerve. None. (Besides maybe a self-park feature, but most of that is done with the car steering itself.)
Therefore, a driving requirement of the design is that it has to scale to a CAR. Leaves Ackermann steering–and the basic design shows their steering mechanism.
I was thinking a Banebot with a 4:1 or a 16:1 ratio- Though considering my suggestion is based off mere speculation, it would be a convenience to then drive an output shaft by way of chain or direct gear contact to allow for a change in ratio (either up or down) if you find you need more power to move the load, or want a higher velocity.
I have been working on a spreadsheet to help with motor and gearbox selection. It may be useful for you. The final speed calculations are optimistic, but relative results should be reasonable.
Suggestions for motors are great and all, but may not be the best for the application. An FRC setup will most likely work for you, but I urge you to define the requirements.
What are the major areas of functionality and concern? These could be things such as maintainability, maneuverability, reliability, durability, etc. For each of the requirements, outline an objective (what you really desire) and a threshold (the minimum that you will accept). The space between the objective and the threshold is the trade space - you will use this to determine if your solution is workable.
Here are some of my first thoughts on requirements:
So far, I have two “requirements”: 30lb weight and 6 mph. Do you have a maximum weight allowed? A minimum speed?
What acceleration is required? This will help define your motor HP and gearing/transmission requirement. It will also help define your electrical power requirement.
How will you get electrical power? Internal combustion engine with a generator-inverter setup? Battery? Fuel cell? Solar? This will help define motor power limits or operating time.
What is the vehicle operating time? How long will it be using full power? What other electrical loads are required?
What ground clearance is required? This will help define wheel diameter and relates to the HP and transmission requirement.
Since 6Mi/Hr is a medium jogging speed, I believe BAG motors with the versaPlanetary gearbox of 10:1 may work. However, this is solely based off the wheels you use! BAG motors run at about 14,000 RPM, so that means that you will get 1400 RPM, which seems great for that sort of speed. You probably won’t drive too fast. The motors should easily achieve more than 6mph!
How about a miniCIM in a versa-planetary? It is surprisingly trivial to modify a CIM to work, all you need is a hacksaw and a tabletop vice. On a 6" wheel you would need a 15:1 gearbox, which would yield you 7MPH and plenty of torque for a 30lb machine
I am a fan of the MiniCIMs. A full sized CIM can draw over 100A and the FIRST ESCs are all rated ~60A. The spreadsheet I posted has a pulldown option for MiniCIM plus you could always add other motor curves. MiniCIMs with 6" wheels and 12.7:1 ToughBox Nano is probably much more than is needed. Bag Motors could also work, but I have not done the math.
Any motor can pull a ton of current if you stall it. If you are actually pulling 100A consistently you are doing something wrong.
I agree that Bags or Mini CIMs are your best option here. Everything else does not handle heat well(which for drive would be awful). However gear ratio again depends on your wheel size.
I’d go with either CIMs or MiniCIMs. BAG motors have such a high RPM that they self-destruct their internal bearings. That is the reason why our shooter needed to have it’s motors replaced so often. Also, if you aren’t stalling the RF-775s, they are fan-cooled so they shouldn’t get very toasty under normal working loads. CIMs are able to cool off all right because of the surface area and size of the motor. BAG motors just overheat all the time!
I agree that pulling 100 amps continuously from a motor means that you are doing something wrong. That is nearly the max threshold of CIM windings!
I doubt speed was the root cause of the BAG motor failure. Improper side-loading on the bearing due to a shaft-mounted shooter wheel or an overtensioned belt was likely a root cause if the bearing itself failed.
I believe bag motors had an electrical problem, not a bearing problem. 550s for one run much faster than them so I would think that the Bag bearings would be fine.
We have found that fan cooled motors tend to heat up regardless of what you do. Both our shooter 550s(100% at 10 seconds at a time) and our drive 550s(constant starting and stopping) all got really hot. I personally wouldn’t reccomend fan cooled motors for a drive application unless you are thinking of using another fan to cool them.
I think that a bigger problem is your Bag motors is your application. You have many CIMs working together to speed up your robot. On the other hand your shooter ratios may not be ideal so you may be working your Bags extra hard.
