Travel speed inconsistency with LV-168 and Tamiya gears/moto

Support A-Star and Orangutan
1

I have bought and used the LV-168 controller to control 2 motors on a robot/vehicle I have made which so far is just traveling a set distance and stopping. I am using two Tamiya Universal Gearbox kits with two Tamiya Narrow Tire sets. The program simply sets the motor speeds to -255,-255, has a delay for 10 seconds, and then sets motor speeds to 0. One gearbox is at the front of the vehicle, the other is at the rear.

However, upon testing the distance traveled after 10 seconds of running, I am encountering quite a large deviation in the measured distance. The left/right movement is relatively small (gearbox alignment issues) but I am getting up to 15cm deviation in a 360cm trial and 18cm in a 420/500cm trial, and it does not appear to directly follow the discharge of the battery i.e. it is rather random, although over long periods of time the traveled distance generally gets shorter. Is there a reason for the deviation? When I was using a BOEBot controller (with separate power for the microcontroller and motors, and a different Tamiya gearbox (4-speed crank axle)) there was a maximum of 1-2cm deviation on aImost a 700cm run. The old setup however used a relay to control the motors as the BOEBot could not output sufficient current.

I have used 2 0.1uF capacitors on each motor to reduce noise, however, while the motors are running, the controller sometimes does not respond to requests to turn off. I also tried running the motors at speed 200, which still had a large deviation from trial to trial. i.e. (255 speed) 487.7->490.5->497.7 then on 200 speed 341.1, 331.0, 345.1, 338, 338.9

I have since ordered 3 Tamiya 4-speed crank axles, as the metal gearboxes appear to shake a lot on running (but despite that they still go remarkably straight). Does anyone have experience with using Tamiya gearboxes with the LV-168? Is the metal gearbox the least consistent in comparison to the others? A problem I experienced with the 4-speed crank axles was that they tended to exhibit ‘wobble’, which would make the robot travel in a circular/parabolic motion. Also, does decreasing the ‘motorSpeed’ decrease the voltage output to the motors? I would rather run the motors at closer to their 3V spec than at 4.5V with 3 batteries. I am also considering converting the vehicle so only one motor would power it and the other would just be a ‘blank’ axle, but I am having a hard time finding a Tamiya kit for a ‘blank axle’ that keeps the axle in place (restricting left/right movement).

2

Hello,

I have not gone and characterized a bunch of gearboxes, but less that 5% variation is well within the range you should expect. The result you got on the BOEBot is rather surprising and probably anomalous.

As for the speed/voltage control, the motors will get almost the full battery voltage regardless of speed setting. We recommend that you use NiMH cells, which would give you around 3.6V.

For your “blank axle” requirement, you could use a low gear ratio gearbox and remove the motor. It’s a bit inefficient, but it might be a quick solution if you already have the gearbox in place.

- Jan

3

I am trying to make a vehicle that does travel a very consistent distance/speed based on a programmed time though. The previous BOEBot based design was sort of anamalous, it stopped working correctly due to electrical noise issues (the controller would not shut off the motors after the given time), but the noise issue could be rectified by using 4 batteries for the BOEBot and 4 separate ones for the motors, and a 10Kohm resistor in the row of the breadboard where the battery + end connected to the motor + end. Under those conditions, the runs were very consistent and only a small negative trend due to discharge was observed.

What would be a better option then? Is there another gearbox/servo control I could use with the LV-168 (and preferably in combination with my current Tamiya axle/tires) that would provide a measured # of revolutions and thus be more accurate? I was thinking about setting up a revolution counter on the axle that would return the information to the BOEBot but I can’t find a good setup that would work with the Tamiya axle and the LV-168 (I have seen custom ones for Lego kits though). I am not allowed to use electronic sighting devices or lasers to determine the stopping point.

4

I don’t have a good alternative to offer. In general, you’ll need some feedback to get good repeatability. Without that, you should look for a precision mechanism (which the Tamiya educational gearboxes are not) and a high gear ratio (so your output speed is more directly a function of your input voltage).

- Jan