Help with designing turret

our team is considering to build a turret for an Offseason project. I have been placed in charge on this project and I wondered if the turret should be rotating by a timing belt , chain or just by gears.
I would be grateful if you could help me understand the benefits and\or of misuse of which way.

it is my first post here and English is not my first language so I would be appreciative if you could forgive me for that

5 Likes

To be honest, it depends.
On the one hand a gear-driven turret is generally easier to manufacture (especially if you’re making it out of aluminum) though without proper gear design and spacing it can suffer from slop that may affect accuracy at long range.

On the other hand, a belt-driven turret is generally less susceptible to slop, and thus can be controlled more precisely, but the manufacturing is generally harder (requiring you to either buy or 3D Print a large pulley) and requires more precise spacing to keep the belt tensioned. Belts are also potentially susceptible to skipping which can throw off sensors used to aim the turret.

Chain is similar to belts but can be more susceptible to stretching over time, so I’d avoid it personally.

In general I think I’d prefer a gear-based turret just for the sake of simplicity, assuming you have the fabrication capability for it (they can be 3d printed too but you have to be careful of the durability if you go this route). In general most games that benefit from turrets don’t tend to use them at a range where you have to be super-precise in your horizontal aiming, so a little slop doesn’t hurt that much.

COTS is certainly an option too, though we modeled a turret off of this particular design a few years back and had all kinds of issues with it, so it wouldn’t be my first choice, personally.

1 Like

So Ive done turrets twice for 2451. We loved them, I’m a huge fan and I think they have their place. I have seen many great teams like 254 use gears on most of their turrets and they seem to be happy with them. However we always ran a timing belt. I think that chain should be out of your consideration due to the weight and small translation of distance. Timing belts are really cool because they allow you to have very little backlash and not have to worry about gear mesh too much. With that said you almost have to use a 3d printer to get this big gear and print in in multiple sections. If you have a markforged or maybe even bambu ur golden, I wouldnt touch it on any other fdm printer. If you dont have either of those go with the gear mesh strategy. If you are 3d printing gears I also encourage you to use a metal insert to add more strength to your hex bores. You will probably round out the hubs if you dont ( I speak from experience). Aside from that bearing methods are a huge thing. Lots of teams use 1 x-bearing in the center of their turret. Go take a look at 2910’s build thread if ur more interested. Ive always used something similar to what you will see in COTS options like the GREY T Turret, but without the top bearings, we made tiny rollers instead. You could also try a low friction material but its all up to you and your teams recources. Cable management is a big component of it too. At the bear minimum use bigus. At the max use a strategy similar to 2451 in 2022. Hope this helps.

1 Like

My team has also considered a turret as an off season project. The biggest challenge we are having though is wanting a 360deg rotation (or continuous rotation. We are having trouble thinking about the wires and such that would need to be rotated. We have looked at slip rings but those usually require them to be centered around the axis of rotation (to my knowledge at least), and that makes game piece manipulation more difficult and time consuming if its not centered.

As for your questions, when my team was at the 118 open house this year they had a turret prototype out on the table. Ill see if i can find that and post it! It might be really helpful to visualize and demonstrate.

IMO slip rings for turrets are pretty unrealistic given the high cost (and possible illegality due to the wiring rules, at least if you plan to use them with motors). Personally, I’ve played around with the idea of using a gearing pass-through like how swerve drives run their wheels on a turret, so that way you could mount shooter motors low and keep the turret light, but then you run into the need for (large) custom gearing and it becomes impractical.

The best solution is probably to just build a 1-full rotation turret (maybe even allow a few degrees of rotation past a full turn if you’re feeling ambitious) and get creative with programming and vision to manage the loop-back positioning.

1 Like

Having built slip-ring turrets for a non-FRC competition, you do not want to built a slip ring turret for FRC. First, turrets typically want you to feed gamepieces through their center which is generally incompatible with slip rings. Second, you are not going to find a suitable slip ring that fits within the cots rules. Finding slip rings rated for 40A current, especially on 3+ pairs of wires with an additional 5 pairs of signal wires (1 pair can wires, 4 pairs ethernet) is not cheap. And you don’t need infinite rotation because your drivetrain is not constantly spinning when you are trying to score with whatever is on your turret.

2910 used chain and sprockets for our turret. It worked quite well. Here are photos including the large machined sprocket.

