High speed Drivetrains

[Dragonslayerend]

I noticed this year that the high scoring robots were ones with extremely fast drive trains. The faster the robot the more advantage the robot had against avoiding blockers and doing laps. Though how come there weren't many robots I saw that couldn't exceed atleast 12 fps. What was your reasoning? Defense? Or was it because of technical issues?

[MichaelBick]

I noticed this year that the high scoring robots were ones with extremely fast drive trains. The faster the robot the more advantage the robot had against avoiding blockers and doing laps. Though how come there weren't many robots I saw that couldn't exceed atleast 12 fps. What was your reasoning? Defense? Or was it because of technical issues?

For one, most of the teams with high speed drivetrains also have shifters. This allows them to run their drivetrain really fast while still having a lower speed to push through and play defense. However shifters are mechanically complex, and can be heavy to integrate. Also the high scoring teams have a lot of drive practice allowing them to decelerate less and play more of the game at their top velocity.

[dcarr]

doing laps

What year is this again? :P

[StevenB]

Elite teams do not just drive fast, they are fast at everything. Why? Because there is only 2:15 in the match to score points. Elite teams put in the design work to get the most out of every motor and mechanism they put on their robot. Most of us are content to get the motor to drive the mechanism at all, but elite teams do the math to figure out what designs and gear/pulley ratios will actually squeeze the most performance out of the motors they have.

The idea that a slow robot has more pushing power than a fast robot is not fully correct. If you have six motors on your drive and wheels with normal-ish friction, you can build a single-speed traction-limited drivetrain that still goes really fast. You should always build things to be as fast as possible (but no faster).

[Jibri Wright]

Team 1781 went with Super Shifters this year. This is why we had some good offensive and defensive capabilities this year. I know that Team 111 had Super Shifters as well but instead of two speeds, they had three! They customly designed a third gear shift that was at a really low gear making them faster than us. As for squeezing the most out of every motor, we didn't take that into account. Then again we're not an elite team quite yet.

[Dragonslayerend]

Elite teams do not just drive fast, they are fast at everything. Why? Because there is only 2:15 in the match to score points. Elite teams put in the design work to get the most out of every motor and mechanism they put on their robot. Most of us are content to get the motor to drive the mechanism at all, but elite teams do the math to figure out what designs and gear/pulley ratios will actually squeeze the most performance out of the motors they have.

The idea that a slow robot has more pushing power than a fast robot is not fully correct. If you have six motors on your drive and wheels with normal-ish friction, you can build a single-speed traction-limited drivetrain that still goes really fast. You should always build things to be as fast as possible (but no faster).

Oh The year I'm talking about is 2013. I noticed that many teams were using mini toughbox with 10:1 gear ratio but extremely slower(like 6 fps) than our 8:1 geared robot(which runs at 12fps). I was wondering how such issues happen. Or whether it was actually planned to be a defensive robot. Actual design goals of teams are what I'm looking for. What was the intent your team had for a drive train and what was the result?

[orangemoore]

What year is this again? :P

That was 2008 Overdrive

[wilful]

If you have six motors on your drive and wheels with normal-ish friction, you can build a single-speed traction-limited drivetrain that still goes really fast.
However with six CIMs you have to take into account the current they are drawing. If you end up stalling all six at once, and each is drawing 40 amps (or more) through its breaker on the PD board, there are 240 amps being drawn through the main breaker, which is rated to 120 amps. At least one team at SVR ran into this problem.

[MichaelBick]

However with six CIMs you have to take into account the current they are drawing. If you end up stalling all six at once, and each is drawing 40 amps (or more) through its breaker on the PD board, there are 240 amps being drawn through the main breaker, which is rated to 120 amps. At least one team at SVR ran into this problem.

You can't stall a traction limited drive

[RyanCahoon]

Another issue to consider is acceleration. Andrew Palardy [apalrd] and others have a number of great posts on this already, but gearing for more torque, in addition to greater pushing ability, also allows you to accelerate faster. Especially given the restricted field (you have to drive around the pyramids) and the level of defense this year, it does no good to have a high theoretical top speed if you'll never be able to reach it. There's always a few rookie teams each year that try to gear their drive trains for 24 ft/s and learn this lesson the hard way when their robot won't move.