Thanks for your replies, everyone, we really appreciate your help!
Here’s a little more information to clarify a few things:
We’re wanting to use normal wheels (not a holonomic drive) and steering like on a car
We had a mechanical engineer give a presentation to our club and he also recommended Ackermann steering, so we’ll be trying to implement it somehow with a servo
We’ll be using a cRIO for the control system
We estimated that the weight should be around 30lbs, but to be safe we should probably plan for twice that, or 60lbs
If we end up using FIRST motors, then we’ll probably get one of the standard batteries that FIRST robots use (12V motorcycle battery?), as long as that doesn’t put us way over our weight limit
Wow, that looks like it might be really helpful. Thank you!
Thanks for your help! Here’s some more information:
The weight we’re shooting for is 30lbs (not sure how realistic this is…), so let’s plan for up to 60lbs
The maximum speed needs to be at least 4 mph, but we’d like it to go up to 6 mph
I’m not sure on acceleration, but it’s not a huge issue since speed isn’t our primary concern
We need it to be battery operated. What type of battery we use mainly depends on how much power the motors draw, so we don’t know what battery we’ll be using yet
We’d like an operating time of at least 20 minutes of running non-stop, which would hopefully translate to at least 40 minutes of on and off use. Obviously a longer running time is preferred, though!
We’ll be using a cRIO as the control system and will have a couple of low power sensors and some kind of camera (possibly an Xbox Kinect)
We need at least 2" of ground clearance, so we were planning on using 6" wheels
So far we’ve had these motors recommended:
Banebot (possibly with a 4:1 or a 16:1 ratio)
BAG motors with the versaPlanetary gearbox of 10:1 (but they might be bad, according to a few)
miniCIM in a versa-planetary
Where do you all buy your parts from? Is there any particular place you recommend?
Standard disclaimer The following is meant to be a guide, and by no means is an exhaustive list of what needs to be done.
I understand your objective weight is 30 lbs, but that’s really hard to achieve; 60 lbs is more reasonable. I would suggest using this number, assuming no mechanisms or devices are mounted to your vehicle. 30lbs will likely be your drive chassis, wheels, and transmissions.
Those are good numbers on speed. You can take these (6 and 4 mph) to determine the velocity against the ground required at each wheel. This speed will be used to get a gear ratio with the motor.
The reason for acceleration is to find horsepower (think in terms of Work). For now, you may want to make an assumption - maybe 6 mph can be reached in 6 seconds (just make something up). The horsepower rating is required to size the motor needed to accelerate your vehicle in a given time. In conjunction with the scrub force on the wheels and your wheel size (width and diameter), calculate the torque required for each mode of operation (forward, reverse, turning). This will allow you to size the motor get torque.
From here, build a “mission profile”, where you lay out how the vehicle is intended to be used: turn on with a systems check, 10 seconds idle, 15 seconds full power acceleration, 85% cruise for 10 minutes, deceleration to 0 mph in 10 seconds, accelerate to 3 mph for 4 seconds at 100% power, drive an 8% grade hill while maintaining 4.5 mph, accelerate to 6 mph at 85% power, etc. This is a simple bookkeeping method that will allow you to size your battery. Using the required current draw from this profile, you can verify or select a battery that will provide the proper current for the time you need. For this vehicle, you may be able to use the standard 18 Amp-Hour 12V battery used by FRC, but it has to be proven first.
Some prototyping of the drivetrain and measurement of each motor’s current draw in a loaded system will help validate your initial assumptions or prove them to be incorrect. From here, you can size your battery appropriately.
Is there a reason for the cRIO? Speed of development (LabVIEW), sponsorship? Ruggedness? If weight is a concern, I would suggest driving this system using an Arduino and coding through LabVIEW or the Arduino IDE to save at least a pound and a half. Additionally, the MyRIO should be entering the market soon (if not already). It is the platform that the 2015 FRC RoboRIO was leveraged off of (as well as the NI cRIO 9068). Either of these should handle the Kinect output in real-time, assuming there is a LabVIEW driver built.
