Subject:
|
Re: Adder/Subtractor revisited
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Sun, 7 Feb 1999 21:57:15 GMT
|
Reply-To:
|
bryan.beatty@&StopSpammers&autodesk.com
|
Viewed:
|
1180 times
|
| |
|
|
David Harry wrote:
>
> > Anyway, the challenge is to build something that can run the straight
> > sections very fast, while optically following the tape. The problem I
> > encounter when trying to build a really high-speed robot is that there's
> > no good way to control the speed of the motors; they're either on, or
> > off. etc..
> > Unfortunately, if the robot is geared for high running speed, even a
> > very brief stoppage to left or right drive power causes the robot to
> > jerk much too strongly to left or right,
>
> Here's what I do and it works quite well. I have added a "partial duty cycle"
> mode to my motor control system. Basically I intermitantly turn off the motor
> for a very small increment of time.... • [snip]
> OH! one very important point: as in the above, rather than turn OFF the motor,
> FREEWHEEL it! This prevents the a jerking motion, resulting in the veering
> behaviour that you experience.
I actually had already tried that, and couldn't get to work in these
particular circumstances. I couldn't get the robot to turn off the
motor for any period briefer than 50 ms, and even that short a period
(coupled with setting the motor to "float" rather than "off") caused too
much of a jerk; a single 50-ms pulse of "float" rather than "on" turned
the robot too much at those speeds. It's a combination of the fact
that:
1. the robot is traveling very fast;
2. the robot has only one light sensor;
3. the tape is only 3/4" wide.
...Anyway, the adder-subtracter seems to solve the problem quite nicely
by gearing the steering motor way down. The fastest I could get a robot
to follow the track using independently-driven wheels was a little over
a foot per second; the differtrans more than doubles that.
One idea I've had for implementing smooth variable-speed motor drive,
but which I haven't gotten around to trying yet, is to use a flywheel.
Do something like this: Motor is attached both to a drive wheel and
(via a 5-to-1 gear-up) to a flywheel (the big spoked wheels from the
Mindstorms set work nicely). Rotation of wheel, motor, and flywheel are
all locked together by the gear train; no differentials or ratcheting.
Hopefully, this will allow controlling the speed of the motor by a
variable duty cycle (as you mention), with the flywheel smoothing out
the "jerks".
I believe the MindStorms motors already are internally geared with some
sort of flywheel to smooth things out; this would just augment the
effect.
|
|
Message has 1 Reply:
 | | Re: Adder/Subtractor revisited
|
| There is a company in Chicago that does a lot with small motors; it's named Bodine Electric, and of course they have a web site at (URL) I can tell the motors in the RCX sets operate with pulse width modulation to control power. Bodine's site gives (...) (25 years ago, 7-Feb-99, to lugnet.robotics)
|
Message is in Reply To:
| | Re: Adder/Subtractor revisited
|
| (...) Here's what I do and it works quite well. I have added a "partial duty cycle" mode to my motor control system. Basically I intermitantly turn off the motor for a very small increment of time. The time-average of a properly synchronized (...) (25 years ago, 7-Feb-99, to lugnet.robotics)
|
17 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
