Calculating motor resistance (NPC-T64)

[misterT]

Nice to see somebody actually finishing a project successfully I've seen too many ambitious projects fail. Good work!

 

[sebgus]

Thanks!

We could improve stability if we could calculate the speed of the vehicle. Unfortunately we don't have time to order and implement a angle sensor for the motors, is there any other way we could estimate the speed? I am kinda out of bright ideas. Thinking about integration of the accelerometer-signal but that would not work when the angle of the vehicle changes (and integration errors if there are noise and static errors). Another idea was to use the output from the microcontroller to the motorcontroller some how.

So anyone with an idea?

 

[misterT]

Measuring the back-emf would be one option, but might be difficult to implement with your existing motor controller. Back-EMF Motion Feedback
Integrating the accelerometer signal should be stable. Derivative can be very noisy and problematic, but integration is stable. If the accelerometer is 3-axis you can compensate for the vehicle angle and get a pretty good estimate of the speed.

 

[sebgus]

Yeah I guess it would be too difficult

Hmm, I've read several threads on different forums claiming that it doesn't work and some claiming that it does work. I guess it is all about how good you design it. I'll try to work something out here.

EDIT: Isn't there a problem with the gravitational acceleration with the accelerometer? Or am I just stupid right now...

 

[sebgus]

Anyone?

 

[cachehiker]

several threads on different forums claiming that it doesn't work and some claiming that it does work.

Helped make a similar system work myself. It was also a cycle skipping system synchronized to the power line, skipping a single cycle between zero crossings, and using a sample and hold to grab the voltage being generated during that 125µs period. I've also seen several "engineers" here fail to make the concept work with modern microprocessors. They can't seem to grasp that having a response (gain) that's 180° out of phase makes it unstable. Duh. I just have a math degree though so I can't help them.