No class monday - my progress - your progress ?
UHCL SenProj 010 spr/fall
at school till about 5pm (yay). This is what I did.
- added battery pack
- mounted sensors
- hooked everything up
- tested with bjt transistors (my mosfets were not here yet) it was
very slow even with extra batteries
- soldered mosfets in parallel (after they showed up) and used them
- lots of testing and rebuilding the circuit.
You can check it out in this youtube video http://www.youtube.com/watch?v=f3nXB5y78g0
it has annotations.
I'm integrating the pic to the chassis next week so we can start
that.
Even though there will be no class monday I still expect everyone to
do their part that was outlined in
http://groups.google.com/group/uhclseniorproj010/browse_thread/thread/39ef9e117f55be9a?hl=en
Duc I havent looked at the wbs in a long time, where are we are we on
track ? I plan on figuring out how to mount my microchip pic board
(http://www.zonemikel.com/wordpress/?p=274) to the bot next week so we
can do some more heavy duty stuff. We need the wbs to reflect what we
are doing and who is doing what, we also need to follow it as much as
possible unless it is just dumb.
Trent im using version 4.038 it is working for me. I'll try and make a
basic program for the pic after i mount it to the chassis, once i've
run it i'll send it to you so you can mess with it. Meanwhile do you
mind making user requirements ? Can you post them in the pages or at
least the disscusion part of google groups, they are pretty
important ! Just make up some decent resonable values for everything
till we see what we can do, of course all user requirments will not be
met. I'm thinking something like this
User Requirements-
- must follow line
- speed ?
- line width
- line color
- background color etc, etc
- must be able to control via remote
- distance (range)
- directions
- etc, etc,
- must be able to navigate by itself
- detect walls
- detect dropoffs
- detect collisions
- keep track of where it's been
- keep track of distance traveled and in what direction
etc, etc, etc, thanks !
Scott, i already made pwm using a 555 timer that might be a option, i
was using it for something else but it would work for driving motors.
It consist of 2 caps 2 resistors and the ne555 ic. You could figure
out all our options for pwm and then use the stuff we learned in sys
engineering to compare and contrast the designs (eg hoq, pugh) some
things you can use to compare are no of pins, cost, complexity and the
range of signals it can produce. Some of the designs so far are
astable multi vibrator (caps and resistors and transistors), ne555,
the pic mcu or that pwm i bought from mouser.
Joseph you can just write up what youve done with the battery
chargers, and how mine is working (or not working). Also
http://en.wikipedia.org/wiki/Inductive_charging you can look up
inductive charging and see if we could use that or not ?
Down the road >>>>
I'm thinking of making a model using the 3d printer we have at
school. Like a chassis and maybe tracks. A body and such. Does anyone
have any xp with 3d modeling using programs like autocad,solidworks or
rhino?
Everyone should be ready to mess with the C code we will be using,
this is going to turn into software as soon as the pic is mounted to
the board.
I hope everyone is spending at least 3-5 hours out of class doing the
assigned tasks, I can only do so much. I spent at least 6 hours this
week spanned across a few days, not to mention the time it takes me to
write this and upload videos and order parts etc, etc.
Duc Tran
zonemikel.com/SeniorProj/ to zonemikel.com/sp; Jeebus MIke!! I thought
I had to make up a whole new wbs/gantt chart because I could not find
the previous one.
Anyway, I've added a couple of things to the charts and moved up some
of the tasks that we have to do. My reasoning is that near the end of
the semester we'll all be busy studying for finals and etc... By the
way, do we have to give a presentation at the end of the semester?
New wbs/gantt chart can be found under files - SP2010
I'll make a new page for the accelerometer tonight and will be
extending what Trent has done for the architectural design later this
week.
On Mar 28, 12:09 pm, "UHCL SenProj 010 spr/fall" <zonemi...@gmail.com>
wrote:
> Hello everyone, there will be no class monday so I wont be showing up
> at school till about 5pm (yay). This is what I did.
> - added battery pack
> - mounted sensors
> - hooked everything up
> - tested with bjt transistors (my mosfets were not here yet) it was
> very slow even with extra batteries
> - soldered mosfets in parallel (after they showed up) and used them
> - lots of testing and rebuilding the circuit.
> You can check it out in this youtube videohttp://www.youtube.com/watch?v=f3nXB5y78g0
> it has annotations.
>
> I'm integrating the pic to the chassis next week so we can start
> that.
>
> Even though there will be no class monday I still expect everyone to
> do their part that was outlined inhttp://groups.google.com/group/uhclseniorproj010/browse_thread/thread...
> chargers, and how mine is working (or not working). Alsohttp://en.wikipedia.org/wiki/Inductive_chargingyou can look up
Joseph Lopez
the side, it might be possible to implement a PID controller into the
control logic. Basically it would monitor how fast the wheel is
turning, and whether that was too fast or two slow and would adjust it
accordingly. I'm not fully sure how all it works, I seem to remember
some reading that said they were sampling from the sensors, and others
seem to be sampling from an encoder that tracks the wheel. In any
event, this would probably mean adding another PIC to the robot to act
as the controller. It might be possible to implement it using
discrete components, it is just a transfer function after all. The
problem with that would be the fact that we'd have to tune it until
the robot responded correctly, which would mean swapping out different
component values here and there until we get it right.
Here's a very basic explanation of programming a PID controller and
how it works. I'm trying to see if I can find one of the better ones
I found earlier this year. http://www.societyofrobots.com/programming_PID.shtml
Also, I'm planning to update the charger page sometime later tonight
with a summary of everything I've read up on. I'll send an email to
everyone when that happens.
> > chargers, and how mine is working (or not working). Alsohttp://en.wikipedia.org/wiki/Inductive_chargingyoucan look up
zonemikel.com
| Ok Joseph. We could also use more than 2 transistors to control it, we could add more transistors in as a sort of breaking function that will stop both wheels when it starts to go off the line then lets the wheel that should turn, turn, but it would start out slower. Or we could just slow it down. Pretty much the concept works, problems were discovered and solved, thats all i wanted from it. It functioned as it should the rest would have been "tweeking" which can be a can of worms. We arent designing a 2 tranisistor line following bot though so we need not spend time tweeking this design. I've never heard of PID, good find. If you want keep looking down that alley to see if it is something we can implement to improve our design, if so how would we implement it. Doe it conflict or replace PWM ? Do good, Be Well, Michael --- On Tue, 3/30/10, Joseph Lopez <jolo...@gmail.com> wrote: |
|