Direct drive or Chain?
 

Our team, tem 599 is debating whether or not to use chain drive or direct drive with four motors. We have heard that chain drive is more forgiving (in terms of strain on the motors) than direct on the motors, but direct drive may be faster. We had some principal concerns with both:
For chain drive:
1. Clearance
2. Are chains really worth all the trouble?
For direct drive:
1. Can we make sure that the motors run at the same speed?
2. If not, how close can we make it?
3. Does it really matter?

with 80/20 frame, chains is so much easier than gears its not even funny. We use 80/20, gonna use chain drive exclusivly (cept for the gearbox on the bosch, that gives us a good reduction we don't need to pay extra for)

edit: Its not hard to get different motors to be withing 10 or less RPM's of each other, we got like 5 or 6, which is about 1 or 2% difference, which will be fine over a period of 2 minutes. Over ten hours, you would probably burn the motors up, but 2 minutes is fine.

The most important thing to consider about using gear is your skills in making a gear box. Usually, when you are using gears meshing with each other, you can only have a 0.001 error, maybe even less than that in order to make the gear box work.

With only drill press, or even a mill without digital read out, I fear you may not have the necessary skills and tools to fabricate a good enough gear box. That's why a lot of teams use chains, so they can tention the chain without making the whole drive system as accurate as a gear box.

If you have the resource and knowledge to make gears work, then by all means do it. Else, chain will work fine too.

As for the speed of motors... They are really about setting up a correct ratio, and both chain and gears will work for that.

Hmmm. Guess me and my drill press with an X-Y table are going to have to go back and redesign our gearbox from scratch. :rolleyes:

Just as a point of reference, in the hands of a skilled machinist (almost an artisan), amazing things can happen. A century ago, toolmakers regularly made precision pieces to within 0.1 mil on a regular basis, by hand. By using the right techniques, it is possible to hand grind a surface plate to one MILLIONTH of an inch accuracy. For 300 years, watchmakers have cut and filed gears by hand, with great precision.

Don't underestimate the ability of a highly skilled technician with simple tools (and don't overestimate the ability of an unskilled fool with exotic tools!).

-dave

we r using a combination of gears and chains

We use gears in our gearbox, but we run chain off of the output. It gives us some flexibility in adjusting speed and power, because it's a lot easier to swap out a sprocket than it is to switch out a gear.

>>Don't underestimate the ability of a highly skilled technician with simple tools (and don't overestimate the ability of an unskilled fool with exotic tools!).

-Amen Brother!

Just remember, if you use chain, make a tensioning devices. Switching rapidly between going forwards and going backwards creates a bit of slack (don't know the technical term for it), and if you're not paying attention, this slack is going to make the chain fall off.

From expierience, I know that had we made a tension device that would have stopped our chain from falling off right at the start of a key match during nats, we would have done a lot better. So don't disapoint yourselves later on - make a tensioner.

In 2001, we (258) used pieces of delrin as tentioning devices on our drivetrain.

By cutting roughly (Thursday morning, at the Los Angeles Regional), and drilling holes wide enough to feed chain through, with another set of holes to suspend the blocks with pullties. Two per wheel, two wheels per side, eight blocks total.

Yes, it was a very makeshift system, but it worked like a dream.

I wish I had a good picture of the system, right now, but you can almost see it at the bottom of this picture. The white cubes are delrin, the white strands are pullties, and the black strands are the tentioned chain.

There's an even easier way to tesion chain. Take a piece of 1/2" dia delrin rod about 1.5" long or enough longer to bridge whatever gap your chain is runing through. Drill a hole through the center of the rod perpindicular to the axis of the rod about 3/8" from each end. The two holes should be parallel to each other. Slide the rod under the chain but up against some supporting structure until it is just a little too tight. Bolt the rod down. The chain will wear into the rod until it is the right tension. It will then stop wearing into the rod as a sort of equilibrium is reached.

It takes 10 or 15 minutes to fabricate and install.

