Newsgroups: comp.robotics
Path: brunix!news.Brown.EDU!noc.near.net!chaos.dac.neu.edu!random.ccs.northeastern.edu!news
From: cleary@everest.ccs.northeastern.edu (Michael Cleary)
Subject: Re: Denning MRV3 Ultrasonics Problems ?
In-Reply-To: moore1@u.washington.edu's message of 25 Feb 1994 02: 25:15 GMT
Message-ID: <CLEARY.94Feb25121110@everest.ccs.northeastern.edu>
Sender: news@random.ccs.northeastern.edu
Nntp-Posting-Host: everest.ccs.neu.edu
Organization: College of CS, Northeastern University
References: <2kjnib$5pv@news.u.washington.edu>
Date: Fri, 25 Feb 1994 17:11:10 GMT
Lines: 24

In article <2kjnib$5pv@news.u.washington.edu> moore1@u.washington.edu (Victor Moore) writes:

	   I have been using the Denning MRV3 robot for some time now and
   have noticed problems with the ultrasonics. It seems that was the robot 
   runs into a wall the ultrasonics get rather flakey (as you might expect).
   But we found that the ultrasonic returns become a function of how the 
   wires are placed in the base of the robot. If the wires are too close to
   the top, or bottom then the we get short returns of about 3.6ft even when
   we should obviously be getting returns of about 12-15ft. As we adjust one
   wire other ultrasonics get thrown off.


Our problem has been that various sonars simply get dropped unpredictably,
resulting in short return messages (22 or 23 instead of 24 readings).  The
cksum (or xor) bytes sometimes correspond to the short set and sometimes to
the full set of readings that should have arrived in that circumstance. Have
you ever had this problem or one like it?

Thanks, Mike

--
Michael Cleary              cleary@ccs.neu.edu              (617) 373-5242
Robotic & Vision Systems Lab,  Northeastern University,  Boston,  Ma.  USA
                            on the web: http://www.ccs.neu.edu/USER/cleary
