Difference in forward/reverse drill motor speed
 

Posted by Patrick Dingle at 2/5/2001 11:38 PM EST


Other on team #639, Red B^2, from Ithaca High School and Cornell University.



Has anyone noticed that the drill motors have a different angular velocity when in forward than in reverse? If so, how have teams overcome this problem if they use the motors for the drive system?

Patrick

Re: Difference in forward/reverse drill motor speed
 

Posted by Matt Leese at 2/6/2001 12:10 AM EST


Other on team #73, Tigerbolt, from Edison Technical HS and Alstom & Rochester Institute of Technology.


In Reply to: Difference in forward/reverse drill motor speed
Posted by Patrick Dingle on 2/5/2001 11:38 PM EST:



Yep, noticed this one awhile ago. You can work around it several ways. The more correct solution (i.e. the mechanical one) is to either mount the drill motors turning the same direction or put an extra gear in one side so they both turn the same direction. The less correct solution (i.e. the programming one -- Matt's Law: never fix something in programming if it can be done mechanically) is to wire a potentiameter onto the control system that will scale down one of the drive PWM outputs. You then drive the robot over a distance and fiddle with the pot until you get the robot driving in a straight line. This does lower the overall top speed of the robot however (you have to scale down the motor turning the forward direction). This takes a little bit of finesse to get right and careful drivers who don't bump the pot. ;) It worked fairly well for us last year but there are a few caveats that I'll let you figure out on your own (hint: it has to do with driving in reverse).

Matt

Re: Difference in forward/reverse drill motor speed
 

Posted by Anton Abaya at 2/6/2001 12:15 AM EST


Coach on team #419, Rambots, from UMass Boston / BC High and NONE AT THE MOMENT! :(.


In Reply to: Difference in forward/reverse drill motor speed
Posted by Patrick Dingle on 2/5/2001 11:38 PM EST:



: Has anyone noticed that the drill motors have a different angular velocity when in forward than in reverse? If so, how have teams overcome this problem if they use the motors for the drive system?

: Patrick

yes the drill motors are not in sync when they are running on opposite directions. this is because one motor is running in forward and the other one is going in reverse.

apparently there are two solutions, one is mechanical and the other is software. the mechanical solution is whats called a "right angle drive" or so they call it. basically what this does is it makes both run at the same direction therefore making the drive more efficient and not having to make the driver compensate for the drag occurring on one side.

check this link out. http://sharingfirst.mit.edu/users/os...n/FRC2000page/

the software solution, is to sync the right to the left via ...software. this will take trial and error as it would vary from robot to robot. but the basic idea is "if right joystick is X, then left should be x+y (y=2-10?)..." with this, it's the left joystick (motor) that drags.

i hope i answered ur question.

-anton
"I think I did."

don't bother
 

Posted by Joe Ross at 2/6/2001 3:28 AM EST


Engineer on team #330, Beach Bot, from Hope Chapel Academy and NASA/JPL , J&F Machine, and Raytheon.


In Reply to: Difference in forward/reverse drill motor speed
Posted by Patrick Dingle on 2/5/2001 11:38 PM EST:



: Has anyone noticed that the drill motors have a different angular velocity when in forward than in reverse? If so, how have teams overcome this problem if they use the motors for the drive system?

We have always used a tank drive system with one joystick for each side. I don't think the drivers have even noticed that they need to slightly compensate. Even if they do, it should become a habit and never be a problem.

Using a mechanical system to compensate may very well be a worse idea than leaving it alone. Even if it means that you slightly increase the efficiency in one direction, you very well may lose that advantage when you consider the complexity of the mechanical system, as well as the extra friction that it introduces.

Using software to compensate is almost as bad. You lose some small range of motion that may come in very usefull later on in competition.

Re: don't bother
 

Posted by Jim Meyer at 2/6/2001 9:05 AM EST


Engineer on team #67, HOT Team, from Huron Valley Schools and GM Milford Proving Ground.