Chain Drive is what will be used by T522

chain v. direct
 

We're going with chain giving 4X4 drive, but we've shied away from casters added to direct drive systems from some unsavory problems our rookie year. if anyone was at the Columbia regional two years ago, they remember a big gold triangle that couldn't go up the teeter totter, and was beaten by a robot that couldn't move. the reason, aside from bad gearing, was the casters caused the robot to spin wildly, as it changed from turning to going strait. the casters would spin into position moving the robot in the process. also in the same fashion as a shopping cart, if one wheel came off the ground and spun, then hits the ground again, the robot will swerve in that direction, and if a caster gets stuck, the robot will spin in large circles!
My advice: stay AWAY from casters! they are evil!

237 uses a sproket chain setup (has for many years now) for our systems. Using 35 chain and sprockets does wonders, we've found it to be a great reliable system. It easyer to change a blown chain than to have to pull a gear bow and do damage cointrol. if the system is over stressed, the chain is usually the first to go rather than the gears being chewed. its more of a user friendly/pit friendly setup, most teams have chain, but it might be hard to find a team with a spare (insert gear specs here)..

Your just asking to burn out your motors with direct drive.. go with the chain.Trust me

Problems with chain
 

We last year used primarly chain drive and experienced a lot of problems with it, for one there is a little slack in all chains so this leads to increased strain on the gearbox, which last year caused us lots of problems. This year we are using belt material that has worked wonders for us...

We are using a combination for direct drive(well direct after the transmission) and timing belt, but for a team without the equipment to do precision machining then you should go with chain. A tiny bit of error with a direct drive system can put you in a world of trouble. Chains are imo the better choice for your situation.

chains also need some precision in their construction. As in all your holes should be centered. We had a lot of problems with sprockets on bending shafts and poor holes in the sprockets. Slowly got worse and worse. So bad that this year we designed an extremely cool direct drive system before the competition got started. then we decided to make a vector drive system sooo we are back to chains. Even our arm uses chain this year. Long strands of 1/4 chain going up a 9 foot arm. looks sweet and is lighter than a solid rod would have been.

Chain is easy
 

I like chain drive. Keep the chain short, try to align it properly, watch the tension and use #35 if you can afford the weight. Chain is strong and forgiving of slight misalignment. We only have access to a wood shop and a couple of drill presses, but have managed to fabricate a pretty good drivetrain.

Our current 'bot has a 17 1/4 wheelbase with 12-tooth sprockets connecting the front and rear wheels (appx 3 feet of chain). Last year's 2-wheel-drive had a really short chain from an 18 tooth sprocket to a 60-tooth sprocket on the drive wheel - perhaps 18" total. That one never needed tensioning, but I suspect this year's will need just a bit, despite the chain being really tight to begin with.

Ken's point,I think, is that unless you have the skill to hold tight tolerences, don't bother trying to build a gear box. You don't need a readout to hold a .0005 tolerance, they just make the job easier. A properly trained machinist will edge find the part, zero it on the readout, move to the XY coordinate, spot the location, remove the part from the mill, setup a angle plate on a granite surface plate, place a gage ball in the spot, measure to the top of the diameter of the ball using a .0001 indicator and Gage Blocks to determine if the location is on center, and if not, measure how much to move to get on center. There is more to tight tolerance work than just haveing the mill. You also have to know how to measure your locations.... Sorry I went on. Thjere is just a lot to building something to that tight a tolerance...

Our team just this year got a mill...talk about an improvement in part production! We built our own gear box with very simple tecniques using only the mill and it's manual gauges...anything is possible with patience, skill, and a little luck!

Mehhh how hard can it be. All you have to be is tediuos. If the wright brothers assistant could build an airplane engine using a natural gas powered lathe and drill, and hand tools to thousandths of an inch how hard could it be to build it using the tools we have today.

Tried and true method
 

I don't get what all the fuss is about these gearboxes. We've built 'em for two years without a mill, or precision machinery.

We have a tried and true method that some may grimace about. First, we order medium-to-large pitch gears that we know we need. Then, once we have them, we take a plate and mount our motor to it, attaching the driving gear to the motor. We locate the second gear, clamp the gear down, remove the motor, and drill straight through the gear into the plate. The hard part is making sure the large gear doesn't move.

We've used this method for two years and it's worked flawlessly. Some more experienced teams may frown on it, but it's not like we're building a robot meant to last for years and years.

Re: Chain is easy
 

I personally prefer a small, tight gearbox, with a single stage of chain and some kind of tensioning adjustment. This allows you to change the final ratio easily if it turns out you've chosen badly for the dominant strategy that evolves per regional, yet have a strong and reliable main gearbox.

