Week 7, Day 9, 11/3/12

Thanks to Vex for allowing us the opportunity to test the new Victor 888's. We've been playing with them for about a week now, and we're impressed.

Here it is in all it's glory. The Victor 888. What's that? You say the picture below is an 884? Well, every single Victor 888 we've received to test has been packaged in pretty much the standard Victor fare we've seen. In fact, some were rebadged 884's. Why mess with a good thing, right?

Don't worry. Once you look at the performance, you won't confuse an 888 with an 884 again.

Victor 884... Errr... 888

Today we worked on Victor 888 Linearity Tests, along with Velocity control Tests on All Speed Controllers.

Below, there are two Victor curves included in the graph. The more curved Victor line, in red, is a Victor hooked to a window motor driving a belt. The belt is slips on a shaft that is locked in place.

We were impressed with the substantially improved performance of the Victor here. However, during the test we noticed another problem. The window motor we were using for the tests had begun to heat up and the torque output was dropping. As a result, the motor was stalling, which it didn't do on any other of our motor controller tests. (This was due primarily to us playing with the Victor 888 more than the other controllers.)

In order to correct that problem, we let the motor cool and performed another series of points. These points are shown in blue/green and are called Victor 888 Cool. They show huge improvements in linearity over the original test, and over a normal Victor. It looks like Vex has really improved the response.

Victor 888 Output With Hot Motor Running Stalled Roller, and then Cold Motor

Next, we zoomed in on the Zero point and used smaller PWM increments. You can see that the Victor 888 has a slightly different slope due to it's slight non-linearity, but you can see in the area around zero it is fairly linear. In addition, the deadband of the Victor is as good as the Talon.

Victor 888 Output Cold Zoomed In

Next, we ran a test based on a comment on Chief Delphi. Team 254 linearizes their drivetrain by measuring the velocity of the drivetrain and using that in their linearization function. We elevated our robot and ran through a series of PWM commands. At each state, we allowed the velocity to steady out then took a reading. At the end we graphed PWM input vs. Velocity output. Velocity is expressed in inches/second. Our drivetrain is a toughbox mini, running 3 wheels.

PWM vs. Velocity for Victor, Talon, Victor 888, and Jaguar

The graph above really tells it all. The Jaguar and Talon win on linearity. The Victor 888 is miles ahead of it's older brother the 884, and not far behind the Jag and Talon. The original test for the Victor 888 was using a window motor, and it appears the change to a Cim with increased resistance improved it's linearity substantially.

The 888 is a significant improvement over the 884, and comes close to matching the Talon and Jaguar for linearity. It's deadband is as small as the Talon's, and it's footprint is much smaller than a Jag. It's another great product for Vex, and we have no doubt you'll be seeing a lot of them on robots this year.