I’ve seen some robots where the shooters are able to rotate independently. How is this done? Like is it done using custom machined parts or is there some COTS parts that can be used or combined for the same effect? Thanks for any answers!
Turrets can be tricky. My team has never made one because we didn’t want to take on the challenge. But many do.
Here is one turret from Team 3880: https://www.chiefdelphi.com/media/photos/45589
You can see it is a custom chain driven shooter assembly.
Here is a video from Triple Helix showing a similar design: https://www.youtube.com/watch?v=9AV4Tij1c5I
1 Like
They can be made with custom parts, but many are made by actuating a Lazy Susan. They can be difficult to build right, and you need to make sure your mechanisms weight is centered over them or they can (and will) pull apart. A few years ago I saw a rookie team build an awesome, and tall, mechanism on top of a lazy susan. Halfway through the event, it pulled apart and dropped a hundred or so ball bearings right into the middle of their robot.
In addition to the increased complexity of a lazy susan compared to a stationary mechanism, they can be difficult to aim properly. If you are good with vision tracking, aiming becomes a lot simpler. Otherwise, you drive to where you want to shoot, and then need to rotate the shooter to aim properly. Doing so may help if someone is playing defense against you, preventing your robot from turning towards the target, but otherwise you might as well just turn the robot properly.
If you go this route, I would highly recommend planning on fixed angles. For example, with Stronghold it may have been advantageous to have a shooter that could shoot forward, left 90 degrees, and right 90 degrees. That way you can line up facing the tower, or just run into the driver station wall and shoot sideways into it.
1 Like
One of the best places I’ve found to answer questions like this is Spectrum 3847’s amazing robot mechanism gallery: https://photos.spectrum3847.org/Robot-Mechanisms/2017/Turret/
1 Like
We did a turret our rookie year using a lazy susan, bringing the ball up through the hole in the bearing. We found out that lazy susan bearings are not designed to take the lateral forces of launching.
We haven’t done a turret on a competition robot since, as we’ve never felt the cost of a real turret bearing (either in $$ for COTS or time for custom) was worth the return. Our air cannon has alternated over the years between turreted and non-turreted design, but none have passed the ammunition through the bearing. One turreted iteration was a TB-mini oriented with the output shaft pointing up. The one we’re building now is being implemented with galvanized plumbing and a bronze thrust bearing (washer).
Edit:
All of our turrets have had a limited range of travel (typically 270 degrees) so we didn’t need slip rings. Assuming you go this path, routing wires (and hoses) near the center of rotation will allow the greatest amount of rotation for a given amount of slack in the umbilical.
My team built a very simular turret to this design in 2017. I do not have any picutures to show but we modeled it heavily off of 125’s turret design that year. Our turret didn’t use any ball bearings and just spun on a sheet of poly-carb. if memory serves me correctly we didn’t have so much of a problem with the turret hood, but the management of wires. Depending on where you run your wires you may need a slip ring or the wires will just bind up and disconnect or snap.
There’s an alternative to using lazy susan bearings, which 148 and 254 employed in 2017 and 2016, respectively. The approach is to use a stacked set of bearings spaced along a circle, with one bearing being a slightly smaller diameter. The larger bearings take the thrust loads, while the smaller bearings take the radial loads. I’m pretty terrible at explaining, but here’s a picture snapped from 148’s 2017 CAD. Hopefully this helps clarify some things.
It seems to me that this implementation is potentially more flexible, assuming you have the machining capabilities to pull it off. You would drive this mechanism the same way as you would with any other turret, that being a custom machined gear or sprocket driven by a 550 or BAG motor.
Be careful with slip rings. It’s hard to find ones appropriately rated, and you need to make sure they don’t contain mercury (which is common!) per the blue box under R08. It’s better (cheaper, quicker, and even easier) to limit the range of motion and build in wire management to work with it.
These are a fairly common bearing to use on turrets:
Wow, that price point would have opened some design doors in 2016, and for our air cannon.
1 Like
Most of the discussion in this thread has been about the mechanical aspects. Jon brings up the control/aiming of the turret that makes the turret useful. The OP needs to research this aspect of the turret too. I took some students through the pits at Houston Champs in 2016 to study a curated list of some of the best performing robot designs. A few of these had turrets, all of which relied on sophisticated control/aiming systems. For every effective turret system, we saw many more (5-6 X) fuel shooters that did not use a turret.
Look into plastics and oil -embedded material. Depending on the weight of of the mechanism, materials can ride on each other. A way to machine flat parts always helps but you can get stuff done with good templates and a bandsaw. The biggest thing is doing one before season.
