Subject:
|
Re: Rotational Sensor & Gearing Down
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Thu, 25 Jan 2001 20:58:01 GMT
|
Viewed:
|
774 times
|
| |
|
|
John Barnes wrote:
> >
> > I'd *hope* not. If I were designing the RCX, I'd make the
> > rotation sensor generate high priority interrupts so that
> > sensors would be accurate no matter what...but who knows?
>
> I just read the very interesting posting regarding testing the rotation sensor
> to evaluate the "missing counts" phenomenon.
>
> The above snippit would indicate that the rotation sensor is handled
> differently from other sensors. It isn't!
>
> It is sampled at the standard rate (one sample every 3 mS).
>
> It's output is an "analog grey code" switching between four well known
> voltages.
>
> The software in the RCX must evaluate the rotation since the last sample
> based on the change in this single analog signal.
>
> Once the angular rate of the rotation sensor approaches one change every
> 3mS, all bets are off.
>
> At low speed, there is a dither problem. The sensor output can be sampled
> right at the edge of changing from one state to another. Since the connection
> to the RCX is not particularly fast, there is a very real chance of catching a
> reading from the sensor of a voltage which is bogus.
>
> The only reason I know this is because I had to write code to handle this
> sensor
> for another hardware platform and thought I had problems with my code. At that
> point I evaluated the RCX in detail and that is what I discovered.
>
> Put a longish handcrank on a rotation sensor and wind it back and forth
> from the
> "edge" of one count to the "edge" of another. You'll pick up lots of
> missing counts.
>
> However:
>
> I use rotation sensors to keep skid steer devices going straight with great
> success;
>
> Clear the rotation sensor channels first.
> Run the motors using float commands to balance the counts.
> Wait until the sum of the counts equates to the distance, appropriately
> scaled, needed
> to be travelled.
>
> If you want to build a angular positioning servo system, use a small
> potentiometer in a
> brick and read it with the input port in passive mode. It works great! And
> its absolute, so
> there's no need to build a home mechanism. (I think there's reference to
> this on Mike G's
> page.)
Just don't go past it's Max or Min areas or you'll have a problem. This is
probably a good place for the white slip gear...
Dean
--
Coin-Op's For Sale!: http://www.akasa.bc.ca/tfm/coin-op.html
Dean's Lego Workshop: http://www.akasa.bc.ca/tfm/lego_wr.html
Vancouver Lego Club: http://www.akasa.bc.ca/vlc
|
|
Message has 1 Reply:
 | | Re: Rotational Sensor & Gearing Down
|
| Thanks to all for the good information that has been related here. Most of my applications were low rpm which seems to be a definite fault area. I like to use them for absolute programming of positions. I now see the problem in using them in these (...) (23 years ago, 25-Jan-01, to lugnet.robotics)
|
Message is in Reply To:
| | Re: Rotational Sensor & Gearing Down
|
| (...) I just read the very interesting posting regarding testing the rotation sensor to evaluate the "missing counts" phenomenon. The above snippit would indicate that the rotation sensor is handled differently from other sensors. It isn't! It is (...) (23 years ago, 25-Jan-01, to lugnet.robotics)
|
19 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
|
|
|
|