If you use chain, IMO it is VITAL to have either:
A) A simple tensioning adjustment (i.e. using an extrusion system and sliding the gearbox slightly along the rail),
B) Having the motor end of the pair spring tensioned, or
C) Having a PAIR of spring tensioners in the middle to eliminate chain slack in BOTH drive directions.

In addition, if you have a long chain run, consider adding idler wheels and/or delron/HDPE grooved CHAIN GUIDES.

The higher the engagement angle, the lower the possibility of chain hop offs. Shoot for MORE than 180 degrees on all end sprockets, if possible, unless the chain is VERY short.

Well, #35 is forgiving, but not #25!

I WISH we've used #35 this year. I assumed we would, but to my surprise our mech team chose #25 for weight savings. Trust me... It was a NIGHTMARE. Way too stiff, too weak, and IMO the alignment was too critical for drivetrain use.

We broke the chain several times in rounds. It also came off more than once, leaving us spinning like a one armed man in a rowboat... The engagement distance is just too small. Therefore, it comes off way too easily from the slightest alignment or tensioning error.

Talking with other teams, #35 chain is a LOT more forgiving of alignment and tension errors, PLENTY strong enough for this work, AND enough other teams use it so you can find parts like master links at a contest in an emergency. My last team used it in the 2001 CDI contest, and it was GREAT! Not a single problem with it.

If we ever use chain again in a FIRST contest (which we probably will), I'll DEFINITELY argue for using #35!

- Keith

chain
 

Actually, I've only used #35. Considering all the info, I'll stick with it. Our 4-wheel-drive bot used 2 pounds of it this year. Plus, you can get #35 links at places like Home Depot and Loew's.

We've had a fair amount of stretch or break-in wear with 35, and I would imaging 25 is just that much worse.

Re: chain
 

OOC, One item that was recommended by one team at GLR was using "standard bicycle chain".

I'm not so sure about that... Does anyone have any experience with this? How does that compare to the "numbered" roller chains size and strength wise? Is it strong enough, and RELIABLE enough for our use? Heck, you can get THOSE sprockets, chains, and links at numerous stores in virtually any mid-sized town! (Personally, I'm still leaning toward #35 for next year...)

- Keith

--
"How did your team break your chain?" "Why, with a 'chain saw', of course..."

Re: Re: chain
 

I don't know how bike chain is sized but it looks to me like #35. I couldn't imagine why it would be on a different standard. Anyway, yes bike sprockets and chain could be a very cheap effective way to get the job done. It just depends on what you are trying to do. We use all hubbed keyway bore sprockets that are not found on bikes.

#25 Chain is plenty strong for drivetrain usage ~900lbs break strength, we used in on this years bot and trust me, we are putting out some serious power in this years drivetrain (CIM&Drills Already have ripped 4 Brecoflex belts in half after two reigonals) As for alignment on the #25 it is much more difficult and critical to keep the sprockets aligned. At Buckeye we were throwing the chains on high speed impacts so we made some delrin circular plates bolted to the sides of the sprocket that make a v for the chain to rest in, worked great never broke or derailed at GLR and we were in the winning alliance. :)

Would I do it again? No Way. All gear drive next year and you never have to worry about it again plus you get even better efficiency.

I'm All for Shimano!!
 

I worked in a bicycle shop for 10 years. There are 2 types of bicycle chain-Standard Coaster Brake Chain and Derailluer Chain. They are both 1/2 inch pitch. The Standard Chain is 1/8 inch wide and uses a master link. The Derailluer Chain is 3/32 inch wide and uses no master link. You need a Chain tool to push the pins out and back in to connect the chain. Bicycle chain is very different in pitch and width and is not compatible with #35 pitch gears.

The past 3 years we have been using #25 chain in our 6 wheel drive set-up. This year's 'Bot has 2 chains per side, 1 short and 1 long. I modified my Park Chain Tool to work with the #25 Chain pins, so Iwe are not running any master links in the drive chains.(This dramatically increases the chains working load) I posted a few photos in the Gallery of how the output shaft is coupled together. The trans can be removed in 1 minute from the chasis without disturbing the chain Adjustments. Just after the last match with 302/67/226 in Ypsi, I pulled the trans' apart, inspected, and reinstalled in the chasis in less than 15 minutes. We have not had any trouble with the chains breaking or jumping off. We played very hard all weekend, and had very little trouble with our drivetrain. The only damage we sustained was qualifying match #102 against 302. They drove us into the wall, which we then drove over them, continued about our way to the rest of the match, until the end when we drove on top of them to claim our territory on the ramp. We did bend a 60 tooth sprocket, which was easily swapped out.

