Drill Motor Adjustments -Rev vs Forward

Hi, my team has run into a quite a problem. We have our gearboxes mounted with the bosch drill motors running in opposite directions(really no way to fix this).

Our problem truly is that the bosch motors are stronger in forward gear than in reverse! It is a topped off value, ie, they run fine up until the power sent to them reaches a pwm value, and then the right one just stops going faster, while the left one does.

We would rather not limit the power to one of the motors, thus killing our top speed, but we would prefer not to limit the speed to half(or so it seems).

Anyone have any ideas how to get around this? is there a setting perhaps on the acutal motors as to which direction is forward, and which is reverse?

Thanks all,
Team 219

This issue has been debated many times on this website. There is no reversing setting. The motors are wound in order to provide maximum power in one direction (the forward direction on the drill is where the power is needed).

Here are the various solutions:

  1. Change your mechanical design so that both motors run to same direction to propel your robot.

The positive outcome of this effort would be that the 'bot would run straight.
The negative is that it takes some effort (more complexity, weight, efficiency loss, etc.) to make it work.

  1. Create software which limits the forward-rotating motor in order to match the speed of the motor moving in reverse.

  2. Deal with it and let your drivers just get familiar with driving in slight arcs.

If you can make an extremely simple and elegant solution to make a mechanical design as said in #1 above, then do that. But, whenever you add mechanical complexity you are injecting chances for breakdowns. If this mechanism is complex, then scrap it and go with #2 or #3. Many teams do well with “dealing with it” and letting the drivers simply get familiar with the way the robot drives.

Andy B.

This is commomly a problem of many teams… I would recomend making a custom mount… or try running a sproket from the opposite direction on you reversed tire… I think FIRST had something on this last year… you might be able to find it in the FIRST archives.

The drivers driving in curves is easy to do… just make sure he/she gets a bunch of practice.

If I find some FIRST drawings on it in some of my old documents… if I do I will e-mail it to you.

Good Luck,

-Greg The Great

I’d agree with both of the previous posts. However, I think that the simpliest solution that you’re looking for (since you don’t wish to reduce power in code, and want your robot to drive straight as possible) lies in basically an extra pair of gears somewhere among your reductions to effectively reverse the direction of one motor. While you might loose a few percent efficency… it’s an easy thing to do.

I’m incredibly temped to CAD out a simple gearbox that’d do just this that would be “out of the box compatiable” with last year’s kit. If I didn’t have another couple hours of homework… I’d be crazy enough to do it.

So… I go ASCII style G’s are gears (5 G’s to a gear), +'s are bearings. Make the housing out of some 1/4 plate, you could get away with 1/2 shaft or less on those gears, and then just couple them to wherever you’re outputing to.

  ---
  |D| 
  |M|
  ---
   |
---+----+---
|  |	|  |
|GGGGGGGGGG| 
|  |	|  |
---+----+---
	|
	|

Anyone feel like adding to my reputation for my artistic skills? Haha, just kidding. Partially.

Matt

I just talked to some of our engineers, and we found that the way we have our gearboxes set up they are directionally universal. So, the repair we will prolly use is to just flip the right one around 180 degrees. Thank you all for your help.

  1. Another way to deal with this is to put a speed measurement on each drill motor (say with an encoder). You can then run the drill motors in closed loop mode.

In 2003, this was just a barely possible solution, especially if you used encoders in your preprogrammed mode. In 2004, with the more versatile controller and tight control over the timing, this may be the best way to accomplish this task.

BTW, can anyone actually explain how they have wound the motors so that they have different power in forward versus reverse? Is there an assymmetry in the coils or in the magnets? I agree that there is an assymmetry. I just don’t see how it is done or how it would be cheaper.

Dredging up very old memories from my school days, I think the reason for difference in fwd vs. rev speed is due to optimization of the angular position of the commutator brushes vs the permanent magnet (stator) field, designed to maximize torque. Think of it as being the DC motor equivalent of cam timing in internal combustion engines.

You can build a permanent magnet DC motor with equal torque in both directions, by “centering” the brushes, but you sacrifice peak torque in doing so. Which would suggest that if you wanted to modify the drill motors’ brush holders, you could adjust for equal speed/voltage in both directions.

Dredging up very old memories from my school days, I think the reason for difference in fwd vs. rev speed is due to optimization of the angular position of the commutator brushes vs the permanent magnet (stator) field, designed to maximize torque. Think of it as being the DC motor equivalent of cam timing in internal combustion engines.

Yup the drill motors are timed in one direction. They are either timed in counter clock wise or clock wise. I believe they are timed in the direction that drills are most commonly used in (forward not reverse). Unfortunately I have no idea what direction that is.

when i look back at last years 312 bot, i’m quite confident that they are timed counter clockwise

Counterclockwise when you are looking at the end of the motor, or looking down the motor barrel? Without specifing that… it could be either way.

Since screws drill in clockwise looking down the “barrel” of a drill, I would imagine that’s where they’re optimized. I’d probably bet my life savings on it. (Which is admittedly negative, since I’m in college.)

Matt

the best way to solve this is to use the yaw rate sensor to close the loop on steering

then when you driver is telling the bot to go straight, the yaw sensor will detect any turning, and adjust the power to the motors automatically

in addition, your bot will drive SO MUCH better with the steering running closed loop - it will act like a servo, instead of a wobbly goblin

if you dont know how to do this, the easiest way is to subtract the yaw rate signal from the joystick X axis signal (assuming your using one stick steering)

this will be the difference between what the driver is telling the bot to do, and what its actually doing.

multiply this error signal by 2 (as a starting point) - normalize it for 127=straight

and use that as your steering command instead of the joystick X command.

If the steering is sluggish, try a gain (multiplyer) of 3

if the steering is jittery, or if it oscillates back and forth, back off on the gain, try 1.5 or 1.

feedback is one of THE MOST powerfull tools engineers have at their disposal!

so… even if i was “wrong”, i’d still be right…:stuck_out_tongue:

I would say looking down the motor barrel (the output shaft pointing toward you).

i could be wrong though; i’m basing this all on the assumption that we ran our drill motors forward last year.

I just noticed a couple things

the transmissions that FIRST gave us last year - the motors spin in the same direction if you have one on each side of you bot, so the FWD > REV problem doenst exists if you use them (we have them on a practice bot we build over the last few weeks)

the other thing - make sure you have calibrated your speed controls - you have to push and hold the little button with a bent paperclip, then move the jopystick full FWD, full REV then back to center, then let the button go - do both sides.

This has been an interesting thread but the proceedure desribed below is completely new to me. Can you provide me with just a few more details… When you mention speed controls, are you talking about the Victor 88x speed controllers? I’m at home and don’t have one lying around so just where is this little button?

the other thing - make sure you have calibrated your speed controls - you have to push and hold the little button with a bent paperclip, then move the jopystick full FWD, full REV then back to center, then let the button go - do both sides. :cool:

The button looks like a little hole next to the LED on the Victor speed control. Thats why you need a paperclip, to push and hold the button thats inside the hole.

The brush position has been beaten to death so far (all of it true)but there is something else that is a factor and that is the brush/commutator switching frequency vs. the speed controller switching frequency. This varies for motor types as the design of the commutator is optimized during the motor design phase. At some point the (high RPM) the two frequencies interfere with each other and the motor efficiency drops off. Also keep in mind that the difference in speed is greater at higher RPM. If you design the drive train and running speed to keep the motor away from running full RPM, you will minimize the effect. There is no substitute for practice, lot’s of practice!