I am writing this for the sake of influencing future rules. I was sad that the rules state only PWM servos, only controlled from roboRio PWM pins. Please consider allowing modern servos to be used in future competitions. No more angle-limited, single-sided, unidirectional-communicating, PWM servos for me, please. I want modern robot-appropriate servos, big and small. I want to be able to do speed control AND position control (and torque control). I want to be able to use the servo as a sensor, get telemetry, know the actual vs commanded position, and measure current draw. I want to chain control lines and have IDs. Please consider modern actuation.
Hiwonder HTD-45H - Crazy inexpensive and easy to run 12V Serial Bus Servo
There are plenty of 5-7V servos from this company too.
Doing it here is 50/50, yes reach out to HQ as well. But I’d be willing to bet a shoe or shoe equivalent that someone at HQ reads these words at some point soon.
The big thing I’d address - talk through safety. How these devices can be integrated to be guaranteed to react to an estop, what level of new development is needed to achieve that… how you would integrate them into the existing electrical system… And, how (with whatever you recommend) would it be easily inspectable to be in compliance and safe.
Weigh that against the availability, and what new capabilities they open teams up to, either technically or from a teaching perspective.
I’ll admit I’m not sure what technical capabilities the links above unlock. For example, the last one is pricey and “direct drive”, but only pumps 0.16 Nm of torque. That doesn’t seem like it would win a arm wrestling match with a hamster. I’m not clear what I would do on my FRC robot with that.
Edit: A thought occurs - is this targeted at FTC, not FRC? Apologies if so, I’m a serial assumer attempting to get better.
Yeah, I was briefly interested in that (for a personal project), because it sort of looks like something that could direct-drive a robot arm joint or something. But it has the same torque as a Neverrest.
OP, I’m loving this. I long for it too. Maybe we will get there one day! There is no real process to make it happen. I’ve been asking for a REV motor to be made legal for 2 years now by emailing them directly and have had no luck.
Post here, get some ideas, and then write an email to HQ, then lobby them at CMP, lobby them at other events, ask some pointed QA questions, and then find the right volunteers to make it happen. It’s a process to get FRC rules changed and posting here is part of it.
I agree with OP. Allowing higher power/cost servos that can leverage the RS-232 bus on the RoboRIO would be very handy. They are common and relatively cheap. That said, apart from Feetech-sized servos, most of the nicer industrial ones have too poor of a power to weight ratio to consider.
The real way to get a motor approved is to get a company to donate a bunch of them. I use Dynamixels for research in my lab, and I think they would have a use case in FRC (e.g. turret, adjustable hood). But I don’t see FIRST going anywhere with them unless they see that a lot of teams want them or they get a large donation of them.
It is supprising hard to find quality direct drive robot actuators. The distinguishing feature is in the low speed precise movement. I understand it takes a high pole count. This is what is in the inside of that actuator
I do have a contact at Robotis. I would probably need a well conceived pitch there too. They were generally receptive to working with universities in the past so this would not be so wierd. I have seen Feetech sales on LinkedIn so I guess we can reach out to them too if we want.
While it isn’t going to happen this year, I wouldn’t give up hope and keep providing feedback to FIRST about this. As FIRST is finally allowing any linear actuator (that’s rated for 12v and 20amp breaker), so we might see better servos in due time with enough feedback.
Buy the goBilda ones instead. They have the same functionality, have about the same price, come in more speed options, and have steel geartrains in them (unlike Rev’s brass) so they don’t explode nearly as easily. They also sell servo blocks which puts a supported bearing on the output shaft which also increases durability and mountability.
Both Rev and goBilda’s servos are manufactured by Feetech, and were designed in partnership with those respective companies. It’s just that the goBilda variant has been widely used in FTC and were tailored specifically for that market (and Rev’s have also been widely used because they are visible in the Tetrix/Rev storefront)
It would be interesting if something could be added to WPILib to reprogram the REV/Gobilda servos on the fly, assuming they use the same protocol. I’m not sure how rules-friendly that is, but it might be useful and add some value to those servos for FRC use.