FRC Super Servo

Hey all,

I was researching BLDC servo drives and was coming across very few robotics-oriented servos for cheap .HEBI Robotics (https://www.hebirobotics.com/) has begun to get into this but they are focused at a more University market. Has anyone put serious consideration into developing a servo drive for FRC? This would be a large gearmotor, encoder, motor control, low level motion control built into one cheap package (~300)? For comparison a versaplanetary with 775pro, a few gear stages, and an encoder stage is in the ballpark of $150, plus an additional $75 for a motor controller, totalling ~$225. .This is essentially abstracting away the design and control challenges a team faces to a black box, letting them focus more on the creative and novel aspects of robot design and programming. Solid controls was a huge challenge for my team, both in designing a mechanism that is robust and straight forward to control as well as perfecting the control algorithms.

What do you all think? Would this be of interest to the FRC community

1 Like

Unless FIRST changes the servo rules, this would be illegal on price alone since it’s not below $75. I think the idea of what is essentially the Limelight of drivetrains has a lot of potential though.

Yeah it would definitely need approval from FIRST. It could be great for getting a drivetrain together or more complex mechanisms.

I think we have to look at the why here. Is the reason less parts? The key behind the versaplanetary parts is the versatility. You can make them work for pretty much anything. Is it the simplicity of programming? Talon software is pretty good. I think with available parts you can make some pretty awesome actuators.

2 Likes

Simplicity of programming I think is one of the biggest benefits. The talon software is great, yeah, but why don’t more teams take advantage of it? From my time in FRC I saw very few teams using sophisticated controls. This I believe is the fundamental building block for robotic systems that move, so by not having that first block you make building more complicated systems (like autonomous routines) difficult.

2 Likes

So why not seek to educate teams about the great Talon software rather than add all new, untested software? What’s to get teams to use new software when they won’t use what’s out there?

1 Like

As others have said, unless the servo rules change it’s unlikely to be legalized. Even then, if you found a high torque low power servo like this, the fact it’d have to be powered from the roboRIO or a REV Servo Power Module makes it essentially useless.

I do wish they’d adjust servo rules, but I’m not sure any changes they’d have would jive with making them easier to use / better.

2 Likes

I spent a few minutes trying to figure out how to do this as a “kit” made of FRC-legal parts. Allowing it to be powered from the PDP, the missing element is the ability to do closed-loop control with a PWM motor controller. You’d essentially have to create a motor controller which runs in “position mode*” and get it approved by FIRST. If you wanted it controlled by CAN, you could do it today with a kit including an SRX or MAX and a class to wpilib or a similar library for labview.

* or could be configured to run this way through a jumper or USB or such

1 Like

A large proportion of teams don’t even know what a PID loop is, let alone how to use one on a Talon. A large servo for simple tasks like actuating intakes or loading balls into a shooter would be a boon for many teams.
Stay tuned for some “super servo” tricks… I’ve been looking into this for a while now.

3 Likes

I was thinking along these lines with this thread… with the pairing of a NEO and a MAX, you just need a gearbox and you’ve got all the hardware. Then it’s just a question of putting together a software package that teams could use to make the PID aspect of a drivetrain easier to set up and tune.

Kinda like this:

This is the actuator that did such tasks as moving the arm that handled the cargo and hatches.

775 and a Talon and that was way more power than needed.

2 Likes

I think a point not mentioned here is having your motor, gearbox, encoder, and controller board tightly integrated in a small unit reduces wiring and mechanical complexity making the system overall more robust and easier for teams to use (less complicated diagnostics/troubleshooting).

2 Likes

This could even be a kit to give/sell to teams. You learn the design principles, control principles, and how they intertwine as you build and program it. It would be good to shoot for no more than $300 per actuator. Cheaper the better. You have to make it robust enough so that teams can use it multiple years, buying essentially a pile of actuators and reusing them every year for different robots, but also cheap enough that the startup cost isn’t debilitating for smaller teams without a lot of resources.