Subject:
|
Re: Maximum sample rate (turn rate) for rotation sensor
|
Newsgroups:
|
lugnet.robotics.rcx
|
Date:
|
Sat, 21 Jul 2001 11:24:39 GMT
|
Viewed:
|
1705 times
|
| |
|
|
Dave and Luis,
Thank you both for the information. I'm big admirer of the
work you've done for the RCX community.
In lugnet.robotics.rcx, Dave Baum writes:
> When using the standard firmware, sensors are sampled every 3 ms. The
> rotation sensor uses a quadrature encoder with a 1:4 gearing between the
> axle and the encoder, so there are 16 "ticks" per revolution of the
> axle. The decoding routine in the software needs to see every
> successive tick, so with 16 ticks per revolution and 3ms per tick you
> get 1250 revolutions per minute.
>
> Note that this assumes everything is absolutely perfect...
I see. (60/0.003)/16 = 1250 And if you were worried about
the Sampling Theorem you would divide that in half just to be
sure.. giving 625 or something like the 500 value published in
http://guide.lugnet.com/set/9756
That 1:4 gearing also explains why the sensors generate
a rather high load at higher RPM. I was tempted to crack one
of the things open to see how they worked. But they were
a bit expensive to be doing that kind of thing.
Incidentally, I just competed in a dead-reckoning contest
using NQC to control a differentially steered robot [wheel-chair
type drive system]. My gear train coupled the sensor to the
motor with a 1:1 ratio, so sampling rate was not a problem.
The motor-to-wheel reduction ratio was 12.5:1, giving me
a slow but accurate robot that could sample 12.5x16=200
times per wheel revolution. It was my first experience with
NQC (and mindstorms), but the language worked like a charm.
gary
|
|
Message is in Reply To:
4 Messages in This Thread:
 
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
