Control System Anomaly

[twiggzee]

We have a simple robot with nothing other than a 4 motor direct drive mecanum wheels system on it. a couple of strange things are happening.

1) it drives faster backwards than it does forward. i doubt it is a code issue and all the wiring seems sound. we can't figure out why this is happening? these are jaguars that we got 2 years ago, perhaps they need to be calibrated if that's even possible? if so, how do we do that?

2) when we are driving at a fast speed (anything more than 30 or 40% power) forward and then suddenly switch to backwards, the robot dies. Meaning, we lose all control through the drivers station, the robot just stops moving. We don't lose connection or anything. When we disable and re-enable it works again.

There is no other programming than the simple mecanum drive code. Any ideas on where else we can troubleshoot?

Thanks!
Wendy
3502

[Cal578]

Mecanum drive is fun. It can be complicated, so it's good to see that you're working on it off-season.

A mecanum system is sensitive to torque and traction. If you're driving each wheel directly from a motor, with no gearbox, then you'll have wheel slippage and/or motors that are at or near stall. This can contribute to different speeds depending on direction. Also, where is your center of gravity? That will also impact traction, and will be different depending on the direction of acceleration. Frame rigidity has a similar effect.

The forward-to-reverse "death" may be caused by the direct drive. When you change direction suddenly, that's a huge current draw, temporarily worse than a short circuit. If the motor controller detects an overcurrent, it will throw an error and stop. You may be able to handle this in software and reset automatically. A better way is for the software to not allow the sudden reversal; it can ramp up to the desired voltage in the other direction. If you're using Jaguars and CAN, there is a command that will change the voltage over a short period of time. Best to do both the voltage ramp and add a mechanical reduction between the motor and wheel, to improve torque.

I hope this helps. Keep us appraised of your progress. Pictures are always nice. 8-)

[Ether]

Originally Posted by Cal578

If you're using Jaguars and CAN, there is a command that will change the voltage over a short period of time.

You don't need CAN to get voltage ramping in the Jags.

There is built-in voltage ramping which can be enabled by moving the limit switch jumpers.

See Q7 on Page 3 of this document.

[twiggzee]

Just to update, we isolated the problem to be the cRio. since we got a new cRio in prep for this year's game, we decided to switch out the old cRio on this bot and it works just fine again. going to call NI tomorrow to see if they will replace/repair this one.

[Greg McKaskle]

Are you sure it is the cRIO and not the power to the cRIO? Specifically, the cRIO is supposed to be powered off of the boosted 24V portion of the PD. If it is powered off of another circuit and the power drops below 9V, then it is normal that it would reboot, meaning it is how it was designed and is a requirement that you supply 9V. The new one will do the same.

Gre gMcKaskle

[twiggzee]

Originally Posted by Greg McKaskle

Are you sure it is the cRIO and not the power to the cRIO? Specifically, the cRIO is supposed to be powered off of the boosted 24V portion of the PD. If it is powered off of another circuit and the power drops below 9V, then it is normal that it would reboot, meaning it is how it was designed and is a requirement that you supply 9V. The new one will do the same.

Gre gMcKaskle

Another update. So when the 4slot crio worked we thought we had a bad crio but then we replaced our old 8slot with a new 8slot so we could use the same mounting holes but we had the same problem again. after lots more investigating we figured out that all this time we didn't have any power going to the sidecar. it must have been getting its power only from the crio and when we used too much power the robot would stop. everything works fine now. pretty embarrassing!!

[Greg McKaskle]

Thanks for the update. That sounds like a more likely diagnosis.

Greg McKaskle

[z_beeblebrox]

You said the wheels are "direct drive". If the CIMs are directly running the wheels without betting heated down, that could be very problematic.

[twiggzee]

Originally Posted by z_beeblebrox

You said the wheels are "direct drive". If the CIMs are directly running the wheels without betting heated down, that could be very problematic.

by direct drive i mean CIM motor into toughbox nano gearbox to wheel. i don't know what you mean by "betting heated down" but we've done this same drive system for the past 2 years with no problems. aside from the wiring issue here. =/

[Alan Anderson]

Originally Posted by twiggzee

i don't know what you mean by "betting heated down"...

I'm assuming an autocorrect failure on the phrase "getting geared down".