Best Drivetrain

[Rick TYler]

Originally Posted by Pat Roche

For flat game (no climbing) the simplest and most efficient drive train that you can build is two powered wheels that are center aligned with "skid wheels" such as hard plastic caster wheels that don't turn.

This was our platform in 2005 -- probably the most interesting year in 1294's history. The robot performed perfectly, all the time -- except when the team forgot to plug in PWM cables, or unplugged the compressor and didn't turn it back on, or when the drive team, which changed practically from match to match, didn't know what to do with the robot. It was a perfect storm of a simple, reliable robot that didn't LOOK reliable due to poor teamwork. A sobering lesson (and not my fault -- really).

[Lil' Lavery]

Originally Posted by CraigHickman

Yeah, I am offering to teach this ROOKIE team how to properly use this system. It's only as simple as using CAD to design your chain runs, and CAD is supplied free to all teams (in any flavor you want, too...). I agree that it's safer for you to avoid giving advice that may result in a bit of failure, but at the same time, if you don't start reaching for the sky early, it takes longer to get there.

CAD isn't going to make the average rookie team have an entirely straight frame, properly aligned sprockets and axles, proper chain tensioning, or the general machining precision necessary to have a fully reliable #25 chain drive. While everything may work fine in the computer, it doesn't always translate to the real robot.
#35 chain is plenty reliable and competitive for any team, not just rookies. While #25 has its advantages, I'd suggest rookies spend their resources and times developing other aspects of the drive and whole robot rather than worrying about #25 chain just yet. Spend more of those resources on better wheels, transmissions, electronics, or manipulators (or about anything else). Get done with your drive a little bit quicker with the supplied #35 and let your programmers have more time.

The best thing for a rookie to do is come up with a reliable, fully-functioning, usable drive with enough time for their programmers and other sub-systems to have access to the bot to do what work they need to do. That way they can spend more time programming, testing, de-bugging, and integrating the systems, as well as the all important training of the drivers.

[Daniel_LaFleur]

Originally Posted by rc_cola1323

I am posting this because a rookie team in Davis was having a hard time deciding and creating a drivetrain for their robot. There have been many different types of drivetrains used in FIRST. So in order to help all the rookie teams for nex year, what drivetrain has been the most dominant from the 2005 season and on?

-rc

ps Pictures please

As a rookie team, here's my suggestion.

Build a drivetrain that:
1> is within your machining capabilities. (If you don't have any machining capabilities, then the kitbot is your friend)
2> is within your mechanical capabilities (If, by the beginning of next build season, you do not know the mechanics of a swerve drive ... don't build one). Mechanical drivetrain failures are the downfall of many young (and some veteran) teams.
3> is within your budget (many drive systems require you to purchace parts ... make sure you've bugeted for them)
4> is within your programming capabilities (Holonomic and vector controlled drive systems require programming expertise ... If, by the beginning of the build season you do not have the programming expertise needed ... don't build one of these drive systems)
5> is easily and intuitively driveable by your drive team.


If you do these things, then whatever drive system you choose will be right for your team

[Brandon Holley]

Originally Posted by Tom Bottiglieri

While I agree with Brandon on most topics, this is one we don't see eye to eye on.

In my experience a fully functioning, simple robot is much more effective and reliable than a figurative time bomb on wheels. Remember that reliability is a part of your scope, as is function. If this doesn't click, think about what you look for in a car. There's a reason Honda and Toyota have risen to the top of the proverbial food chain.

The enemies of scope are time and cost. (FIRST has a third enemy: experience. But that is for a different discussion!)

Broken robots = stressful = no fun = less inspiration

No one said the robot had to be broken...just because it is more complex, or took more time to develop doesn't mean it is going to break.

I dont know tom, i guess we'll have to agree to disagree here...it seems to me that learning some more complex engineering substance is just as valuable as any other part of FIRST.

[Chris Marra]

Originally Posted by Brandon Holley

No one said the robot had to be broken...just because it is more complex, or took more time to develop doesn't mean it is going to break.

I dont know tom, i guess we'll have to agree to disagree here...it seems to me that learning some more complex engineering substance is just as valuable as any other part of FIRST.

I have to back Tom on this one, mostly because I think he just has another take on the same issue. Teams sometimes need to make a decision about where they aim to provide the most inspiration to kids - in the workshop through innovation, or at the competitions through empowerment when they see their creation. Sure a complex design is an amazing feat to accomplish and everyone can take pride in it, but how much fun is it if you don't finish it or you don't get a chance to utilize it because your focus was on making it happen at all. If you can finish your design early you get to show kids another very important aspect of engineering - testing and training. If you took any talented driver and told them they would be driving a Wildstang robot for the first time at a competition right after the coders finish testing, they would flip out. Not everyone can take a positive message out of a complex design if they don't pull it off in time or they don't get to use it to its fullest. I don't want to come off saying that winning is the inspirational experience, but sometimes competing needs to take a little more precedence over designing and learning how to tackle a complex project like any robot within the constraints of the competition is something teams ought to consider too. So there is nothing wrong with attempting a complex design, because there is no reason it shouldn't succeed, but teams ought not lose sight of their final goal: to have a competitive, fully operational robot completed within six weeks (without forgetting about their drivers or coders!)

