Worm gears
 

I was thinking about worm gears today, for a winch application for the robot, when I remembered that there are some commonly available plastic gears, that are used in garage door openers. They work with a 1/2 HP motor, so they're reasonably strong. If you search for "chamberlain gear kit" you should be able to find them for sale from several vendors for $20-30 for the illustrated set which includes some spacers, thrust washers, and lock clips.

Re: Worm gears
 

wow, nice find! Even comes with Grease!

That'll be used for a prototype elbow in the off season.

Re: Worm gears
 

I have a friend who installs garage doors, thus he fixes and replaces openers too. He donated a few used gears to the team, so I got to play with them. The worm gear has two teeth, spur gear has 32, for a ratio of 16:1. Apparently it's not a steep enough ratio to prevent backdriving (he said a garage door with balance spring problems will fall closed when connected to one of the openers that uses these gears).

Still, it's a nice light weight 16:1 gearset that fits 1/2" shafts, and looks relatively easy to make a box to hold the parts.

Re: Worm gears
 

How does that work? Aren't worm gears only 1 tooth gears? Or does this one have two separate threads?

Re: Worm gears
 

There are two separate threads.

If you study the linear actuator thing in the KOP you'll see a similar screw with more than one thread.

Very strange...but that's how it is.

Re: Worm gears
 

Thanks! That makes sense.

Yeah, that lead screw in the kit is just weird! I've always wondered why it has more than one thread. This year's has six, and last year's only had three (I counted :p).

Re: Worm gears
 

and the year before that it was 4 threads. Odd that they change it... any Ideas of were someone could obtain a cheap gear system like this that doesn't backdrive?

Re: Worm gears
 

McMaster Carr sells a hand powered worm drive winch, perhaps you could find what company makes it and see if you can get spare parts for it? just the gears?

Some circular saws us a worm drive, for example the Skill model 77. might be some adaptable parts in that?

I've been trying to think of other things....

Re: Worm gears
 

y not use a paw and ratchet? to prevent back driving?

http://www.mcmaster.com/#gear-ratchets/=5g9esx

At the bottom.

Re: Worm gears
 

That's what we'll probably end up doing. i doubt those stainless ones would be easy to adapt, but we might find one that is pretty easy to adapt....maybe even a part from something as simple as a ratcheting tie down strap would work? Or a socket wrench ratchet?

Re: Worm gears
 

thats what we have pretty much settled on. lol. Ur going to here a loud clanking noise we we go up but only for about 2 seconds ;) if that lol.

Re: Worm gears
 

If there were an FRC award for posting the most useful ideas, images and videos on Chief Delphi over the past three years, Jim, your team would have to be at the top of my list of nominees.

This is another great piece of info.

Jason

P.S. Those gears must be tough... one of my garage doors had been a bit sticky lately when it was coming down and when I checked it out, it turned out that the spring on one side had come disconnected when a fastener failed. It must have made a bang when it went, but the opener just kept chuffing right along. A quick fastener replacement (er... "UPGRADE PART" to use FRC speak) and some retensioning and it was as good as new... never missed a beat.

Re: Worm gears
 

Thanks, Jason! We try to help where we can....

I've also learned that posting our crazy ideas here is a great way to find out quickly if we're headed in the wrong direction.

Re: Worm gears
 

ya, I saw that there was a worm gear in a circ saw, ill see what i can do about finding a broken one and post its ratios and stuff.

Re: Worm gears
 

Nice find. I had more or less ruled out a worm gear since they generally are extremely low efficiency (relative to a spur gear and ratchet) in order to prevent backdriving, but using a higher efficiency worm drive with a ratchet might be an option. Any idea of the efficiency of this set?

Re: Worm gears
 

We used that exact gear set for our winch in 2004 (Frenzy)

We were very pleased with the results. Lasted 2 regionals with no significant wear.

If it is possible to back-drive it, one robot is not nearly enough force to do it in our experience. If you were lifting 2 or 3 I suppose you'd have to try it to find out.

Re: Worm gears
 

Worm gears are fraught with peril for FIRST teams.

Well, actually, they are fraught with peril for any application, but some times you need to face your demons!

In one off applications like FIRST robots, you really need to know what you are doing to use a worm gear (other than the one in the window motors of course).

Getting the center distances right is a harder task, managing the thrust loads, etc. etc. And don't get me STARTED on efficiency (you flush half to 2 thirds of your power right down the drain -- and that is if you do things right you can waste 90% easily without even trying! But... that is is why they are not backdrivable - so that can be a plus but...)

My advice: find another way if you can.

As for me and my house... ...I am thinking this is a job for NBD -- Nothing but Dewalts -- Worked for me and Chief Delphi in Raising the Bar, I don't see why it shouldn't work for Breakaway

Joe J.

Re: Worm gears
 

We used and box end ratchet wrench in 2007 to make our lift winch non-backdrivable.

All it required was a hex of the appropriate size welded to the winch shaft, and then prodding the pawl with an allen wrench to let it back down.

worked like a charm

Re: Worm gears
 

It appears the door opener gears probably have better efficiency than most worm gears, because they don't have a very steep ratio, and apparently are backdriveable. The sliding tooth contact is not as efficient as the mostly direct contact of spur gears, so the efficiency must be less, thanks for pointing that out.