I have considered running Bicycle chain for next year because so many ratios are available in very lightweight sprockets. The tensile strength should be about the same as #35 chain. In fact, I don't think 130lb robot can exert as much force on a chain as a 220lb rider cranking up hill would. :cool:

I have noticed that gearboxes can be a little... touchy. This year, we went with the FIRST supplied gear boxes, with slight modifications. Never again. Something was seriously wrong with the right hand side of our drivetrain, and we ended up dissassembling it 6 times. Just before our last match, we blew the side out of the plastic gearbox. We believe that it was a combination of overtightened bolts, slightly oversized bushings, and excessive force from the motor. Unfortunately, our expierence with chain drives hasn't been much better, as we had a fair amount of trouble with the chain jumping last year that was never really resolved. One of the major lessons was that the whole gearbox needs to be aligned correctly, and everything test fitted and measured several times. I'm sure we would have caught the problem had we had more time to practice, but I believe it is finally resolved. I will definitely be making a tour of the pits this year looking at the drivetrains!

S cubed

Yeah I'm sure they built the plane in six weeks too.

You saw my first post extolling the virtues of chain. Yes. That was before we actually built it. I don't like chain for these reasons now:
Heavy.
Dirty.
Prone to falls off.
inefficient
space consuming.
liable to jam.

First of all, in our drive system there were these stuffs:(i kan spel.)
2x 60 tooth sprockets
8x 10 tooth sprockets
2x 40 tooth sprockets
2x 45 tooth sprockets
2x 22 tooth clutch sprockets (big thingies that have built in friction clutch)
20 steel bearning blocks

all were 5/8th inch bore.
all was #35 chain.

so we had ungodly amounts of 5/8th inch shaft (heavy) lotsa large carbon steel sprockets (heavy even with holes in em) lotsa heavy steel chain, lotsa heavy steel bearing blocks.


and it was so inefficient and liable to jam . . . . . well, we fixed the jam part by locktite-ing everything where it was supposed to be, but still. Next year we are getting access to a good digital or CNC mill and making us a gearbox. actually we have several ideas for next year, but they all assume we will need a drivetrain at all. Which we probably will, but we will wait and see.

ooo, yes, I forgot space consuming. our bot was a box of 80/20, most of the innards were taken by our 4 motor drive system, while 45's drive system had 6 motors, but they kept it on one far end of the bot . . . hardly took any room at all. There is so much we learned this year. Its not even funny. And if thats not funny, there are lots of things that aren't even funnier . . .. nooo, no Catch22 here today.

We have never had a chain fall off or jam. For the ineficiency, I'm not sure. Gears can be just as ineficient if not meshed and aligned properly. As for space and weight, I can really debate that. But we like chains and sprockets just because of the simplicity and robustness. It is so easy to set up a chain drive with simple tools and make it very effective and robust too.

Our robot:
2 Drills
2 Chias
-- All sprockets carbon steel #35--
4 10T
4 32T
2 35T
2 48T

42 inches in total of 5/8 keyway shaft

Very heavy but if you saw our robot in action at Phoenix you wouldn't complain about the weight.

Being tedious is my point. Gage blocks and indicators were created in the mid 1800's. So precise measurement has been around a while. Laying out a drive setup with a height gage and drilling holes on a drill press will work, but you still need to precisely measure the hole locations if you want a smooth, frictionless system. I'm sure the Wright Brothers used Gage Blocks and indicators to build a gas powered motor. You can't get a precise piston to wall measurement from a pair of dividers and expect that the piston will not sieze or rattle itself out of the bore. Or correct oil clearance on the bearings without a bore gage.

You can only hold as tight a tolerence as the machinery that you have to work with. The better the tools, the better you can repeatedly hold a tight tolerence. Modern machinery dramatically reduces the time required to Accurately machine a part, and beinng we are only given 6 weeks to design and build a robot....all the time you save...:cool: