Anyone ever try a screw drive? Was it worth it or nah? Thanks!
Hmmm. I haven’t been around here long enough to know that nomenclature, and Google isn’t helping. Got a picture or description of what you mean by “Screw Drive”?
Yes.
Editted to remove ginormous picture.
You can see the image. Basically the threaded rod spins which moves the carriage. Some 3D printers also use this.
Oh gotcha, just a worm-gear centered mechanical power transmission?
We did a lead screw setup for our elevator in 2015. Great because it was hard to back-drive so we didn’t need to worry about our motors burning up holding the tote in place. Horrible because it was hard to backdrive, which made positioning by hand pretty sucky unless we greased it up just right.
I think it has its place in FRC, but the big question to answer: for your application, do you care about being able to backdrive the mechanism easily?
But if you use the worm gear with a threaded rod, shouldn’t it be able to back drive?
Like this
I want that. But yeah like that.
You can see it here too in this 3D printer
Ah I see
To keep friction low, the edges will have to be pretty sharp. I think it would cut up the carpets. You’d be risking being banned all the time.
\
I was thinking in term for elevators or lifts.
Ahh, you mean a ball screw or a lead screw.
Many teams have used ball screws for motion. Many choose to use a standard threaded rod with bronze nuts to cut cost. Actual ball screws can be quite pricey. Examples used to come with the kit of parts - I’m not sure if they still do anymore.
Andymark sells larger versions of them them (FRC Robot sized):
You can see why they get prohibitively expensive when you need them for long lengths.
What you’re describing is a ball screw or lead screw. Teams certainly have used them in their robots, but there are often better ways to get the motion you want. (Of course, that depends on your design priorities; in some cases they are the best way)
Ball screws are often prohibitively expensive for the common lengths of FRC elevators. They are almost always back-driveable, and have up to 90% efficiencies.
Lead screws are not suggested because they are highly inefficient, usually 20-25%. That means that your motor needs to be 4x more powerful than otherwise, and it draws 4x more power from the battery. They are reasonably cheap though, and are usually not back-driveable.
Instead of these, you can use a “standard” chain-driven elevator. This style is likely more efficient, lighter, and cheaper than either option.
708 frequently use lead screws and/or worm gears in their mechanisms. Their 2015 robot was named “Banshee” for the sound their screw-driven elevator made when it moved. Their 2016 and 2018 robots both used screw-driven telescoping arms to climb.
Team 3512 used the AndyMark Worm Box in 2016 and 2017 as our climbing gearbox.
Overall, we were pretty satisfied with it’s performance. A neat thing about it is that because it’s made to be driven by a CIM, you can instead use a VersaPlanetary w/ CIM output which gives you a myriad of different ratio and motor options. Because of tight tolerancing, the Worm Box can be a struggle to put together sometimes and in general definitely has a learning curve before you get good at quick assembly and disassembly. It’s a fairly large gearbox, and especially with a VP mounted to it, it’s a bit of an awkward shape. It doesn’t back drive when unpowered, which can both be a blessing and a curse depending on the application.
Overall it’s a solid product, however now that VEXpro has released a ratchet for it’s VersaPlanetary ecosystem combined with the power of the 775Pro motors, my team probably won’t be using the Worm Box w/CIM in the future for any applications were we need to prevent back driving in a subsystem. The VP ratchet + 775pro combo is lighter, cheaper, and smaller.
Hope this helps!
Sean is correct, we have used lead screws and worm drive gearboxes 3 times in the last 4 seasons with varying degrees of success.
Ball screws would be the best solution but as Ari said they can get very expensive and you will lose a ton of efficiency if you get the wrong type of acme screw. We purchased our lead screws in 2015, 2016, and 2018 from McCaster Carr but made the mistake of purchasing an “acme lead screw” in 2015. But in 2016 we found the “precision acme lead screw” and that made all the difference.
The standard 1/2-8 or 1/2-10 acme screw is single thread so you will get .125" or .1" of travel per turn respectively. The precision acme screws are multi-thread, in our case 8 and travel 1" per turn. The single thread screw in 2015 was more or less direct drive off a mini CIM to “get the right speed” but ended up creating really bad vibrations and we replaced it with a chain driven system for MAR Champs.
Multi thread screws can back drive depending on your gear reduction. This year we have a shifter on our lead screw drive with roughly 10:1 and 30:1 reductions and it will back drive when its in 10:1.
I highly recommend this screw and nut if you want to try it. We would 100% use this again on one of our designs. It would be really interesting do do an elevator like 1640 with a screw like this.
We have used worm drive gearboxes in 2016, 2017, and 2018. I would say our 2016 implementation was the most successful. We were recommended a Boston worm gear set from a mentor in the area in 2016 off amazon, unfortunately I don’t have the part number handy. We used a similarly designed gearbox in 2018 but I think we made a miscalculation in our loading somewhere along the line and ended up breaking a lot of gears this season. In 2017 we tried direct driving the worm on a CIM for our climbing mechanism but over time we had issues with the gearbox and ended up replacing it with a 775pro on a 90:1 versa planetary with a ratchet wrench. Its likely that we were side loading the CIM shaft and damaging the bushings but we didn’t bother looking too deep into it.
Our 2016 and 2017 robots posted on our website and we are planning to get 2018 posted some time before 2019.
And we are also competing at Ramp Riot in 2 weeks, feel free to come by and take a look at the robot up close.
Where did you use the lead screw in 2016 and 2018?
EDIT: Do you also mind explaining how you got the motor to drive the screw?
Our team used a screw in 2016 for the climbing mechanism. From what I can remember it was effective. The only problems if I remember correctly were a lack of speed and backdriving!
Ballscrews and leadscrews have their place in FRC, but if you’re new to designing I strongly advise you avoid them. It’s very easy to load up a single-lead leadscrew with thirty or forty pounds and then wonder why it’s killing your 775 motors and moving really slowly. Ballscrews are expensive if you don’t know where to look, and can have long lead times.
A more robust solution is to use belts or roller chains for your motion. Much lower friction, and thus faster, although it can be somewhat more expensive depending on how you implement the leadscrew. The distadvantage is that it’s backdriveable, but it’s not difficult to just add a small amount of power all the time to keep it in the same spot and offset gravity.