Differential Steering RWD

<fieldset class="swiftfieldset"><legend>RoboteQ AX2850HE</legend> <table border="0" cellpadding="2" cellspacing="0" width="100%"> <tbody><tr> <td><span class="smalltext">I would like to use the AX2850 motor controller to control 2 ETEK-R motors in a rear wheel drive setup on a dune buggy. My concern is the control of the motors to prevent skidding while turning. Can this controller simulate the motion transfer of a differential (i.e. outside wheel spins faster then inside wheel while turning?)

I understand the controller allows for Mixed mode control both in open loop and closed loop, but that this mixed mode control is in place for "Tank Style Steering" meaning that one motor would be reversed while the other one is moved forward.

Would the motors naturally allow themselves to speed up upon turning, and I should not be concerned with this?

Any input on my concern would be appreciated.
Thanks
Mark</span></td></tr></tbody></table></fieldset>

Mixed mode would not work well to achieve what you need. Left and Right motor speed is set by the value of the steering command. One motor will not turn faster as the other one turns slower.

You could wire the motors in series, this way as one motor slows down and tries to draw more current, the other would spin faster. However, it would be harder to get the max power for each motor.

In any case, if you are trying to drive a vehicle in the sand, the Etek motor will draw potentially very large current and it is not likely that our controller will perform under this kind of extreme stress.

Thank you for your prompt response.

By wiring the motors in series, would I still need a 2 channel motor controller?

It was reccomended to me not to run 2 motors off of one channel.

Please explain.

The vehicle (dune buggy) is just a test bed for future development, it was purchased cheap and will be only driven for testing purposes to show proof of concept.

What he's saying is that the Eteks have a stall torque rating of something like 900amps.

That is to say that if you were to keep the motor shaft immobile and current draw would rapidly shoot up to a level that would likely smoke even the HE versions of Roboteq's controllers. And indeed nearly every other controller on the market.

I imagine that the recommendation about running both the motors on one channel is related to overtaxing a controller as well.

As far as I know only the Zilla controller for performance EV's is rated to handle the kind of current you seem to be looking to control. It's hard to get (There is a waiting list), they expect you to water cool it for extended operation, and it's well over $2,000...

And you would need 2 of them as they are single channel type of affairs to eliminate the mechanical diff... And a very beefy contactor to do reverse.

That all being said there use to be another controller listed on robotcombat.com that was intended for electric forklifts that people used with the Etek motors... They don't list it anymore and I can't recall the name off the top of my head... But I can find it if need be... Not a good solution though as it required either $300 software or $300 hardware programing interface to set any parameters. And was about the same base price as Roboteqs controllers... And the specs seemed pretty similar as far a capacity...

Hope this helps. Bottom line... Building a vehicle is a lot trickery then it seems :-)

"That is to say that if you were to keep the motor shaft immobile and current draw would rapidly shoot up to a level that would likely smoke even the HE versions of Roboteq's controllers. And indeed nearly every other controller on the market."

Please explain this scenario further... is this in an idle state, when the vehicle is not moving and the throttle is at ZERO?

Very new to this and very thankful for your input.

I've probably phrased what I said wrong... Every time I try to really understand how all this stuff relates I end up with a headache :-)

But the issue that I think cosma was warning you about is an scenario like this:

Your going fast in your buggy, lets say for the sake of example that the throttle is at or near 100% the motor is drawing a good amount of current, but nothing your controller can't handle.
One or both of the rear wheels gets stuck. As long as the throttle is still wide open the motor will draw more and more current out of your battery's and through the controller. This spike of current can go very high very quickly, faster then your controller or even a fuse can catch it and your controller catches fire and or melts.

As far as "stall torque" goes I've always interpreted this as the maximum amount of amps a given motor could draw. It's a figure you plug into formulas to get other figures that let you model your motor and by extension the rest of your vehicle. This all lets you figure out how it's going to perform on paper... If your into that sort of thing :-)

I started out getting the fastest motor I could afford (a magmotor s28 400) and worked the rest of it out by building and testing.

Well... Obviously I'm still not cranking out equations :-) And I'm still testing by the seat of my pants. But the AX2550 is my _3rd_ speed controller. And I'm still saving for some better batteries... I'm having a hell of a time, it's lots of fun, but if I knew how much money I was going to end up spending I never would have started... And that's why I hate doing the math :-)

Check out both of the mailing lists below... Both are good in there own way. Both can have upwards of 100 messages a day so set up a filter.
www.evdl.org/
evtech.org/

P.S. I'm making an electric skateboard. No brakes, no steering wheel... It should have been done by now :-)

I fully understand now what you were trying to convey across.

However the next question is.... isn't that the purpose of current limiting, in order to prevent this from happening?

Shouldn't a controller be able to instantly detect the amount of current that is passing through it and protect itself?

read some more on this topic....

Perhaps even something simple like a fuse inline with a motor would suffice?

Current limiting works quite well in normal operating conditions. In a stall, current will be extremely high. With an Etek motor, that current is akin to a short circuit current. Current limiting, which works by throttling back the power when current exceeds a preset limit, will not perform well in such a situation.

The latest revision of the Roboteq controllers now feature short circuit protection and stall protection.

Short circuit protection will detect when instantaneous amps in the MOSFETs are of several hundred amps and cut the drive within microseconds.

Stall protection looks for the very high motor current that can be flowing through the motor at low PWM cycle, even though the battery current will be much lower at that time.

These protections are not bulletproof and will work when short/stall are exceptional conditions. With the Etek on a sand buggy, stall and near-shorts are likely to be very common.

As far as using a fuse in line with the motor, this could work except that 100+A fuses are not common and not very quick to blow.

Just in case ya' got the wrong idea... I have not managed to do anything bad to my AX2550 yet :-)

A dinky little IFR Robotics controller (This thing shipped in an envelope! I knew I shouldn't use at as soon as I put my hands on the thing... But I did anyway :-) It was hooked up with a little 60mm fan on the heat sinks that was drawing air over and out of the enclosure it was mounted in. The first time I popped the little electrolytic cap on the thing and this plume of dense white smoke starts spewing out the back as I come coasting to a stop. That was just from going back and forth to much... I replaced the cap and then blew a mossfet going up a grade the next day.

My next victim was an OSMC... This bad boy comes ratted very generously but is just the H bridge... The "power amplifier" you need a separate board to supply a variable voltage in the form of a PWM stream and some other logic for forward / reverse etc. I tried to design a brain myself but somehow I fucked up the power supply... Though to this day I'm not really sure how... But after about 30 seconds there was a high pitched whine and then the FET's and the controller chip started burning.

My AX2550 has been through much more extensive use and has yet to disappoint me in any way. But this stuff is trickyer then I thought... Just making a simple vehicle has given me a whole world of new respect for a conventional car...

Follow up on your idea... But just remember that as much as bigger is always better in my book too, the engineers job quickly becomes one of balancing trade-offs. When I build the next power board it will have a smaller motor so I can get more run time off the same amount of battery's.

-PM