Home arrow Forums
914 PC Bots Forum  


<< Start < Prev 1 2 Next > End >>
edaniels
User

Platinum Boarder
Posts: 240
graphgraph
Karma: 2  
Re:How to Config the IR calibration file - 2006/11/30 05:37 The sensor calibration data file should also be in your C:/BRIAN3 directory. You need to hand edit your c:/BRIAN3/BrianConfig.ini to tweak the sensor settings the way you want. All the sensor calibration data file is really useful for is to tell you what the "normal" value is for the 3 downward-pointing sensors -- basically it's telling you the distance from the sensor to the floor, and the values can vary depending on what kind of floor you have. That's why you need to do a calibration.

James' post above mentions the two sections in the manual that you will need to read in order to understand the process better. I tried to describe the process in the manual, and then in the back there is an annotated configuration file that highlights the areas you need to change. Check the "914 Owner's Tips" forum for some more posts that describe the process as well.

-Ernie

Post edited by: edaniels, at: 2006/11/29 21:39
  | | The administrator has disabled public write access.
ajruth
User

Senior Boarder
Posts: 14
graphgraph
Karma: 1  
Re:How to Config the IR calibration file - 2006/12/05 07:42 Any chance on getting that when you get back from your trip ?

Thanks
Andrew
  | | The administrator has disabled public write access.
Ulf
User

Expert Boarder
Posts: 27
graphgraph
Karma: 0  
Re:How to Config the IR calibration file - 2007/03/25 19:20 When i compare the distance of one IR-sensor shown by the GUI to the real distance i found a big (23cm to 20cm -> 15%) difference.

By stepping through the source i found that:
tmp = ((6787.0 / ((float) uiData - 3.0)) - 4.0);
//tmp = (1/(uiData*0.000147 - .00042))-4.00;
//tmp =(int) (27 * pow(((uiData/1024.0)*5.0),-1.1));

Some theorie:
(As far as i know)

Usually you calculate the distance of a sharp sensor by

D = A / ( V - B )

Where
D is the distance (in cm)
V is the value of the AD Converter (M3)
A is a constant of the curve rising
B is a constant of the curve offset

To get the values for A and B you have to do a calibration.
(Has nothing to do what happend in the calibrate_IR function in Brian !)

A = ((V' - V) * D' * D) / (D - D' )
B = (D' * V' - D * V) / (D' - D)

V and V' are measured values for a first (small) distance at about 20cm (D) and a second (large) distance at for example 60cm (D')

This calibration (two measurements with fix known distances) should be made for each sensor and A and B should be individual for each IR to get the best results.
(The mathematics of cause should be done by progam)

Is there a further interest in such a new calibration function or can you live with the present results ?
  | | The administrator has disabled public write access.
GoIrish
User

Platinum Boarder
Posts: 184
graphgraph
Karma: 3  
Re:How to Config the IR calibration file - 2007/03/26 00:51 Ulf,

Thanks for the post - I look forward to trying again with the sensors when I have the bot back together.

-Go Irish
  | | The administrator has disabled public write access.
<< Start < Prev 1 2 Next > End >>