Shop OBEX P1 Docs P2 Docs Learn Events
Motion recording problem — Parallax Forums

Motion recording problem

AImanAIman Posts: 531
edited 2010-04-29 03:02 in General Discussion
How can I get information from various items to record exactly what happens when they are NOT being used to create motion?

This is part of a weather sensing device. In short this looks like a jointed flag pole. The top is where the cups are attached to the motor, the flag part is where the panels are to collect info on wind resistance, light, sound, humidity and electricity and the joint is where the servo goes about halfway down. At the base is a swivel so the stuff can turn with the wind.

It has to be able to track when the cups attached to the motor move to track wind speed, it has to be able to track the rpm as it accelerates or decelerates and for the correct lengths of time and the correct durations at each rpm. It has to be able to collect how far a servo moves back and forth to record wind gust effects on the various resistance panels, it needs to collect the pressure and strain put onto the panels and also collect info about light, sound, humidity and electricity (for static build up).

Heres the catch, the stuff has to be able to reproduce exactly what happens for lab replication. So if a wind comes and bends the pole at the joint to a 15 degree tilt, then the servos need to be able to record a 15 degree tilt at the same speed and for the same length of time that the pole was pushed by the wind. If the panels show dusk conditions with static build up then these conditions need to be able to accurately detect the same conditions in a lab. If rain is detected, then the correct levels of moisture need to be recorded, over the correct time frames so they can be duplicated in the lab.

I have experimented with using servos and motors, but can't get it to record info or to replicate the original motions with the original times.

Post Edited (AIman) : 3/20/2010 3:09:14 AM GMT

Comments

  • kwinnkwinn Posts: 8,697
    edited 2010-03-20 03:40
    For recording the info would you not want to use optical sensors rather than motors and servos?

    It sounds like a simple project. Install sensors to take readings regularly and record the time and sensor readings.
  • FranklinFranklin Posts: 4,747
    edited 2010-03-20 03:42
    First get the data storage part working then worry about the rest. Accurate data stored with an accurate timestamp is your first project.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    - Stephen
  • AImanAIman Posts: 531
    edited 2010-04-05 17:58
    I guess thats where I get confused.

    How can I make an accurate time stamp?

    My past projects were simply take an ongoing time reading, when something happens put a time stamp there and continue.

    This however seems to me baffling because the time stamp has to be able to log with small movments. In other words if an IR is used to measure how far away something is from its start point it needs to log the distance and the time so that speed can be determined. However, loging both eludes me.

    Shouldn't I be able to do a simple time stamp with every measurment change?

    For data storage, it seems to me an EEprom will wear out to quickly so will it be best to hook up a thumb drive or go straight to a desktop?
  • FranklinFranklin Posts: 4,747
    edited 2010-04-05 19:11
    Could you show us the code you have now and list the sensors you are monitoring and the outputs from each? If you have an accurate time you are halfway there. Also remember anything you write to will take time to complete the write so your racording has to be slow enough to accomidate that.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    - Stephen
  • AImanAIman Posts: 531
    edited 2010-04-27 21:08
    I am crazy busy right now so I can tell you what I have come up with to this point.

    The IR and Sonar are all set with basic start points - a neutral if you will - the sensors get checked by and each time they are checked if there has been any movement in either direction from neutral a time and distance reading is recorded.

    This way I don't need a running time or distance record, and by recording the time and change for each sensor as it occurs I can duplicate the exact info by using the recorded info as a log for a subroutine to use as a base to duplicate info.

    There are issues but for the most part its working ok.

    Likewise a Hall sensor gets tripped when motion in a circle occurs and by measuring RPM and using a time stamp I hope to get an accurate reading of speed.
  • legoman132legoman132 Posts: 87
    edited 2010-04-29 03:02
    You might be able to use a small DC motor and an ADC (and maybe a few discrete components).
Sign In or Register to comment.