View Single Post
Old 11-07-2014, 02:54 AM   #1295 (permalink)
Permanent Apprentice
Join Date: Jul 2010
Location: norcal oosae
Posts: 523
Thanks: 351
Thanked 312 Times in 214 Posts
Originally Posted by Cyruscosmo View Post
First thought that comes to my mind is powering into a corner with both wheels trying to do the same speed independent of the other. If the steering is compromised in any way you will go where the motors tell you to go which is straight. Think tank type steering.

A traction device in a differential is the way I will go over trusting any form of code to get it right. All you need is one bad rainy day and you could spend the rest of your life in a wheelchair or worse.

Now that being said I know there will come a day when wheel motors will be the new thing and the control devices will get so good that the mean time between failure is a moot point but why make something that complex when a simple mechanical device will do it right now the right way every time.

Use the sensors already in place to manage wheel spin or Rev limit the electric motors so they don't blow up when a wheel slips and simply let go of the throttle.

The important thing here is that it is NOT a speed control issue. This makes the whole thing significantly different than attempting to match the speed of two independant motors. In fact, from a controls perspective, speed control requires another complete control PID loop. So with only torque control the motors won't be fighting you like tank steering. If things go really awry, one motor may produce 25% more torque than the other, but this is much different than the fight that could happen when one motor tries to go 25% faster than the other. In that case, the control loop may command FULL torque until the speeds match, or negative torque on one motor with increased torque on the opposite - which could have awful results like you are warning about.

I too am leary about putting too much computer (code) between the driver and control. This is why I'm trying for a simple analog circuit with a bypass switch. The bypass switch would make the commanded TORQUE from each motor identical.

I'm not completely convinced locker differentials are a simple solution. Putting one in an MGR would be a very expensive custom job.

When I was growing up, my dad had a Chevy Malibu with a "limited slip" differential. We went skiing every opportunity, and I can say my dad's excellent driving made up for the differential's weird behavior. If one wheel slipped, it would clutch and force all the torque to the opposite wheel, which could slip because it all off a sudden got all the torque. Then it would clutch to the opposite side. Back & forth, fishtailing down the road because of the "help" from the differential. According to a friend with lots of off-road experience one needs to be very careful with locker choice, because the on-off behavior is pretty common.
  Reply With Quote