[Jeff Waegelin]

Originally Posted by Tom Bottiglieri

While I agree with Brandon on most topics, this is one we don't see eye to eye on.

In my experience a fully functioning, simple robot is much more effective and reliable than a figurative time bomb on wheels. Remember that reliability is a part of your scope, as is function. If this doesn't click, think about what you look for in a car. There's a reason Honda and Toyota have risen to the top of the proverbial food chain.

The enemies of scope are time and cost. (FIRST has a third enemy: experience. But that is for a different discussion!)

Broken robots = stressful = no fun = less inspiration

This was probably one of the biggest lessons I took away from this year. Over the course of my FIRST career, my designs have ranged from the super-simple to the super-complex, and everywhere in between. This year was by far the best robot I've ever worked on, and it was also probably one of the simplest. Coincidence? I think not. Everything was simple and reliable, and it worked every time (barring a mishap with a wire and the terminal block). After 4+ years of constantly having to fix things at every event, I had two stress-free competitions, both of which were greatly successful.

The takeaway from all of this: especially for drivetrains, most teams will benefit from something simple and reliable. The less experience and assistance you have, the more critical this becomes! The last thing you want is to spend your whole competition season getting the robot to move. A 6WD chain drive may not always be the perfect drivetrain for every competition, but it's easy, simple, and reliable, and it will always be serviceable and successful for any game. (Until we get that water game, or Dave punishes us by banning all wheels.... )

An addendum on the #25 versus #35 question: I consider myself to be a pretty experienced designer, but I will ALWAYS use #35 for drivetrains. It doesn't matter how well you can CAD a drive base, if your fabrication tolerances are not tight enough, you WILL have problems with #25. I will take the reliability and forgiveness of #35 over the weight savings any day. I've seen too many #25 systems fail due to misalignment to do it any other way.

[Brandon Holley]

Originally Posted by Chris Marra

I have to back Tom on this one, mostly because I think he just has another take on the same issue. Teams sometimes need to make a decision about where they aim to provide the most inspiration to kids - in the workshop through innovation, or at the competitions through empowerment when they see their creation. Sure a complex design is an amazing feat to accomplish and everyone can take pride in it, but how much fun is it if you don't finish it or you don't get a chance to utilize it because your focus was on making it happen at all. If you can finish your design early you get to show kids another very important aspect of engineering - testing and training. If you took any talented driver and told them they would be driving a Wildstang robot for the first time at a competition right after the coders finish testing, they would flip out. Not everyone can take a positive message out of a complex design if they don't pull it off in time or they don't get to use it to its fullest. I don't want to come off saying that winning is the inspirational experience, but sometimes competing needs to take a little more precedence over designing and learning how to tackle a complex project like any robot within the constraints of the competition is something teams ought to consider too. So there is nothing wrong with attempting a complex design, because there is no reason it shouldn't succeed, but teams ought not lose sight of their final goal: to have a competitive, fully operational robot completed within six weeks (without forgetting about their drivers or coders!)

Chris/Tom

I agree with you guys, but I think your missing my point. I am not saying to go out and build some ridiculous design...but as a rookie team I feel it is good to set a precedent that you are going to do your best to make a robot that is to your ability every year. Some teams simply cannot machine an entire drivetrain, that is understood...but if a team has the means to make a part of their drivetrain, a little complicated, or innovative, why not?

I think it is important to balance the idea of "being competitive" with learning science and technology too.

I guess what I fear is a team saying, well yeah we could do that, but why don't we just buy this and call it done...

[Doug G]

I have to chime in here as well.

1) There's no rookie team in Davis, maybe they were referring to a rookie team that attended the Davis Regional.

2) Not to stiffle creativity and the thrill to innovate, but for a rookie team or 2nd year team - I highly suggest a robust 6WD setup with kit transmissions (if supplied).

Pic... http://www.travisusd.k12.ca.us/vande...n/DSC05175.JPG

Last year was probably our most reliable drive system (yes - even with banebots). Kit transmissions, chained to a center-traction wheel (kit wheels), the center wheel chained to the front and back wheels. Tensioned with UMHW off-center cams. The whole drive system was put together using nothing more than a cordless drill and a band saw (neglecting the welded frame which could be replaced by a kit frame).

THE MOST IMPORTANT THING FOR A ROOKIE TEAM IS TO GET THEIR DRIVE SYSTEM WORKING BY THE END OF THE THIRD WEEK. Our team was very competitive last year, simple because of it's dependablity and driver practice. Our drive team had over a week of practice before ship.

