gear ratio

[Kevin Sevcik]

Originally Posted by Madison

...operating with the unwritten assumption that all other things are equal.

A robot with a 9:1 ratio and 6" wheels will be slower than one with a 9:1 ratio and 8" wheels.

I give presentations about drive train design now and again and I sometimes assume a level of understanding that's not there yet, so I just wanted to explicitly note that gear ratio alone does not determine a robot's final speed.

To further the point, the final speed also depends on a lot of other factors like the efficiency of the gearbox, friction and losses in the drive train, and the inertia the motor is trying to accelerate. If the rule were simply that a lower ratio gearbox = higher final speed, then we'd all be direct driving CIMs to our wheel. But, of course, a direct drive CIM is unlikely to generate enough torque to even get the robot started rolling, would take forever to accelerate up to top speed, and would probably blow a breaker attempting to get there. That'd be another trade-off you're making with the gearbox. A lower ratio gearbox is going to take longer to get up to its top speed. It's even going to take longer to get up to the top speed of a higher ratio, lower speed gearbox. So, in summary, simple rules of thumb are usually highly simplified and approximate.

[kramarczyk]

If you want to get a quick look at how gear ratios impact speed I recommend taking a look at this infamous piece of work... http://www.chiefdelphi.com/media/papers/1469
Take long look at the drivetrain modeling sheet, especially the velocity vs time plot.
Deconstruct it, learn it, build it better.

[Dick Linn]

Does anyone have some actual numbers of how fast their robot managed to go as a percentage of the max unloaded motor RPM? I mean as-built with gearbox loss, etc. I'm not interested in feet/sec. figures. I was just wondering if something like 80% was realistic.

Originally Posted by squirrel

The lower ratio (9:1) will give a faster moving robot.

Originally Posted by Madison

...operating with the unwritten assumption that all other things are equal.

Yeah. 9:1 will (probably, depending on the motor...) give you a higher top speed. Acceleration is a totally different ball game, and requires some really fun math that no one ever got around to teaching me. (to be totally honest, I'm not a math guy.... Hence the Sociology major)

I seem to recall from my lurking there being a spreadsheet floating around that calculates acceleration as well as top speed for FRC bots... Am I mistaken? Anyone know where this tool is?

[Cory]

Originally Posted by Dick Linn

Does anyone have some actual numbers of how fast their robot managed to go as a percentage of the max unloaded motor RPM? I mean as-built with gearbox loss, etc. I'm not interested in feet/sec. figures. I was just wondering if something like 80% was realistic.

It's pretty close. 968 did a test of actual ground speed vs calculated theoretical speed of our 2008 robot. The actual speed was approximately 80% of the no load 100% efficiency speed.

[Brandon Holley]

Originally Posted by Cory

It's pretty close. 968 did a test of actual ground speed vs calculated theoretical speed of our 2008 robot. The actual speed was approximately 80% of the no load 100% efficiency speed.

Was that calculated from the ~5500 free speed CIM RPM ? Or were you assuming some sort of amperage being pulled?

[Kevin Sevcik]

Originally Posted by Brandon Holley

Was that calculated from the ~5500 free speed CIM RPM ? Or were you assuming some sort of amperage being pulled?

Originally Posted by Cory

The actual speed was approximately 80% of the no load 100% efficiency speed.

No-load = free speed. That is, the speed the motor turns at when nothing at all is connected to it. Note, that this doesn't mean no amps are being pulled. There's still a small amount of friction to be overcome etc. Plus, the rotor coils are constantly changing orientations and making and breaking contacts. This means the magnetic fields are constantly collapsing and reforming, which means the back emf generated by the motor will vary over time. Since the voltage supply nominally stays the same and the coils have some resistance, current is flowing whenever the back emf doesn't exactly equal the input voltage. So all DC motors have a no-load current as well as a no-load speed.

Which brings me to another silly technical point. Since the CIM is drawing current it has some non-zero input power, even at no-load. However, since there isn't a load output from the shaft, it has zero output power. So the motor's efficiency at no-load is actually 0%, not 100%.

[Cory]

Originally Posted by Kevin Sevcik

No-load = free speed. That is, the speed the motor turns at when nothing at all is connected to it. Note, that this doesn't mean no amps are being pulled. There's still a small amount of friction to be overcome etc. Plus, the rotor coils are constantly changing orientations and making and breaking contacts. This means the magnetic fields are constantly collapsing and reforming, which means the back emf generated by the motor will vary over time. Since the voltage supply nominally stays the same and the coils have some resistance, current is flowing whenever the back emf doesn't exactly equal the input voltage. So all DC motors have a no-load current as well as a no-load speed.

Which brings me to another silly technical point. Since the CIM is drawing current it has some non-zero input power, even at no-load. However, since there isn't a load output from the shaft, it has zero output power. So the motor's efficiency at no-load is actually 0%, not 100%.

I was referring to no load in the sense of robot sitting there with wheels in the air-ie:no load on the gearbox, not no load on the motor. I was a little unclear there.

As to the second point, I meant 100% efficiency as in zero frictional losses in the drivetrain, not 100% motor efficiency.