We had some fun with DeWalts our rookie year, and are a bit leary of them. I suppose if you are good at machining an 8mm hole in a pressed metal gear that has a 9mm minor diameter, then can get the gears pressed on the shafts (and have it survive a few regionals) without breaking they are OK, but that shaft to gear coupling problem makes me shy away from them for use with a CIM. A FP driving them should not be so bad.

Thanks for the comments!

Re: Worm gears
 

My experience is nearly universally positive. The shifters have had some problems (mostly due to programming issues that put them out of gear in autonomous) but other than that, I have been very happy with them.

Of course, I have always had access to either a good wire EDM source or a retired tool & die maker that made jewelry (figuratively) so getting things made were never an issue.

AndyMark (and to a lesser extent Banebots) has pretty much put me out of the NBD business, BUT... ...for an efficient high ratio (roughly 50:1) with a non backdriveable output, it is tough to beat it. imho.

Joe J.

Re: Worm gears
 

I don't know much of anything about this kind of thing, but a DeWalt was one of my initial thoughts this year for hanging. It's not too heavy either, and it's stood the test of 2004. Once I figure out the easiest way to get my hands on them...

Re: Worm gears
 

Very interesting....

http://www.andymark.com/product-p/am-0917.htm

Re: Worm gears
 

These were available to play with at IRI - very cool!

Just a note to those who may be wondering - these gearboxes are indeed backdriveable (I backdrove one by hand). Design accordingly :)

Re: Worm gears
 

Just another thought to consider with garage door gear strength is the fact that garage doors are counterbalanced with either coil or torsion springs, so the opener is not lifting as much weight as you might think. Backdriving can be eliminated with a ratchet/pawl operated by a servo to lock the transmission in position(s).:) :)

Re: Worm gears
 

Good find, Jim! And, thank you for starting this thread.

I must say that a couple of things inspired this design. One was this thread, and another was the failure of my home's garage door opener this past spring. After replacing the gears in my own garage door open twice, I figured that these gears may be good to try on a FIRST robot.

Also, I recall that there was an FRC team in St. Louis who used these same gears in their drive train this year. I forget who it was.

We did some testing on these gearboxes, and found that it takes about 85-90 ft-lbs of torque to strip the plastic teeth. We actually broke the large gear hub with a previous design at 60 ft-lbs, but then we beefed up the support around the hub so that the weak link in the design was indeed the gear teeth.

We have found a few things about this design that we like:

1. It is lightweight. The entire gearbox weighs 1.1 pounds.
2. The gearbox is quiet.
3. The CIM motors are out of the way, leaving the center of the robot with plenty of room for game object pickup (if needed)

Here is a video of a 4wd mecanum system using these gearboxes.

Sincerely,
Andy B.

Re: Worm gears
 

Andy - did you measure an efficiency?

Re: Worm gears
 

Neat!

It appears to me that the common failure mode in garage door openers is wear due to lack of lubrication. So make sure you provide plenty of grease, and instructions to use plenty of it, with the gearboxes.

Re: Worm gears
 

Andy - We like these a lot and plan on getting some in the next month or so. Are you planning on releasing a CAD file of this at some point?

Re: Worm gears
 

Guess who's making this into an arm gearbox? :)

With some earlier reduction (say, an RS-775 into a CIM-U-LATOR and AM Planetary), you can easily get your robot down to arm speed.

Re: Worm gears
 

Awesome.

103 used a worm gearbox this past year for the tower (though it was not designed by us). It worked really well and the only thing that had problems were the spur gears driving the shaft that the worm was on. :p

Re: Worm gears
 

Our team used a McMaster-Carr steel and cast iron worm gear set on last year's robot arm. We used one of AndyMark's nanobox's and a CIM to drive it. This drove a bicycle sprocket and chain to another larger sprocket mounted directly to the arm. Our worm gear set alone did not backdrive and with all the reduction we had very good control and speed, plus plenty of power. The system gave us no problems at all and we didn't even give a thought to efficiency since we had more than enough torque

We did notice some severe action when going from full up to full down, so we spring loaded the arm to reduce the "slam" factor. The arm was balanced by the springs when horizontal, so we actually drove the arm both up and down from horizontal. It worked quite well, and our rookie team was able to hang more than its share of tubes.

Though the weight of the steel.cast iron gears is greater than the plastic, I think I would go with them again, particularly on a drive system. I know the plastic gears work well in the garage door application, but I want to see some data on durability in our robot applications.

Dr. Bob

Re: Worm gears
 

294 used a 20:1 single threaded bronze gear/steel worm combination made by Boston Gear in both 2010 (for our kicker pullback) and 2011 (for our lift winch).

In 2011 we made the mistake of going with a cast iron worm gear (for cost reasons) on our practice robot; turns out we were operating with too high of a torque for cast iron and shredded the gear; cast iron is only rated at something like 25-50% of bronze, and our practice robot telescope wasn't as smoothly operating so it put more load on the gear. We never had an issue with it on our competition robot (as it had a bronze gear).

I'll post pictures/CAD of our 2011 gearbox if anyone is interested in more details. We used two banebots into a ~4:1 pinion/spur reduction, and a 20:1 worm reduction directly driving our 4" diameter drum. Good speed and absolutely no backdrive (we notched a slot in the end of the worm shaft so a power screwdriver could be used to manually raise/lower if necessary).

Re: Worm gears
 

reported ^