There are many teams with rotatable shooters (118, 2910, and 254, to name a few), but none of them can rotate continuously. Wouldn’t implementing slip rings dramatically improve performance?
This is an identical post from 20 years ago.
I don’t believe they’re illegal, but getting one that has enough connections, rated for the current load and reliable would be quite expensive.
3 Likes
Once people have experience using slip rings in a harsh environment, they make every effort to avoid slip rings.
7 Likes
For a long time they were illegal. My understanding is they currently are legal, but they mustn’t be of the mercury wetted variety.
But these days it’s not worth the trouble in most cases. We’ve either made coaxial drives (Modern Swerve), designed the device to never cross a set point when rotating from point A to point B (many turrets), or use a programming algorithm that “unwinds” the device if it has rotated too far (some older non-coaxial crab/swerve drives, such as 1986’s 2010 drive).
3 Likes
Maybe in theory, but I’d argue “probably not, at least historically”.
Building in a slip ring removes the need to perform an “unwind” operation if the desired angle passes through the hard-stop.
The cost of this operation is just whatever time it takes to swing the turret around. For fast turrets, my casual observations say this is a 0.25sec operation on average. In this sense, a slip ring might take 0.25 sec out of the cycle time.
Let’s assume a fairly fast paced game where the cycle time is 10 seconds. The slip ring represents a maximum possible cycle time improvement of 2.5% - not nothing, but I’d argue not dramatic either.
But it’s also important to consider that there’s going to be driving time during the cycle. If care is used in picking the driving path such that the orientation of the bot at the scoring position will be known, that turret-swing-around can happen during the drive portion of the cycle. A slip ring doesn’t help in this case, because there was no “cost” to moving the turret and the drivetrain simultaneously.
Note the complexity is nonzero too. Slip rings (or at least some of the wires) have to end up concentric with the rotational axis. And, for most current FRC designs, that’s also where the gamepiece flows through. Not impossible to design around, but not super trivial either.
Also, Anything that can handle both 40A+ and CAN-bus frequency signals is gonna be veeeery pricy.
Between overall complexity and $$$ associated with a slip ring, it’s pretty reasonable to me that most teams shy away from them. The juice just isn’t worth the squeeze.
But these are all assumptions based in history. A super-fast-paced game next year might invalidate those assumptions, to where it suddenly makes sense to implement a slip ring.
8 Likes
Most slip rings are only rated for something like 6 wires at 30A or 3 wires at 40A. Powering two shooter motors and CAN comms would be difficult without spending a lot on a specialty slip ring with a lot of high current and low current wires, or stacking hollow bore slip rings.
CAN-FD probably won’t live through a slip ring (2.0B is fine though). Same goes for Ethernet for a LL.
Mounting it would be difficult, as it would need to be above your entire shooter assembly to work properly.
Just seems like a lot of money and hassle overall, even if it would be cool.
2 Likes
I haven’t looked at the specific rules on this in a couple of years, but my impression has been that slip rings were fine for electronics like sensors or lights but couldn’t be used with motors due to how the motor wiring rules are worded. This may not be the current rule wording/interpretation though.
That said, in general many teams with turrets just use cable tracks and limits to spin the turret back around the other way. Given the cost of large, high-current slip rings that you’d need for motors, it’s probably a lot cheaper to avoid them too.
I’ve played around with the idea of making a turret that works kind of like a swerve drive where the motors are mounted fixed on one side and power is passed through to a shooter wheel via a shaft that spins with the turret, though it would likely require a lot of custom machining and heavy parts.
1 Like
I think 6907 found a legal slip ring for their custom swerve, but it is only for one motor. However, just like Swerve, I think it is possible to remove all motors on the turret to make it continuous, and I remember seeing a differential turret from somewhere.
1 Like
Finding slip rings that can carry the current necessary for shooters or drives that are small enough for packaging, rules legal, and affordable is a challenge.
Cable wrap and fixing that in software saves a lot of slipring headachs and heartaches.
For mechanisms that are small enough to use the little COTS sliprings: generally they don’t benefit from the continuous rotation, or there is an alternative design path to avoid the complexity.
So basically, in the current paradigm of FRC, slip rings have little value due to time investment, complexity, and rubustness tradeoffs.
Having said all that, if TTB wants to magically source a 10" ID 12" OD that is less than a half inch thick and can carry 40 amps on 4-6 channels, my wallet is yours.
2 Likes
Being someone who actively participates in a robotics competition that heavily uses slip rings, you don’t want to use them for FRC.
The problem you will find is that for a shooter, you really want at least 2 pairs of 40A wires and 5 pairs of signal wires (2 pairs motor power, 2 pair can, 4 pair ethernet), where 40A slip rings are expensive and hard to find. FRC does not allow you to combine lesser rated wires for higher current ratings, so you must buy the expensive slip rings with 40A disks. Except that slip rings with these specifications are hard to find. You can try looking around on moflon or senring, but the last few times I checked, they don’t stock slip rings suitable for FRC. And for that matter, even if you do meet the electrical specs, I doubt they can withstand 100Gs of deceleration.
Additionally, the packaging doesn’t work out in FRC, because most turrets wants to send gampieces through the center of the turret, which is where the slipring also wants to go. There are ways around this, but they are not light, small, or easy.