In Reply to: don't bother
Posted by Joe Ross on 2/6/2001 3:28 AM EST:



: Using a mechanical system to compensate may very well be a worse idea than leaving it alone. Even if it means that you slightly increase the efficiency in one direction, you very well may lose that advantage when you consider the complexity of the mechanical system, as well as the extra friction that it introduces.

Mechanical solutions do not have to be complicated or less efficient. Right angle drivetrains with bevel gears are not the only solution either. For the last 2 years the HOT team has always spun our motors the same direction and will continue to do so. This year we are using a single stage gear reduction from the drill motors using only spur gears.

It's probably a little late this year for this type of change but stop by our pits and check out how easily this can be done. I would explain it but I don't feel I can do an adequate job without a picture. I'll work on getting a picture to post.

Re: don't bother
 

Posted by Matt Leese at 2/6/2001 9:35 AM EST


Other on team #73, Tigerbolt, from Edison Technical HS and Alstom & Rochester Institute of Technology.


In Reply to: don't bother
Posted by Joe Ross on 2/6/2001 3:28 AM EST:



I'm not sure how you had your robot setup last year but ours running tank drive with two powered wheel chair wheels in the back and two swivel casters in the front would track to the left rather badly. Our solution was the software one rather than a mechanical one because we discovered it too later. Our solution wasn't perfect and it had to be recalibrated rather regularly (you might've seen me in FL trying to find a long space to drive the robot...) but it did work.

Matt

larger fish to fry...
 

Posted by Joe Johnson at 2/6/2001 1:16 PM EST


Engineer on team #47, Chief Delphi, from Pontiac Central High School and Delphi Automotive Systems.


In Reply to: don't bother
Posted by Joe Ross on 2/6/2001 3:28 AM EST:



I am still convinced that this is a small problem that can be easily compensated for by the drivers when they have the time to practice.

Focus on getting your drivers more time behind the wheel and you will be farther ahead in the long run.

Just one man's opinion.

Joe J.

Re: Make 'em go the same direction!
 

Posted by Mike Gray at 2/7/2001 11:55 AM EST


Engineer on team HOT from Huron Valley Schools sponsored by GM Milford Proving Ground.


In Reply to: don't bother
Posted by Joe Ross on 2/6/2001 3:28 AM EST:




We discovered the speed difference in the drill motors in 1998 while trying to determine why the machine pulled to the left. After exhausting all the software/joystick/speed controller/mechanical possibilities, only the motors remained. We measured the forward and reverse speeds with a strobotach.

We found that that the speed difference is a huge 15-20% depending upon the applied voltage. Motor-to-motor variation is a low 2 percent.

With the drill motors rotating in the same direction, anyone can handle the machine well with a few minutes of practice. When the machine performs as expected, driver confidence soars.

The speed difference is caused by the motor, not the gearset.

After much thought, we found a pretty easy way to rotate them in the same direction. If my team members don't object, I'll post an AutoCAD rendering or photo.

Beating a dead horse...
 

Posted by Joe Johnson at 2/7/2001 9:59 PM EST


Engineer on team #47, Chief Delphi, from Pontiac Central High School and Delphi Automotive Systems.


In Reply to: Re: Make 'em go the same direction!
Posted by Mike Gray on 2/7/2001 11:55 AM EST:



We have NEVER seen such a difference as 15-20%

I know that many folks have very strong ideas about
this but I am serious when I say that we have never
done a thing to try to compensate for this reported
difference.

Just as another point of data, I called my contact at
Johnson Electric (I often use them in my day job). He
was able to look up the print of the motor for me (he
is sending me a pdf file of it soon). He tells me that
the motor should have the same performance in CW and
CCW directions. Yes, some motors do have differences,
but not this motor.

Now I am between a rock and a hard place because I have
conflicting reports from usually reliable sources
(Johnson has ZERO incentive to mislead me,
Raul-the-Magnificent is almost NEVER wrong, and who
argues with ANYONE from the Heros-Of-Tomorrow team?).

Believe it or not, I am thinking of having our lab do a
gov't job for me and test the motor on our dyno.