-Ronnie
For stronghold, 1501 had a great turret shooter on their robot. We got to see it in week 1 here in NC and then again at Worlds. They had made some changes but the overall system was largely unchanged. They had a really good control aiming system and could shoot from just about anywhere on the field. Thy also had good feedback to the drivers with a series of LED lights that would let the drivers know when they had a boulder, when they had acquitted the target, etc.
https://www.chiefdelphi.com/media/photos/43187
We have talked about adapting our swerve module integrated bearing design to a turret shooter - basically upsizing the ball bearing ring to be big enough for the game piece to fit through. Those conversations always end the same way - we have swerve drive, so the whole robot is the turret. Consequently, we have never done it. Still, it would be fun to try.
We just used 25-chain on the edge of 1/8 aluminium sheet because we don’t have access to a water-jet to cut teeth in it. In this picture the chain was not on the pulleys probably because software was working on it still at that time.
Couple videos:
Bottom LED was boulder loaded, Middle LED was shooter up to speed, Top LED was turret locked. All three had to be lit up before controls would allow the driver to shoot.
Video to show results of filtering on field.
One of the Einstein videos
For wgorgen:
Pictures from NC! We had so much fun! You brought back some good memories.
I remember that chain drive on the turret. So cool. That design was really great. I remember when I figured out (about halfway through quals) what the LED light strips were signaling. Very smart.
It was great to watch you guys on Einstein that year.
Ha! I remember you guys using our signs to block the sunlight to help your vision system. Good times. Hope you guys come back down to NC one of these years.
I’m seeing some in here about not using a lazy susan and custom machining some parts - I’m curious to learn a bit more about that, I assume you would have to machine it as two parts, the chain driven bit and then a bottom plate that’s kept in place by some kind of rollers or something? That’s what I’m kind of taking from the picture from 148’s CAD. I haven’t had a chance to download and open it up yet, but I’m trying to make sure I have the right idea.
Turrets were big for the Aim High Game (2006). A lot of good info can be found in the book published that year.
Take a closer look at the 148 CAD. The “top plate” is the part that rotates. The “bottom plate” is fixed to the frame of the robot and does not rotate. The bottom plate has a larger circular hole in it that forms the race that the turret rotates on. There are a series of bearing stacks that are attached to the top plate. These bearing stacks form rollers that ride on the ID of the large circular hole (race) in the bottom plate. These rollers are formed by one large bearing on top, one small bearing in the middle and one large bearing on the bottom. The middle bearing is what rides on the inner edge of the race resulting in a radial constraint. The upper and lower bearings in the stack form “end plates” to this roller to form vertical constraints. This bearing stack is held in place with a bolt passing through the center of the bearings which is also used to attach it to the top plate.
By placing a series of these rollers in a circle attached to the upper plate, you allow the turret to rotate. The diameter of the “bolt circle” for the roller bolts would need to be smaller than the diameter of the hole in the bottom plate by the diameter of the small bearing (and probably a tiny bit more just to add enough freeplay to limit any binding). You would need a minimum of 3 of these rollers to form a stable turret, but you would probably want more than that.
Ultimately, those are all the features you need to have a functioning turret. There is a decent level of precision that is needed to make these parts, but if the turret was small enough, it could be made on a fairly simple lathe. It should be relatively easy to fabricate the plates for a larger turret with either a waterjet, a CNC router, or even a simple end mill with a turntable fixture, but you should not need anything more advanced than that.
The details of whether you have chain sprocket teeth or gear mesh teeth in the upper plate are to facilitate the rotation is really a matter of personal preference and capability. Alternately, if you had a simple circular OD to the top plate, you could have a single grippy wheel to rotate it. What you add to the top plate and whether you have a large hole in the top plate to pass a field element through are also matters of personal design preferences.
Hope that helps.
1 Like
Did you mean your chain driven piece and the bottom plate as being both on the turret? Most turrets have the bottom plate fixed to the chassis, and the top part rotates. For simplicity, the motor which spins the turret is normally on the fixed side (why run those wires through your wire management system?) and it engages the turret through belt, chain, or friction (e.g. a rubber wheel spinning a metal drum).
Bottom line - the parts on opposite sides of the rotation would be machined separately. Parts on the same side could potentially be machined together, but it may be more convenient to machine them separately and fasten them to each other.
I know the bearing stack was used by Team 27 - I have a fuzzy remembrance of seeing it in the 2006 behind the design book.