[MichaelBick]

Another issue to consider is acceleration. Andrew Palardy [apalrd] and others have a number of great posts on this already, but gearing for more torque, in addition to greater pushing ability, also allows you to accelerate faster. Especially given the restricted field (you have to drive around the pyramids) and the level of defense this year, it does no good to have a high theoretical top speed if you'll never be able to reach it. There's always a few rookie teams each year that try to gear their drive trains for 24 ft/s and learn this lesson the hard way when their robot won't move.

The other way of accelerating faster is adding more motors(something many teams have done this year). The field is actually pretty open for robots that are less than 30" tall.

[efoote868]

Most years teams limit themselves to 12 fps or less because 12 fps is a sweet spot when it comes to acceleration in drive trains.

You can design a drive train that will go 40 fps; it'll never reach that speed on an FRC field. A better way to design a fast drive train is to figure out the maximum straight line travel your robot will ever have to go (be it 10, 20, or 40 feet), and then design the gearing around that.

I'll have to look into it (and do a search), but if I recall correctly there was a model that showed that a single speed 4 cim drive geared to a maximum 12fps went from dead stop to 20 feet in about the same amount of time as a drive geared to a maximum 14fps, and faster than a drive geared to 16fps. The 12fps drive had the advantage of more pushing power than the 14fps drive.

EDIT:
Was this thread I was thinking of.
http://www.chiefdelphi.com/forums/showthread.php?t=110382&highlight=drive+model+simulation

[Dragonslayerend]

Most years teams limit themselves to 12 fps or less because 12 fps is a sweet spot when it comes to acceleration in drive trains.

You can design a drive train that will go 40 fps; it'll never reach that speed on an FRC field. A better way to design a fast drive train is to figure out the maximum straight line travel your robot will ever have to go (be it 10, 20, or 40 feet), and then design the gearing around that.

I'll have to look into it (and do a search), but if I recall correctly there was a model that showed that a single speed 4 cim drive geared to a maximum 12fps went from dead stop to 20 feet in about the same amount of time as a drive geared to a maximum 14fps, and faster than a drive geared to 16fps. The 12fps drive had the advantage of more pushing power than the 14fps drive.

EDIT:
Was this thread I was thinking of.
http://www.chiefdelphi.com/forums/showthread.php?t=110382&highlight=drive+model+simulation
We will do a drag race with multiple neighboring teams that runs at 6 fps to 14 fps to see which actually is a faster robot for accel and distance travel. N post it up.

[efoote868]

We will do a drag race with multiple neighboring teams that runs at 6 fps to 14 fps to see which actually is a faster robot for accel and distance travel. N post it up.

Unfortunately there will be factors that are difficult to control with different robots, for instance the amount of wiring the robot uses, the size and weight of the wheels and gearboxes, as well as the weights of the robots.

Regardless, the results will still be interesting. I'm sure we all would appreciate as much information as you can post!

[IndySam]

We started with a shifting drive this year but after problems we switched to single speed. We have a 4 cim drive geared to 17.3ft/s and 14.01 adjusted speed (acording to the JVN DesignCalc.) We were also light at 85lb's.

We were the fastest robot at both regionals we attended both in top speed and acceleration.

A few key points to going fast:

Keep the weight down, This kind of speed is much more difficult to achieve at 120lbs. If you want to go fast you have to watch every ounce from day one of your design.

Unless the game screams for it, we will never shift again. If the rules are the same next year, we will use a six motor drive.

You must have a practice robot. A real fast robot is difficult to drive. Our driver Nick made it look easy because he has many hours of practice driving a fast robot.

Pushing match, what pushing match? If they can't catch you they can't push you around. It's difficult to play effective defense against a real fast robot. Most robots would line up to block us but by the time they were ready we were already gone.

[RyanCahoon]

A real fast robot is difficult to drive. Our driver Nick made it look easy because he has many hours of practice driving a fast robot.