BOTTOM LINE:

I don't know what to believe. But again, my advice is
to do nothing, focus on getting your drivers more time
behind the joystick(s).

Joe J.

Re: Might be sleeping...
 

Posted by Mike Gray at 2/8/2001 8:02 AM EST


Engineer on team HOT from Huron Valley Schools sponsored by GM Milford Proving Ground.


In Reply to: Beating a dead horse...
Posted by Joe Johnson on 2/7/2001 9:59 PM EST:



I know its hard to believe, and I can't explain why they do it, but turning them the same direction dramatically improves performance, especially when launching from a dead stop.

I would love to see some data from a real lab! Think you can do this? We got some EMI data on these motors also. Both radiated and conducted emissions are very high, but that is expected.

The drill motors are the only motors exibiting this odd behavior. The Delphi motors are especially quiet, both audibly and electrically. They rotate symetrically too!

I am wrong.
 

Posted by Joe Johnson at 2/8/2001 2:39 PM EST


Engineer on team #47, Chief Delphi, from Pontiac Central High School and Delphi Automotive Systems.


In Reply to: Re: Might be sleeping...
Posted by Mike Gray on 2/8/2001 8:02 AM EST:



I just got a frantic message from my Johnson Electric contact. It seems that he was wrong yesterday when he told me the motor was not biased.

When he confirmed with his manufacturing sources in Hong Kong, they told him that the magnets are oreinted 15 degrees with respect to the brushes, making the motor performance asymmetrical.

So... I was wrong. Sorry for misleading folks.

But... I am still of the opinion that this is not such a big deal. Again, I believe driver skill can compensate very effectively.


Yet... I have to confess that because the design on Chief Delphi 6 is such that we can change the motor orientation without changing any parts, we will probably have them installed to run in the same direction.

Bottom line:

If it is easy, I say do it. If it is hard, I say don't bother.

Joe J.


P.S. I KNEW that I should never take sides against The Heros of Tomorrow (and especially Mike Gray & Jim Meyer) and/or Raul-the-Magnificent -- ah well, live and learn.

P.P.S. I am going to try to get Johnson to give me the CCW and CW motor curves. I am very curious to see the differences on paper.

Re: I am wrong.
 

Posted by Mike Gray at 2/8/2001 3:07 PM EST


Engineer on team HOT from Huron Valley Schools sponsored by GM Milford Proving Ground.


In Reply to: I am wrong.
Posted by Joe Johnson on 2/8/2001 2:39 PM EST:




If its easy to alter the rotational direction of the drill motors in your machine, you might try a quick experiment:

Evaluate the quality of the launch with the motors running in opposite directions. Turn one around and do the same thing. I'm convinced that you will want them to run the same direction.

Please let us know how the operators respond to the A/B test.

We Compensated Last Year
 

Posted by Bill Beatty at 2/9/2001 11:05 AM EST


Other on team #71, Team Hammond, from Team Hammond.


In Reply to: Re: I am wrong.
Posted by Mike Gray on 2/8/2001 3:07 PM EST:



The difference for us last year was as high as 15 percent and for the first time we compensated with software. When we changed motors a quick check with the tach and we would change the percentage factor. Not a biggie, but it seemed to help the drivers somewhat.

Mr Bill

we did in '98, but not '99-'01 (EOM)
 

Posted by Ken Patton at 2/9/2001 8:06 PM EST


Engineer on team #65, The Huskie Brigade, from Pontiac Northern High School and GM Powertrain.


In Reply to: We Compensated Last Year
Posted by Bill Beatty on 2/9/2001 11:05 AM EST:



sdkldfjkdf

Re: I am wrong.
 

Posted by Alan Ing at 2/9/2001 9:31 PM EST


Engineer on team #368, Kika Mana, from McKinley High School.


In Reply to: I am wrong.
Posted by Joe Johnson on 2/8/2001 2:39 PM EST:



: I just got a frantic message from my Johnson Electric contact. It seems that he was wrong yesterday when he told me the motor was not biased.

