Shop OBEX P1 Docs P2 Docs Learn Events
The Great Outdoors — Parallax Forums

The Great Outdoors

macmanmacman Posts: 20
edited 2005-09-09 23:46 in Robotics
OK, Picture me a newbie, I have just managed to get my BoeBot running around the living room using its IR sensors to avoid any obstacles it may encounter, great stuff and congrats to Parallax on superb documentation.

OK so now we turn to the Great Outdoors, Summer is coming and I would like to swop out the living room for the back yard. The plan to transplant my BOE board onto a platform equipped with knobbly all terrain tires, and two dc motors, a larger battery pack and off we go...

Well I know that is not quite that easy, and I realize that I would need to invest in a motor controller to handle the two DC motors. but would the IR sensors still work OK, would I need to return to 'whiskers' how much extra work is involved and what pitfalls can I expect to encounter? Can/Should I just start my swopping out my servos for DC motors and work from there. I am hoping that those of you with experience in such matters could help me set my goals with any advice along the way. OK I am trying to avoid obvious mistakes that could so easily be avoided by just 'asking'.

I have looked at the variety of speed controllers available but would prefer to stick with Parallax associated products because I know that somewhere there is code and advice available and as a newbie I find that aspect of staying within the Parallax family a great help and a valuable resource.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Regards Nigel

Success is the ability to go from one failure to another with no loss of enthusiasm.
Winston Churchill

Comments

  • Chris SavageChris Savage Parallax Engineering Posts: 14,406
    edited 2005-04-12 01:48
    Nigel,

    ·· There are, of course, any number of roads you can take here (No pun intended).· Basically you could make the decision based on the type of terrain you expect to traverse.· We do have the Tank Tread mod, which works nicely outside, just not in tall grass (Even some short grass).· You can add a Memsic Accelerometer for tilt detection, etc.

    ·· If you decide you need more speed and bigger wheels, then yes, a motor controller is in order, but you won't be able to choose until you find suitable motors and know their requirements.

    ·· As for the IR detection.· It will probably have to be slightly adjusted based on the height of the Bot and what type of terrain you expect.· You could also mix sensors.· Or, add a PING sensor.· Then the IR could be used solely for lower object/edge detection.



    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Chris Savage
    Parallax Tech Support
    csavage@parallax.com
  • Paul BakerPaul Baker Posts: 6,351
    edited 2005-04-12 01:54
    I can't help you on your mechanical choices, but gearbox DC motors and H bridge for each sounds like a smart choice. Remember the larger the wheels you use, the higher the stall torque required from your motors. And yeah, your going to have to scrap your IR sensors the IR component in sunlight would drown out your sensors. I would go with an ultrasonic ranger, this should also give you greater and more accurate distance measurements which could come in handy if your bot is traveling at a nice clip.
  • Chris SavageChris Savage Parallax Engineering Posts: 14,406
    edited 2005-04-12 01:58
    Paul,

    ·· I think the IR sensors could still be used aimed down for edge detection, or, at a lower level for close-up object detection below the line-of-sight of the sonar.· I will have to try this in practice, but I would imagine the usefulness would go down, but I wouldn't think to zero.· I could be wrong though, my bots have yet to see outdoors.· I will be trying soon now that I am in better weather.· It's still several feet of snow in my old neighborhood!


    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Chris Savage
    Parallax Tech Support
    csavage@parallax.com
  • macmanmacman Posts: 20
    edited 2005-04-12 11:42
    Thanks for the info, that really helps me start to focus on the specifics. Paul you mentioned stall torque, the motor I had in mind is a 7.2v Gearhead with a 50:1 ratio, 175RPM with a stall torque of 7.1Kg-cm. I had intended to pair each motor with a 5" r/c off road wheel would that work OK?
    If I wanted to run more than two motors say 4 or 6 could I link them in parallel and run them that way?
    As for the Ultrasonic would the new Ping))) Ultrasonic from Parallax work and would I need one or two or more?
    Thanks again guys, I really appreciate the help.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Regards Nigel

    Success is the ability to go from one failure to another with no loss of enthusiasm.
    Winston Churchill
  • JonbJonb Posts: 146
    edited 2005-04-12 12:34
    Paul Baker said...
    ·And yeah, your going to have to scrap your IR sensors the IR component in sunlight would drown out your sensors.

    Paul,

    Have you tested this? I was under the impression that the reason we modulated the IR signal(~38khz) was because we needed to eliminate interference from, among other sources, sunlight. Most remote controls work well in sunlight and I wonder what your source was for this information.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  • bishopbishop Posts: 82
    edited 2005-04-12 12:39
    i've tried what you're basically about to do and let me give some observations that I had. The boe-bot even with larger wheels is not really suited for outdoor travel. Because its basically a tail dragger, you may find that the butt-ball will snag on grass, rocks and bumps and slow down the bot. The one ball that i found that would travel the grass with the larger wheels was a wiffle ball (no holes kind) that I cut a slot in and popped on to the rear of the bot. that seemed to work o.k. the tracks worked out better for me, but I had to 'rig' them into a proper alignment and tension to get them to not slip so much.
    good luck!

