Subject:
|
Re: Ooops! NXT Software Comparison correction / I2C Messaging Speed
|
Newsgroups:
|
lugnet.robotics.nxt
|
Date:
|
Wed, 5 Sep 2007 23:28:47 GMT
|
Viewed:
|
22329 times
|
| |
|
|
In lugnet.robotics.nxt, John Hansen wrote:
> I'm not talking about getting bus errors - just invalid
> distance readings from the sensor.
I think I understand. In this case, the limiting factor is the US sensor itself,
how fast it can ing and interprete a return. Or from my standpoint (a
physicist), it's a fundemental limitation of the physics of sound, not a
limitation that has anything to do with the I2C speed or the FW speed issues.
> if you are really interested in valid comparisons between
> different development platforms.
You seem to have a point here John - Steve has accidentally selected the wrong
sensor, one that if the comparision is completely analogous across all platforms
(one of which slows down the reads to make sure a "fresh" reading, ultimately
out of date by something like 15+ ms), unfairly penalizes those solutions that
can out-read the US sensor (a rather slow sensor at that). A better comparision,
one that would properly show the performance of the various environments, would
be to remove the bottleneck of the bit-banged US sensor if that's what is really
the problem, subsituting in some other sensor that will not limit the I2C speed.
Of course, and only from my point of view again, it seems obvious that a
limitation in the way a certain sensor responds is not a good measure of a
limitations in the SW reading it. I think even the US sensor & LEGO seems to
admit to this, as it will return the same reading from an on-sensor buffer when
polled faster than the actual sensor can update the position.
Truthfully, I'm still not sure what the major issue is here - the comparision
seems to do just what it was designed to do, make a reasonable (not perfect)
comparision between different solutions.
--
Brian Davis
|
|
Message is in Reply To:
13 Messages in This Thread:
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
Active threads in NXT programmable brick
|
|
|
|
|
|