: When he confirmed with his manufacturing sources in Hong Kong, they told him that the magnets are oreinted 15 degrees with respect to the brushes, making the motor performance asymmetrical.

: So... I was wrong. Sorry for misleading folks.

: But... I am still of the opinion that this is not such a big deal. Again, I believe driver skill can compensate very effectively.

:
: Yet... I have to confess that because the design on Chief Delphi 6 is such that we can change the motor orientation without changing any parts, we will probably have them installed to run in the same direction.

: Bottom line:

: If it is easy, I say do it. If it is hard, I say don't bother.

: Joe J.

:
: P.S. I KNEW that I should never take sides against The Heros of Tomorrow (and especially Mike Gray & Jim Meyer) and/or Raul-the-Magnificent -- ah well, live and learn.

: P.P.S. I am going to try to get Johnson to give me the CCW and CW motor curves. I am very curious to see the differences on paper.

Joe, thanks for the last minute motor seminar at the kickoff, it was really informative.

In regards to that drill motor. My guess is that the motor curves are going to be significantly different. Our robot uses one motor in the forward direction and one in reverse. We are pretty sure that friction is not an issue. Our robot definitely turns right faster than left. We put a meter on the motor and made some continuous turns in both directions. When turning one way, the motors are drawing 10 amps (both motors running forward), but in the other direction (both motors running backwards), it is drawing about 18 amps. This is the case for either motor. Seems pretty significant. Some robots might pop their relays while turning in one direction and controlling the stretcher but not in the other direction. Anyway, I would recommend measuring the current through both motors while turning to make sure your gearing is appropriate for both directions.

Re: Difference in forward/reverse drill motor speed
 

Posted by Al Skierkiewicz at 2/6/2001 1:46 PM EST


Engineer on team #111, Wildstang, from Wheeling High & Rolling Meadows High and Motorola.


In Reply to: Difference in forward/reverse drill motor speed
Posted by Patrick Dingle on 2/5/2001 11:38 PM EST:



I am having a tough time with this one. I have not heard that there is a difference between forward and reverse. (although I'm old and miss a lot.) The motors are electrically identical in either direction, the transmission is the same, the output shaft is the same and the speed controller is the same. That leaves only two places outside of drive train for the anamoly to occur. 1. The joystick pot is not symetrical (but this should be corrected in the calibration of the speed controller) or 2. the full data is not being sent to the controller (software correctable.) We had a problem a few years ago until we realized that the controller calibrates itself (stores the full swing data points) only when you release the button. I will have to check with Raul on more info on this.

Re: Difference in forward/reverse drill motor speed
 

Posted by Matt Leese at 2/6/2001 5:54 PM EST


Other on team #73, Tigerbolt, from Edison Technical HS and Alstom & Rochester Institute of Technology.


In Reply to: Re: Difference in forward/reverse drill motor speed
Posted by Al Skierkiewicz on 2/6/2001 1:46 PM EST:



The whole reason this exists is because of the brush formation. The brushes in the window and seat motors are symmetrical in both directions which means they will go the same speed forward and back. However, in things such as the drill motors and fisher price motors, the brushes are aligned to be more efficient in one direction. I'm not sure why they do this, whether it increases power or not or it's just cheaper, this is the way they are designed. So it's not something that can be worked around.

Matt

Is this true?
 

Posted by Joe Johnson at 2/6/2001 9:07 PM EST


Engineer on team #47, Chief Delphi, from Pontiac Central High School and Delphi Automotive Systems.


In Reply to: Re: Difference in forward/reverse drill motor speed
Posted by Matt Leese on 2/6/2001 5:54 PM EST:



I have looked at these motors for years now and I have
never noticed that the brushes are actually not 180
degrees apart from each other on the Fisher Price and
Drill motors.

Will someone please confirm this for me?

Joe J.

Re: Is this true?
 

Posted by Matt Leese at 2/6/2001 9:29 PM EST


Other on team #73, Tigerbolt, from Edison Technical HS and Alstom & Rochester Institute of Technology.


