DPRG mailing list archives: [DPRG] Placement and interpretation of ultrasonic

Tue Jun 5 13:06:58 CDT 2007

On the Tankbot, we had mounted a small array of Maxbotix sensors.
While they _are_ very narrow beam width sensors, they will still cross
talk.

The simple solution used on the Tankbot was to poll each sensor one at
a time.  We used the pulse width output, and simply waited the full
255ms time period for each sensor, regardless of when the pulse came
back.  This gave us a constant known sensor rate (all four sensors
return each second.

You could also simply wait till one sensor returns its data, and then
pulse the other. For two sensors, this would allow for a high refresh
rate from the sensors.  This usually means you can detect objects and
the associated collision faster, and therefore your robot can drive
faster.

Ed

On 6/1/07, David Peterson <robodave1 at tx.rr.com> wrote:
> Hi Brian,
>
> The Maxbotix sensors state having a narrower detection area than other sensors,
> but still might have the possibility of interference with each other. As a
> single transducer sensor, placement might be as easy as a simple 5 degree angle
> between the 2, as dpa's SR04 makes use of with Polaroid transducers (not quite
> parallel, but almost appearing so). Some sonar sensors need to be placed at bit
> of a height off the ground, maybe 4 or 6 inches up, to avoid any false ground
> reflections (picking up ground cracks and such). Your placement could also
> depend upon what you want to do with your robot. If just to wander around in a
> area, having both sensors facing forward (with a small angle between) would work
> to allow the robot to detect most obstacles in its path. If wall following was
> an expected, then one sensor could be aimed straight ahead for oncoming walls,
> with the other aimed to the side, to maintain a distance from the side wall. Be
> prepared to change mounting to get better results in what you wat to do. It
> might even be good to mount them on servos, so that you can have them aimed as
> the robot needs in different situations.
>
> On crunching numbers, figure out which method you are using to read them, (pulse
> width, variable voltage, or serial data) and how the numbers could be used to
> change your motor output. You could use the returned info as just a setpoint,
> causing the robot to do some action if the numbers are greater than or less than
> a certain amount. Or you could use the changing return to produce a variety of
> responses, such as special types of turns when the numbers show the robot is too
> close to something. This could be a number of hard determinations, or the
> numbers could part af a formula that directly determines PWM to motors for a
> continuously variable response. The best would probably be to experiment and see
> what gets the response you want from the robot.
>
> -robodave
>
> ----- Original Message -----
> From: "Brian Clough" <bclough at banksofthesusquehanna.com>
> To: "Dprglist at Dprg.Org" <dprglist at dprg.org>
> Sent: Friday, June 01, 2007 8:52 AM
> Subject: [DPRG] Placement and interpretation of ultrasonic
>
>
> Guys:
>
> I have two MaxBotics ultrasonic sensors, that I want to use to give the
> Hexbot some "eyesight".  I need a good way to place these two sensors,
> and some information on how to crunch the incoming numbers, to give the
> processor a good picture of surrounding obstacles.
>
> I don't know if I should place both facing directly forward, at slight
> angles, one forward and another at an angle, or something else.
>
> Does anyone have some general advice on this topic?
>
> Brian
>
>
> _______________________________________________
> DPRGlist mailing list
> DPRGlist at dprg.org
> http://list.dprg.org/mailman/listinfo/dprglist
>
>
> --
> No virus found in this incoming message.
> Checked by AVG Free Edition.
> Version: 7.5.472 / Virus Database: 269.8.5/826 - Release Date: 5/31/2007 4:51 PM
>
>
> _______________________________________________
> DPRGlist mailing list
> DPRGlist at dprg.org
> http://list.dprg.org/mailman/listinfo/dprglist
>