Hello Fred,
I will try to clarify as best I can
That link that you put up is actually the site I got my calibration program
from.
I have attached a file of my simple PID navigation program, the calibration is from the NXT Line Follower Website.(ran in NXT -G inviroment)
My Vi is fairly simple, it uses 2 simple line equation formulas, you enter
in X = (scale Value of light sensor) and it will output desired motor speed,
It uses 2differant formulas one of which is for when the Scale Value is >
or = to 50, the other formula is for when the /scale Value is <50,
The reason that I need/want the calibration is that having a full range of
100 scale values will give me greater resolution for motor corrections. My light
sensor with ought the calibration is giving me a scale value, with the min of
40 and about a max value of 62, and that will be dependent on ambient
light conditions.
So if the light changes with uncalibrated values I would in turn have to
change my formulas every time I pull the robot out, which is why I would much
rather just run a calibration program before each run.
Now that I feel that I have given sufficient background now to dive into the
actual problem I am having.
The calibration program that I have been running works great. It creates a
light sensor.cal file on my brick and when I run the "view light 3"
program the scale value stays calibrated to the 0 to 100 scale indefinitely.
(one thing when I go to class tomorrow I will be making a labview vi to simply
display the light values like the "view light 3" nxt-g program so see
if it is a problem with labview overriding the calibration) I will update you
after class about that
When I run my portable robot.vi it works great also, for the first 10 - 15
seconds, reading from the calibration file and producing the correct values of
0 for the black line and 100 for the white area around the line. after about 10
-15 seconds with MY vi running the displayed scale values revert back to the
original scale values of 40 for the black line and 62 for the white area,
this does not seem to be a problem with the calibration file, because when I
restart my vi, the same thing happens, it starts out with calibrated values, then
reverts back to default readings.
I have tried various ways of using the actual labview calibration block from
the "NTX Native I/O" pallet. I port the code over to the brick and
get an error when I try to run the code, there is no error message other than
"error"
I have tried taking the calibration file off the brick and viewing it, with
no success, to find out later that I need to view it with a binary editor. I
will be doing this soon; one of my partners took the robot home to redesign the
chassis so I will be doing that tomorrow also.
I have also contemplated trying to simply creating my own calibration file through
labview once I view the layout of the original light sensor.cal file
in summary,As of now I am thinking that Labview is overriding the use of the
calibration file, I do not know how or why. also I think being able to make an actual labview calibration program would help out alot, I will look into the math behind the calibration in order to create a program.
I tried to articulate my problem to the best of my ability (writing is not
my strong suit)
Any help would be great and I hope this added info will be helpful to you