Pushing match, what pushing match? If they can't catch you they can't push you around. It's difficult to play effective defense against a real fast robot.

Did you do any "drills" that you found particularly effective, especially in teaching your driver to avoid and react well to defense, or did you just run practice matches?

[thefro526]

You can't stall a traction limited drive

Just because you can't stall a traction limited drive doesn't mean that it won't draw near-stall current.

We've popped at least one main breaker this season due to an aggressive pushing match, and I wouldn't be surprised if a handful of other teams have as well.

[karomata]

The more experienced teams tend to better understand the large advantage that can be had when you drivetrain is fast yet maneuverable. Take 148 for example, although they do not always have the best scoring mechanism in FIRST, they continue to do exceedingly well partly because of their amazing nona (http://www.youtube.com/watch?v=_hTyXQUgYLE) drivetrains.

[PayneTrain]

Just because you can't stall a traction limited drive doesn't mean that it won't draw near-stall current.

We've popped at least one main breaker this season due to an aggressive pushing match, and I wouldn't be surprised if a handful of other teams have as well.

422 built a friction-limited 6CIM drive and popped the breaker twice at their first event after getting in an intense pushing match with the feeder station wall thanks to a miscommunication with the drive encoders.

[efoote868]

422 built a friction-limited 6CIM drive and popped the breaker twice at their first event after getting in an intense pushing match with the feeder station wall thanks to a miscommunication with the drive encoders.

Any idea how long you were stalling the motors? I wonder what went first, some of the 40a's or the 160a.

[fb39ca4]

Did you do any "drills" that you found particularly effective, especially in teaching your driver to avoid and react well to defense, or did you just run practice matches?

Our team used our logomotion robot with our practice bot to practice playing and avoiding defense.

[AlecMataloni]

I know that Team 111 had Super Shifters as well but instead of two speeds, they had three! They customly designed a third gear shift that was at a really low gear making them faster than us.

We actually used 2 speed VexPro Shifters with a 3rd stage provided by WCP which matched our desired speeds for 6in wheels. Sorry if you were misinformed. :D

[its da PAT!!!]

Did you do any "drills" that you found particularly effective, especially in teaching your driver to avoid and react well to defense, or did you just run practice matches?


We don't really focus on defensive drills. But our driver knows the robot and I simply tell him whats going on with other robots. Then in his crafty mind, he'll decided where to go. But to tell you the truth, he'll sike me out a lot....

[pmangels17]

This year 271 went with the most intense custom transmission we have ever done. 6 cims, on a two speed shifter, geared for roughly 16 fps in high speed and 6 fps in low speed. Also, we have a Power Take Off for our climber, but our drive is ridiculous. We push robots sideways with all our torque, and we can race around the pyramid in high speed. Also, we went up against a bot with mecanum wheels, and pushed them twice all the way around the pyramid. Simple 6 wheel dropped center, with double wide center wheels, all 6 with roughtop tread. We peaked at 314 amps one match, but it was only for a split second, so we were fine. We have ammeters on two of our drive motors, so just multiply by three to find total amps. Oh, the pushing. I'd say that for any team that even just wants to upgrade from the kitbase, add a two speed shifter, the versatility lets your robot do so much more, and if used right, it gives you a lot more options. We have used two speed drives many years. We didn't in 2012, but for the years that I participated in so far, 2011, 2012, and 2013, we had 2 speed drives. We haven't had many problems at all.

[TheMadCADer]

The more experienced teams tend to better understand the large advantage that can be had when you drivetrain is fast yet maneuverable. Take 148 for example, although they do not always have the best scoring mechanism in FIRST, they continue to do exceedingly well partly because of their amazing nona (http://www.youtube.com/watch?v=_hTyXQUgYLE) drivetrains.

I should point out that 148 has been using a slightly different version of that style of drivetrain in 2011 and 2013 (in 2012 they used a fairly standard tank drive), which they call a butterfly drive. It's pretty much the same as a nona-drive, but there is no perpendicular omni-wheel. It allows for similarly 'drifty' driving, but no strafing. In 2011 they also liked being able to toggle their traction wheels down in order to drive in a very straight line during autonomous.