Newsgroups: comp.robotics
Path: brunix!sgiblab!spool.mu.edu!uwm.edu!caen!destroyer!nntp.cs.ubc.ca!newsserver.sfu.ca!sfu.ca!lye
From: lye@fraser.sfu.ca (Bill Lye)
Subject: Re: Polaroid(tm) Sensor Update
Message-ID: <lye.747502428@sfu.ca>
Sender: news@sfu.ca
Organization: Simon Fraser University, Burnaby, B.C., Canada
References: <1993Sep7.164326.23820@merlin.dev.cdx.mot.com> <TIM_DEAGAN.4.0009567C@fenain.us.dell.com>
Date: Wed, 8 Sep 1993 15:33:48 GMT
Lines: 47

TIM_DEAGAN@fenain.us.dell.com (Tim Deagan) writes:
>There is an unclear portion in the documentation that says (I left
>the docs at home :-() to leave INIT high for 100ms and then low for 
>100ms.  They actually refer to the INIT line by a different symbol,
>MD something or other.

There are two versions of the Polaroid board.  From your description,
it sounds like you've got the older version, the one that you trigger
essentially by applying power to it!  If this is the case, your 
timing diagrams will include a little note in the time between INIT/MD
going high and TRANSMIT going high, saying that the timing of this is
not guaranteed or some other such stuff.  To switch INIT/MD, Polaroid
recomends using a darlington arrangement.  Not too nice.

The newer version is capable of being switched on directly from CMOS/TTL
logic gates.  Much nicer :)  Also, the timing between triggering INIT and
the start of TRANSMIT is guaranteed, i.e. immediately.

>Is this the way everyone who uses the module well does it?  This limits
>the unit to a 5Hz sample rate.  Not as fast as I'd like.

Well, in our project, we're multiplexing 3 ranging boards 4 ways to
12 transducers using relays.  We use a single 68HC11 to switch the relays,
trigger the INIT's and time the ECHO's.  We get a composite throughput
of just over 7 Hz, triggering each ranging board a bit faster than 28 Hz.

>I feed the INIT and ECHO line into an AND gate to get a pulse who's 
>width represents the distance.  I drive the INIT line from a 555 timer.
>I do this to get as much accuracy as I can from a Basic Stamp.  It's
>pulse measurement capapbility is in 10us increments, since it does
>not have an internal timer I have to count a loop variable which gives
>only approx 1ms resolution.

>I'm interested to hear wether everyone else inthe world knows about the
>100ms up 100 ms down timing.  I'll get the docs from home and quote the
>section verbatim if needed.

We were a little apprehensive when we saw this in the Transducer training
kit, and were much relieved when we the boards we ordered turned out to
be the newer model.

>--Tim
-- 
| Bill Lye, lye@sfu.ca    |   
| 5th year ENSC Undergrad |
| Simon Fraser University |
| Mobile Robotics Group   |
