Question on Motor Bias

[IntensifiedAmp]

With a finished robot, we're not so finished.

We seem to have a slight problem at extremely low speeds, with our drive train. We're using 4 CIMs, Banebots planetary transmissions, and KOP AndyMark wheels.

At low speeds, when moving the joystick forward, one wheel moves faster than the other. However, when moving the joystick backwards, the other wheels moves faster than the original one.

This implies that at higher speeds, the motors move in sync.

Could this be a possibility of motor bias? Or is it a trimming issue (which by the way, may take forever due to the ultra-sensitive touch) ?

Thanks for the help in advance!

Ah, and yes, our wheels were off the ground at the time.

[Miner]

We have found that the wheels, really the motors, go faster while going backwards. The victors can be calibrated or your programming team can create an auto calibrator for the victors. We did last year, it was cool, and pointless.

Trimming is extremely hard with the joysticks, im assuming, and I read in an earlier thread that a team modified their joysticks by putting stronger springs in their joysticks. This made them less sensitive and a lot easier to control.

Good Luck
Jacob

[David]

Originally Posted by IntensifiedAmp

This implies that at higher speeds, the motors move in sync.

Don't be so sure. They may appear to be unbiased at high speeds, but the can be very different we tested to and found that the can be up to 25% biased). My team went on to further test the motors and found that the motors themselves are slightly (<2.5%) biased, but not otherwise different (also the position of IFI). A word of advise is to check your speed controller output, you speed controllers can be giving very different outputs even if the all have the same PWM input. Calibration does help speed controllers become more consistent, but also try to use speed controllers purchased at the same time to avoid manufacturing variation over time.

[LieAfterLie]

I have a relevant question...

Is the CIM motor bias linear? Or rather, is the (change in rpm)/(change in pwm input value) constant over the entire 0-254 range?

I understand that the "not moving" pwm value for the individual motor might not be exactly 127, and that it's 254 rpm might be faster than it's 0 rpm or vice versa. (The victors have a built in dead band centered around 127 I think, so testing for the exact standstill value might be hard...)

Say your wheel goes (unloaded) 800 rpm one way (254), but 700 rpm the other way (0). Could I correctly assume that the wheel's rpm will change about 5.9 rpm per pwm value? And thus that the standstill point is at a pwm output (without dead band) of about 119 not 127?

Same thing for victor bias, does its and the motor's bias together behave like this?

[IntensifiedAmp]

Thank you David and Miner,

We're probably going to go with changing the values around a bit with each set of motors. I hope that helps.

But you see, I'm not in the programming team. So fate rests with them!

Good luck.