Shop OBEX P1 Docs P2 Docs Learn Events
Project idea-want feedback and-or sugg. — Parallax Forums

Project idea-want feedback and-or sugg.

MacGeek117MacGeek117 Posts: 747
edited 2006-12-13 17:17 in Propeller 1
I'm not sure if I should leave this here or put it in the Sandbox, so here it is.

··· I am designing a model railroad controller that uses the Propeller as a brain. It will control everthing-right down to the power supplied to the trains.
··· The interface will consist of a VGA monitor and PS/2 keyboard. The monitor will give all nesessary info on what's going on(i.e., speed, switch config, idiv. train info, etc.).
····The interface should be easy to use. I will post the control commands later.
RoboGeek

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"I reject your reality and subsitute my own!"

Adam Savage, Mythbusters
www.parallax.com
www.goldmine-elec.com
www.expresspcb.com
www.startrek.com
·

Comments

  • GavinGavin Posts: 134
    edited 2006-12-09 04:30
    DCC?
    Seperate boosters, one for programming?
    Autoreverse loops?
    Pull a mouse apart to get three encoders for 3 loco throttles?
    Touch screen VGA

    Could be interesting, should be easy to use waitcnt to generate the 10khz pulse train.

    Too many things to do, not enough time.

    Gavin


    ·
  • MacGeek117MacGeek117 Posts: 747
    edited 2006-12-12 15:59
    >>DCC?
    What is DCC? (Haven't worked with this stuff before.)
    RoboGeek

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "I reject your reality and subsitute my own!"

    Adam Savage, Mythbusters
    www.parallax.com
    www.goldmine-elec.com
    www.expresspcb.com
    www.startrek.com
    ·
  • edited 2006-12-12 17:21
    Hey Gavin and RoboGeek... I see that you suggesting using the wheel encoders and the mouse hardware for the controlling the motors... do you remember that·we were talking about·a similar idea a couple weeks ago on another thread?.. I remember you liking the idea...· well, I have done some tests and it really works!!.. I still have to take pictures and perform some more tests,· but I·think the idea is perfect for hobbyists with limited resources (like me :-)·· talking about the mouse hardware, I've got curious about how the mouse determines the direction the wheel is rotating, and it is similar to a quadrature encoder, but instead of the typical 90 degrees separation of the sensors, the assembly simply uses a common infrared emitter, and the receiver is a little package with two infrared receivers closely put side by side, so, to determine the direction, it simply detects who of the detectors senses light first, so if the left sensor sees light first, then the wheel is rotating towards that sensor, and viceversa (I know you get the idea).. :-)· .. one thing that I like is that the mouse IC controller actually buffers the count of the pulses, so, at 8 bits per encoder, that means it can count 255 pulses before the information is read by,·let's say mouse.spin, otherwise·it overflows the counter and generates an overflow error,· I saw that the mouse.spin objects states that it samples the readings at 200 samples per second,· so that would be around 255 x 200 = 51000 pulses per minute maximum, I am probably wrong on this estimate, but I am certain this idea can still count a whole lot of pulses per second without missing pulses... perhaps we can join forces in modifying the mouse.spin to be more adecuate for RPM and motor position measurement... smile.gif

    What I have done so far, take two identical mice (logitech);·cut off the infrared emitter/receiver sensor assemblies for both X and Y axis·with a dremel tool, reconnect the assembly to the remaining working mouse with 10 inch wire, print 36 and 72 tines wheel encoders on transparency sheet, hook it to a cordless drilling tool (not high RPM)· and then connect the mouse to the Propeller and run mouse.spin· and show on the TV constant readings of delta_y· / delta_x;· not high tech but I think if proofs the idea· smurf.gif····· I also hooked the original mouse wheel encoder to the drill with same results...

    One final note is that the Z-Axis (the Scroll Button)· does not seem to·have the same resolution as the Y or X axis, at least on cheaper Logitech mice, you need to check the Mouse IC Controller Datasheet for the particular mouse you are using, however, for low speed motors it should still be fine, or you will have to create wheel encoders with less tines...



    I hope this helps....



    Post Edited (Joe "Bot" Red) : 12/12/2006 6:21:40 PM GMT
  • GavinGavin Posts: 134
    edited 2006-12-13 07:56
    Joe,

    Cool stuff.
    Little Machine Shop actually sells CNC type jog wheels.
    Wonder if it is compataible to the mouse hardware?
    Mini numerical keypad, nice box and I have a manual cnc pendant.
    mouse Z axis is normally a little switch make/break type, not optical, could be converted to optical?

    Robogeek,
    DCC - Digital Command Control, what most modeltrain guys use for digital train control.

    Gavin
  • edited 2006-12-13 17:17
    Hey Gavin...·· thanks for the link to www.littlemachineshop.com... tons of cool stuff there...· About the Jog Wheel... are you talking about the Jog Dial Wheel??... excuse my ignorance... but I do not know much about this type of machines (not yet at least :-)···· so I browsed the catalog.. and I ran into the Jog Dial Wheel... which generates quadrature A / B· encoder signals...· I guess this wheel is used to finely control certain aspects of the machine..·· well.. I believe you can easily connect this Jog wheel to the mouse PCB...· look at the·"HT82M398A Photo-Coupler Connection Diagrams.jpg"· attached·mouse squematic...·· you can remove the Infrared Receiver sensor and then connect the A and B outputs from the Jog Wheel to lets say X1 and X2 in the IC... presto...

    About the Z-Axis (Scroll Wheel)··you are right about the some mice using switches... but look in that squematic again... it shows that it also supports infrared sensors.. in·fact..·the·old mice I am using (logitech 2-button with scroll wheel)· uses infrared sensors for the scroll wheel...· :-)

    But the particular issue with the Z Axis is that the time gap between phase A phase B· in order to detect the pulse is much longer than the X and Z axis....· look at the "HT82M398A Photo-Coupler Timing Diagrams.jpg"·· the bottom note states that Z axis·requires a much longer·time gap between the raise of pulses in phase A and B compared to the other axis... that definitely affects maximum measurable RPM·of the motor and the number of tines per wheel...

    Perhaps if you are interested in making the Jog Wheel yourself..· here is an idea...· http://www.makezine.com/blog/archive/2006/08/how_to_make_a_spinner.html

    I hope this helps... I wish I had more time for this mouse hack.. but I am working on giving my Lynxmotion Extreme Hexapod III a brain based on my beautiful Propeller Chip.... :-)
Sign In or Register to comment.