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    **************

    daniel woolston
    Teksystems Inc.
    www.danwoolston.com
    **************
  • Chris SavageChris Savage Parallax Engineering Posts: 14,406
    edited 2005-04-12 15:10
    macman,

    ·· At 5 for $100.00 you could afford to mount one on each side, and still have an extra for another project.· Perhaps a garage parking indicator...wink.gif

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Chris Savage
    Parallax Tech Support
    csavage@parallax.com
  • Paul BakerPaul Baker Posts: 6,351
    edited 2005-04-12 15:23
    Jonb said...
    Paul Baker said...
    ·And yeah, your going to have to scrap your IR sensors the IR component in sunlight would drown out your sensors.
    Paul,

    Have you tested this? I was under the impression that the reason we modulated the IR signal(~38khz) was because we needed to eliminate interference from, among other sources, sunlight. Most remote controls work well in sunlight and I wonder what your source was for this information.
    Not specifically, I am approaching it from a noise in linear systems perspective. The IR is putting out a fixed energy in candela (or luminous flux). Sunlight has for the most part is a continuous spectrum (with a few darker spots where molecules in the atmosphere absorb particular wavelengths) the luminous intensity picked up by the photodiode or transistor will be the aggregate of the signal plus sunlight. At a certain point the signal will be undetectable because the signal·is swamped out by the noise. I cannot say at which point this becomes an issue because this is dependent upon how its mounted etc., but Chris is correct that as a line following module it should still operate correctly, though Im not sure how useful that would be for this application. The insensitivity to noise will depend on the exact circuit use for the receiver and its method of signal amplification. If the bias point (caused by the sunlight) is accounted for and the photodiode/transistor is used as a small signal amplifier it would be possible to still detect the signal if the photodiode/transistor is not in saturation. If the receiver becomes saturated the small signal amplification is greatly diminished and detection becomes a problem.

    WRT the motors, those sound strong enough for your application.
  • Chris SavageChris Savage Parallax Engineering Posts: 14,406
    edited 2005-04-12 16:12
    ·· In layman's terms (In case Paul lost you... tongue.gif , the light does interfere with the IR detectors sensitivity to a degree, and there are a lot of factors that may or may not come into play.· The best thing you can do is try it, see if it works, and if not, find another option.· If you're in the Arizona desert, or in Alaska (Or the antarctic) you will probably need something else!· shakehead.gif

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Chris Savage
    Parallax Tech Support
    csavage@parallax.com
  • Paul BakerPaul Baker Posts: 6,351
    edited 2005-04-12 16:43
    Lets see if I can draw an analogy, say you are taking an analog to digital conversion of the sensor, and the total range of voltage is from 0-5V. And say that the reciever when in close proximity to an obstacle produces a signal on the detector with a 1V (.5V P-P) 38kHz sinusoid. If the sunlight component places the line at 4.8V, .3V of the signal will be cut off because a voltage greater than 5V is not possible. This reduces the ability of the receiver to detect the signal because some of the signal is chopped off (4.8V+.5V>5V). There are circuits which can extend the usefulness of the IR detector since·the sunlight component can be thought of as a common mode signal. By using two detectors in a differential pair where one receives sunlight plus the signal and the other detector receives the same amount of sunlight with no signal present, the second one will subtract out the common mode (sunlight) from the other. But saturation of the detector is still an issue (in lay terms saturation can be correlated to a speaker being driven too hard and causing distortion, you're trying to operate it in a region it was never designed to be used)
  • macmanmacman Posts: 20
    edited 2005-04-12 17:12
    I think given the layers of complexity I would prefer to try the ultrasonic sensors, because I have not tried these before and it would appear that they may be bettersuited to the task.
    I take the point that daniel woolston made about the BoE Bot trailing ball, and in order to rectify that I would like to add another two wheels giving me a 4WD base, but the motors I had marked for use have a 200mA draw with a stall rate of 3.80A so If I equip my platform/base with two motors on each side wired in parallel that would give me a stall rate of 7.6A checking through the stats there doesn't seem to be a motor driver that can handle that much. Is there another driver or alternate option other than running four individual driver/motors. I am willing to make a custrom H Bridge but would need a good schematic and directions. Any ideas!!

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Regards Nigel

    Success is the ability to go from one failure to another with no loss of enthusiasm.
    Winston Churchill
  • FireHorseFireHorse Posts: 5
    edited 2005-09-09 23:46
    I wonder if anyone tried to·hack 'e VAC' by 'Sharper Image'. It seems to be maybe the cheapest robotic platform for outdoor applications (I paid $70-, including shipping for mine). It got 5" wheels, good motors and sensors. I bet you can hook up BS2 or Javelina and other Parallax stuff to it.

    VK
Sign In or Register to comment.