The WCP 7" diameter turret also uses chain & sprockets. It’s a bit different than ours, but works well and is certainly easier to build if you don’t have access to CNC machinery. This WCP architecture can be scaled up pretty easily to have a larger diameter if you have the CAD know-how.

1 Like

Here is a video of the wire management prototype many saw at the 118 open house.

This allows us to run all of the wires into the turret for:
5 Krakens + Spare set of conductors
1 Falcon
Can wires + spare
2 Ethernet cables + spare for 2 Limelights
Set of wires + spares for beam break sensors

This energy chain setup gave us 420 degrees of travel and worked perfectly the entire season with no issues with any of the wires going to the turret.

254 and 2910 both had simpler cable management implementations that allowed them to get more travel out of their turrets.

For actually driving the turret we usually use custom waterjet gears that clamp around a single thin section bearing. I posted some more info about this seasons implementation here.

The CAD of our 2022 robot can be found here and has a very similar turret implementation as this years robot. We epoxied all of the gears to the shafts in the turret gearbox to reduce backlash.

7 Likes

I don’t disagree that high power slip rings might not be good or necessary for turrets but there are reasonably priced options from the small-wind-turbine market:

There are cheap Aluminum turntable bearings available on Amazon. The hole patterns are wonky, so mask it when you re-drill them. More slopping than nice thin section bearings, but far cheaper. And they just bolt in place, which is a big plus. No fancy machining for bearing fits required.
We built one based on this with a machined Aluminum gear and the matching 10 tooth 10DP on 1/2" hex steel gear.
I never did get my kids past the “how do we get wires to this crazy thing” part, so it eventually got pinned in place :frowning: Bit off more than we could chew…

The slip ring you listed is not useful for FRC. For most FRC games, you want 3 pairs of power minimum and 5 pairs of signal minimum to run 3 motors, CAN, and a limelight which are significantly more expensive than just a singe 3-phase slip ring.

Imo any slop on turrets are unacceptable, a not stiff turret is almost always worse than no turret and is often one of the main reasons that most turret bots would do better without one, even when ignoring the other numerous drawbacks of turret.

I’ve been wanting to see someone use these in a robot.

or these multi-turn cable carriers :slight_smile: ?

Completely agree. For your first turret, it would be useful to use the WCP (or other COTS) turret which is usually cheaper than making one yourself anyways. This also allows you to focus on understanding the turret design by building something that is already known to work properly, and focus on getting the programming and control down, instead of spending your time focusing on making a mechanically robust system.

If you want to make a custom one afterwards, you can then use the control system you developed for your COTS turret on the custom one. Mechanically, you can design it based on either the WCP one (or whatever COTS turret you decide to pursue), or based of the single large thin section bearing style that 118/254/etc. uses.

In fairness, pretty much any turret slop is negated by mounting your targeting camera on your shooter. If mechanical slop causes it to be off for some reason, you’d see that in the camera and it would just continue to rotate until targeting correctly.

No they are not, and the number of teams with bad turrets (aka building the turret caused the team to perform worse than if they didn’t use one) throughout the years shows why this is the case. System dynamics significantly overpower any methods of trying to work around slop, with the only way of taming them is to eliminate the slop entirely.

Shooters involve extremely high forces and those reaction forces have to go somewhere. For a poorly built turret, it goes into recoil and deflecting the entire assembly which causes shots to miss. Slop or backlash in any form also makes controlling the turret’s position significantly more annoying if often not impossible because as disturbances coming from the drive train are able to influence the shooter’s position and that position control on systems with backlash is extremely annoying if not possible.

1 Like

Personally I’d pretty much just clone 254’s 2022 turret. It’s driven by a 10DP gear, supported by a big X-contact bearing (source from Lilybearing) and clamped sort of like an SDS modules bearing, pretty straightforward.

Torrence talked about it here.

Also, these days you can actually get a bit better backlash by using 20DP gears (instead of 10DP) and getting a service like Fabworks to lasercut the big gear profile. When doing this it’s a good idea to order 2x 1/8 thickness aluminum and stack it because the tolerance will be better than a single piece of 1/4.

4414 did that on their turret this year, notice the seam between the plates and the 20dp gear pattern.

13 Likes

Pretty sure 3339 used one (or something similar) this year for their limelight turret

3 Likes

We did use one this season, to allow the limelight to rotate continuously. Overall the slip ring was great with no major issues.

2 Likes