For signals, you can find little Rx/Tx units that use one color of IR light in each direction to transmit Ethernet. You couple them with a clear plastic light pipe and arrange things so that the two ends are right next to each other, almost touching. This has to go in the exact center, but it’s a proven way to send Ethernet across a “slip ring” (this is used in some LIDAR units, others use inductive coupling or RF). In FRC, you could do all this as a custom circuit.
Now, the problem is power. With FRC rules, you need one pair of conductors for each motor/actuator. Plus, you need power for the signal transfer electronics. Solvable, but not generally worth the trouble. Even then, you might want to add capacitors to help cover for little glitches in the smooth transfer of power, but this would probably run into rules prohibitions for motor controllers, etc.
For something like a LIDAR, you can make a toroidal transformer that has the two halves separated by a small gap and mount one half on each side of where you want to allow rotation. So I think there’s a fairly elegant proven solution that could work for everything but motors/actuators. But at this point, it’s not clear it fits with the needs of anything in FRC, short of a custom LIDAR 
.
1 Like
They do make network rated slip rings. My employer has several turntable-based tooling setups that run comms through a slip ring. All use CC-link (Mitsubishi’s flavor of RS485, used for remote sensor inputs and solenoid control outputs), but (at least) two of them use an ethernet connection through the ring as cc-link specific slip rings are hard to find and the ones that do exist (from what I’ve seen in the past) do not offer any non-comms connections (for power delivery, 5 24V connections minimum in this particular case).
Said units are a bit bulky (around the size of 12oz pop can) and I don’t remember if they’re mercury wetted or not (if they are, not legal for FRC). And I don’t remember how much power they can pass-- IIRC they’re 16-18 gauge wires, fine for 24V control and solenoid power but probably not so much for a twin-motor high-power shooter.
For Rapid React (2022), the Bionic Black Hawks (2834) designed and implemented a continuous rotation turret that uniquely used a slip ring.
The design mounted a limelight at the center of a slip ring that had 8 wires to carry the network/POE to the limelight. The limelight would continuously track the center goal independently of the turret.
The turret could continuously spin, and only had a fixed deflector to direct the shot. This limited the range you could shoot from, but it was accurate within a 15-20 ft range by varying the speed.
The limelight slip ring turret had the hex bore encoder to read its position. What was left to do was program the limelight turret to stay locked on the center goal, read the encoder position, and then sync the shooting turret position to the limelight.
Unfortunately, the reflective target, shooting while moving, processing speed was too slow and the shots would miss enough so the drivers preferred manual mode. I think the current co-processors and April tags would make this system much faster.
All of the comments and warnings above I 100% agree with and agree
with our research. The 8-wire slip ring for ethernet/POE from Amazon ($20) is not spec but it did work. If you search for a RJ45 ethernet slip ring, they are around $250 dollars.
Dave Polito
4 Likes
I’ve been surprised that nobody has adapted automotive alternator slip rings for FRC power applications, since they are very inexpensive, very high current, and very available (as “alternator rebuild kits”). Is this sort of not-quite-COTS not-quite-CUSTOM thing disallowed?
I think the trouble lies in the number of contacts. A quick Google shows these to only have 2-3 contacts?
So perhaps enough for power for 1 motor - what about signals for a bldc motor? Or any sensors? Other motors on your rotating mechanism (shooters may have 2 motors, and indexing mechanisms).
I think simply the # of electrical signals (in addition to long term signal reliability), is what’s holding slip rings back.
You’re looking at R623 for an answer to that. Also R622 (for sizing/rating), plus as I recall automotive stuff is often frame ground (a massive R611 no-no).
There are ways to build custom slip rings without mercury (the other typical death knell for slip rings), but those have other issues including getting the ratings correct.
Perhaps you can use coaxial slip rings, where you use a through-bore slip ring with rings rated for high current, and use a cheap slip ring in the center with however many signal pairs you need. 24 wire signal-only sliprings are cheap on moflon, and even cheaper on taobao, however, I’ve not checked the price on through-bore power slip rings, and chances are, they are pretty expensive.
Even still, doing that is really annoying and doesn’t solve the packaging issues (slip ring and indexer no worky), or the fact that these things aren’t rated for 100Gs of deceleration.
In a branch circuit from a circuit breaker to an actuator, 2024’s R623 explicitly allows “COTS slip rings”, ostensibly as passive conductors (a defined term). They’re not wires (to which the gauge specification in R622 exclusively applies) and they’re not custom circuits (although they might have been in some previous years’ rules).
What’s more, R623 is not necessarily exhaustive. It has vague criteria like “appropriately gauged/rated elements”, and refers to “COTS” in 3 of the 4 examples, but ultimately, membership in the set of “intermediate elements” requires interpretation.
I would say there is a substantial risk that a team and the inspectors won’t see eye-to-eye on the meaning of the rules, and that that is possibly more of a disincentive than the actual performance limitations of these devices.
The specific slip ring type noted was a COTS kit. That makes it more gray area than you might think.
The question to be asked there, in my mind, would be “does the manufacturer provide sufficient instructions to build this to the proper rating?” If so, I’d put it as more COTS than not. If the manufacturer shipped a box of parts with no instructions, I’d be more skeptical.
The second question, and probably more important, is “How would a team show that their slip ring is correctly rated?” Completely ignoring gauge spec and COTS kits, if a team shows up to competition with a homemade slip ring, how do they show it’s appropriately rated? Do they run the current through for 5 minutes and hope it doesn’t smoke?
I suspect the difficulty of assessing the “appropriately rated” part is at least some of what keeps the usage of slip rings low. (That and the mercury note.)