In Reply to: Is this true?
Posted by Joe Johnson on 2/6/2001 9:07 PM EST:



In all honesty I have no idea where I got that bit of information from but I'm fairly certain I picked it up somewhere last year and I seem to remember getting it from a reliable source. So I may just be spouting off random stuff.

Matt

Re: Is this true?
 

Posted by PAUL GIANNOSA at 2/6/2001 10:10 PM EST


Engineer on team #27, TEAM RUSH, from OSMTECH and TEXTRON AUTOMOTIVE CO..


In Reply to: Re: Is this true?
Posted by Matt Leese on 2/6/2001 9:29 PM EST:



: In all honesty I have no idea where I got that bit of information from but I'm fairly certain I picked it up somewhere last year and I seem to remember getting it from a reliable source. So I may just be spouting off random stuff.

: Matt
HEY. LAST YEAR, SOMEONE FROM TEAM 33, WHO I WORKED WITH AT WORK, TOLD ME THE MOTORS PRODUCE MORE TORQUE FORWARD THAN IN REVERSE. I TOLD HIM HE HAD BEEN SMOKIN' SOME FUNNY STUFF AND HE HAD MORE THAN A FEW SCREWS LOOSE. WELL, TO MY SUPRISE, WHEN WE WERE AT THE GREAT LAKES REGIONAL, THE BOT PULLED TO THE RIGHT. GUESS WHICH SIDE HAD THE MOTOR RUNNING IN R(FOR RACE? I SMILED, WITH FEATHERS IN MOUTH, AND TRUDGED ON FOR A REASON WHY I NOW HAD EGG ON MY OAKLEYS. TURNS OUT THAT THE WINDINGS IN THE ARMATURE ARE WOUND A PARTICULAR WAY TO PROVIDE MORE TORQUE IN FORWARD THAN IN REVERSE. TAKES MORE GRUNT TO DRIVE A SCREW THAN DOES TO PULL IT FROM ITS HOLE. BEST WAY TO OVER COME THIS"PROBLEM" IS TO RUNN BOTH MOTORS IN SAME DIRECTION. JUST PACKAGE THEM CREATIVLY. PMGRACER

Re: Is this true?
 

Posted by Michael Betts at 2/7/2001 7:20 AM EST


Engineer on team #177, Bobcat Robotics, from South Windsor High School and International Fuel Cells.


In Reply to: Is this true?
Posted by Joe Johnson on 2/6/2001 9:07 PM EST:



There is no doubt that there is a small difference between forward and reverse on the drill motors. This was also true of the Milwaukee motors we used years ago.

However, I agree with Joe's earlier post. Team 177 has never done anything, mechanically or electrically, to try and "fix" this difference. The drivers learn to compensate almost immediately as soon as we get them practicing.

Build it and drive!

Mike

Re: Is this true?
 

Posted by Brian Cholerton at 2/7/2001 12:12 PM EST


Engineer on team #311, Red Jammers, from East Islip High School and KeySpan Energy, Multiline Corp., Computer Assoc..


In Reply to: Is this true?
Posted by Joe Johnson on 2/6/2001 9:07 PM EST:



The brushes are at 180 degrees from each other. The alignment has to do with the timing of the brushes and the commutator in relation to the magnets. Electric motors are usually set to have the timing advanced by a few degrees so that the motor can make more power (torque) more efficiently in one particular direction. If a motor were set up to be the same in both directions it would definitely affect the torque curve. Similar to an internal combustion having it's timing before top dead center but obviously not for the reason of burning fuel.

Best of luck!

Re: Difference in forward/reverse drill motor speed
 

Posted by Al Skierkiewicz at 2/6/2001 1:46 PM EST


Engineer on team #111, Wildstang, from Wheeling High & Rolling Meadows High and Motorola.