3) Use innovative ideas to develop a cool manipulator or strategy. There's absolutely nothing more troubling than to have your robot unable to drive.

I think back to the days of 2000, 2001, 2002 when there was no simple drive system that came in the kit. You had some cordless drill motors, transmissions, and skyway wheels that you had to make work. That took our team the bulk of the 6 weeks.

[K-Diddy]

We had a 3WD drivetrain with crab drive.

[Tom Bottiglieri]

Originally Posted by Brandon Holley

I guess what I fear is a team saying, well yeah we could do that, but why don't we just buy this and call it done...

I don't think anyone is doing that. I just think the drive base aspect of FIRST robots has gotten to a point where it is more economical to "drag and drop" components. Design considerations for "learning" can be made in other areas of the robot.

You have when to hold your cards and when to fold your cards. Take the ace that FIRST has given us and build the kit chassis or a simple 6wd setup with COTS. The time you save will allow you to invest more into your functional designs, which are usually the things that make or break a robot design.

[thefro526]

We used a six wheel drive based off the KOP and it worked great and never failed during a real match. It was easy to build and easy to service (not that we ever really had to). I would recommend any rookie team try a six wheel drive with the KOP Wheels and Transmissions just because of the reliably aspect. My favorite feature I integrated into the drive this year was using bolts as axles and pvc spacers to keep everything in place. It made dropping a wheel out take all of 2 seconds and we saved at least a pound or two by not using collars.

[=Martin=Taylor=]

The Following is Fact:
The purpose of the drive train is to get the manipulator in position to score as many points as possible in the minimum amount of time.

That’s all you need to know. The best drive train is the one that accomplishes this task the best. Period.

It doesn't mater how many wheels, chains, motors, or gears you use. If 2WD will get the job done, use 2WD and use the time/money/resources saved to work on the manipulator.


The Following is my Opinion:

Historically, 6WD with the center wheel lowered has been the fastest system.

If you need to climb a ramp, 6WD is necessary to prevent bottoming-out.

But if the playing field is flat, why bother with 6WD at all? I mean, two of the wheels aren't even touching the ground. So why bother? If the playing field is flat, save yourself some weight and only use 4WD with the wheel axels spaced 12" apart and casters on the front.

Some will argue that swerve, mechanum, omni, and tank treads offer advantages over 6/4WD. But when you consider how many teams consistently win regionals and championships without these systems it becomes hard to see what the real benefit is.

As for the 25 vs 35 argument...

2006 Our team used #25 in 2006 with the kit frame and it was a COLLOSAL disaster. We couldn't keep the chains in line, or tensioned and they kept falling off.

2007 Deciding we would never use tensioners or #25 chains again we switched to #35 with movable axels for tensioning. This dive train was very robust, but also very heavy.

2008 Deciding that maybe we had been too rough on #25 we switched back, implementing the same movable-axel tensioning system we used in 2007. Worked great. Much lighter and we never lost a chain.

So it really comes down to tensioning. #25 needs to be really tight, so you need a good tensioning system. #35 can be much looser. Some teams like 766 and 330 have gotten away without using any tensioners.

[Lil' Lavery]

Originally Posted by Hachiban VIII

The Following is Fact:
The purpose of the drive train is to get the manipulator in position to score as many points as possible in the minimum amount of time.

That’s all you need to know. The best drive train is the one that accomplishes this task the best. Period.

It doesn't mater how many wheels, chains, motors, or gears you use. If 2WD will get the job done, use 2WD and use the time/money/resources saved to work on the manipulator.

What about defensive robots?

[JVN]

Originally Posted by Hachiban VIII

But if the playing field is flat, why bother with 6WD at all? I mean, two of the wheels aren't even touching the ground. So why bother? If the playing field is flat, save yourself some weight and only use 4WD with the wheel axels spaced 12" apart and casters on the front.

Two Words:
Normal Force.

If you want to build a robot capable of pushing, you want every ounce of weight sitting on top of the highest traction drive wheels you can find. If you have weight sitting on non-driven wheels (like casters) you are reducing the amount of normal force available.

In some games or for some strategies pushing is not an important consideration. In these cases a 6WD may not be important.

Brief Aside:
In my opinion drivetrain design is very formulaic. "If you want to do X then you need to do Y." It is all about the functionality requirements your team has, and the design trade-offs you are willing to make (these trade-offs may involve things like weight & team resources, or things like pushing power & top speed.)

I should write a paper on this sort of thing.

-John

[=Martin=Taylor=]

Originally Posted by Lil' Lavery

What about defensive robots?

You can win playing defense. But you can't win without playing offense.

Defensive robots depend on offensive robots to win. Building a robot that can't score, or even one that isn't designed with scoring as the primary strategy, is risky.