paper: Cyber Blue Drive System Test Report

[Chris Fultz]

Thread created automatically to discuss a document in CD-Media.

Cyber Blue Drive System Test Report by Chris Fultz

[Chris Fultz]

This fall we completed acceleration and top speed testing on 3 single reduction gearboxes to compare the performance of each one.

This report summarizes the test methods, data, results and conclusions from the testing.

We welcome questions and "peer review" of the process and results.

[Michael Hill]

Great paper Chris!

Can you explain a little more what the "Block" time is?

[Chris Fultz]

To determine top speed, we used the data for the 40 - 50 foot segment of the test. The robot was at top speed by then in all cases.

So, we took the time required to travel that 10 feet (in seconds) and converted that to feet per second to determine "top speed". For the report, we called that 40 - 50 foot segment the "block" time.

In reality, it is the average speed for that final segment of the test run.

[Hugh Meyer]

Hi Chris,

Nice paper! Looks like you had a fun summer.

Was the drive run open or closed loop? Did you use Victors, Jags, or Talons for the speed controller? Did you collect any data on the current draw during the runs?

Did you have the encoder on the CIM motor or transmission output shaft?

-Hugh

[Chris Fultz]

Open Loop. When engaged, we turned the motors on to full speed.

Victor 888's.

No current draw data. We only captured time stamp and encoder count.

Encoder was on the output shaft of the gearbox.

[Alpha Beta]

Enjoyed reading the paper. I teach experimental design in IB Chemistry, and appreciate the detail that goes into getting meaningful results. Thanks for spelling it all out.

Would love to see the results repeated with a 3rd CIM on each side, and expanded to some more aggressive gear ratios. Hopefully this paper will encourage more teams to conduct similar tests and publish their results.

The wisdom of the statement "uninterrupted travel distance" is spot-on for applying these results. If you are proficient at scoring, expect defensive pressure to break up what might initially seem like a wide open field.

[Chris Fultz]

We have a full set of AM gearboxes and multiple sprocket sets that are part of a larger design project (will have to be in 2014). The hardware we have from AM will allow us to take additional data and use a wider spread of gear ratios. We didn't have time to do it this fall.

The VEXPRO sets made for a simple test and easy swap-out but we were limited to the three reduction options since we did not want to add in the variable of different belts and sprockets.

[DampRobot]

Thanks to everyone from 234 for doing this awesome research and paper! I really enjoyed reading it.

I'm glad that this confirms my recent belief that "more acceleration is better in most cases" really isn't the way things work. It appears, given the data, that a slower geared robot will not accelerate noticeably faster, but will have a slower top speed. Within parameters considered reasonable for FRC, you really do go faster when you gear faster.

Of course, there may be arguments about drivability or the ease of coding autonomous, but I'm glad to have some evidence about whether or not gearing slower can get you to a distance faster in real world conditions.