In Reply to: Difference in forward/reverse drill motor speed
Posted by Patrick Dingle on 2/5/2001 11:38 PM EST:



I am having a tough time with this one. I have not heard that there is a difference between forward and reverse. (although I'm old and miss a lot.) The motors are electrically identical in either direction, the transmission is the same, the output shaft is the same and the speed controller is the same. That leaves only two places outside of drive train for the anamoly to occur. 1. The joystick pot is not symetrical (but this should be corrected in the calibration of the speed controller) or 2. the full data is not being sent to the controller (software correctable.) We had a problem a few years ago until we realized that the controller calibrates itself (stores the full swing data points) only when you release the button. I will have to check with Raul on more info on this.

Re: Difference in forward/reverse drill motor speed
 

Posted by Patrick Dingle at 2/6/2001 3:04 PM EST


Other on team #639, Red B^2, from Ithaca High School and Cornell University.


In Reply to: Re: Difference in forward/reverse drill motor speed
Posted by Al Skierkiewicz on 2/6/2001 1:46 PM EST:



It has to be something with either the motor or the gearbox. When I noticed it, I was hooking up the battery directly to the motor. You can hear the difference (and measure it with proper equipment) in speeds from forward and reverse.

Patrick

: I am having a tough time with this one. I have not heard that there is a difference between forward and reverse. (although I'm old and miss a lot.) The motors are electrically identical in either direction, the transmission is the same, the output shaft is the same and the speed controller is the same. That leaves only two places outside of drive train for the anamoly to occur. 1. The joystick pot is not symetrical (but this should be corrected in the calibration of the speed controller) or 2. the full data is not being sent to the controller (software correctable.) We had a problem a few years ago until we realized that the controller calibrates itself (stores the full swing data points) only when you release the button. I will have to check with Raul on more info on this.

Yes, there is a difference.
 

Posted by Al Skierkiewicz at 2/7/2001 9:11 AM EST


Engineer on team #111, Wildstang, from Wheeling High & Rolling Meadows High and Motorola.


In Reply to: Re: Difference in forward/reverse drill motor speed
Posted by Al Skierkiewicz on 2/6/2001 1:46 PM EST:



I have checked with Raul and Jim B and they HAVE experienced a variation in forward/reverse speed. In past years a thorough investigation showed as much as 10% variation in speed from unit to unit and forward/reverse. As Joe has suggested, (we agree) getting drivers to invest time behind the wheel is the best way to overcome the problem. I am still having a problem in analyzing why this occurs. Production variations can explain unit to unit changes but not for/rev. Any other input?
Al

Re: Yes, there is a difference.
 

Posted by Dave... at 2/8/2001 9:03 PM EST


Engineer on team #5, Robocards, from Melvindale, MI and AFL (Alcoa Fujikura Ltd).


In Reply to: Yes, there is a difference.
Posted by Al Skierkiewicz on 2/7/2001 9:11 AM EST:



DC motors operate by inducing a magnetic field created by the armature within the magnetic field of the permanent magnets in the housing. As any elementary science teacher can attest to, a wire wrapped around a nail with DC current running through it will create a directional magnetic field, and by changing the polarity of the wires, the magnetic field will change 180° (N becomes S and vice versa).

The armature is wrapped with magnet wire from the inside out, and so running current in opposite directions will not only change the direction of the magnetic field from the armature, but will interact differently with the magnetic field from the permanent magnets, which does not get to change its directional field.

I imagine that a DC motor which does not use permanent magnets but rather windings in the housing to create the second magnetic field is not as susceptible to power loss in forward vs reverse. Both fields would change 180° any time the polarity was reversed, keeping roughly the same power in either direction.

We simply limit the power to the stronger drive motor by a trial and error method in the programming. If the robot steers to the right, we decrease the maximum power available to the left drive motor (e.g. p1_y = p1_y MAX 228) until the thing goes straight. Crude, but effective and easy to change.

My two cents worth.

: I have checked with Raul and Jim B and they HAVE experienced a variation in forward/reverse speed. In past years a thorough investigation showed as much as 10% variation in speed from unit to unit and forward/reverse. As Joe has suggested, (we agree) getting drivers to invest time behind the wheel is the best way to overcome the problem. I am still having a problem in analyzing why this occurs. Production variations can explain unit to unit changes but not for/rev. Any other input?
: Al