Usage of IR and what it takes

[Phil Roth]

AM I missing something here???. After review of the Navigate project for IR receiving, it seems that 4 external interrupts from each sensor are used to check the frequency and then count IR pulses from the beacon. The real time ISR then accumulates how many pulses occur every 26.2MS.

The questions I have are:
The code refers to the EDU controller, which I see has external interrupts clearly marked on the silkscreen, the FULL SIZE controller does not have "seperate" interrupt pins avalable. Can the PIC be configured to use the GPIO pins on the controller as IRQs??

Some other comments:
Also the system is loaded with "fudge factors" which ultimately will determine final accuracy.

There is a foward autonomous run to come off the wall and stop, looking for IR synch. Then hopefully the robot finds, turns, and goes straight.

I'm no IR expert, but the approach seems elegant but painfully slow, having an articulated servo system working the IR sensors then feeding back error information to the drive wheels. I'd like to get there in under 5 seconds.

I noticed K.Watson has been hammerring on this code since November.. good thing, I'd hate to write it from scratch. By the way. Who the heck is Brian, Gail, Patricia and Patrick.????

[Kevin Watson]

Originally Posted by Phil Roth

The questions I have are: The code refers to the EDU controller, which I see has external interrupts clearly marked on the silkscreen, the FULL SIZE controller does not have "seperate" interrupt pins avalable. Can the PIC be configured to use the GPIO pins on the controller as IRQs??

They're just not marked. It's discussed here.

Originally Posted by Phil Roth

Also the system is loaded with "fudge factors" which ultimately will determine final accuracy.

These are parameters that will need to be tweeked for your specific application.

Originally Posted by Phil Roth

I'm no IR expert, but the approach seems elegant but painfully slow, having an articulated servo system working the IR sensors then feeding back error information to the drive wheels. I'd like to get there in under 5 seconds.

Yes, we know it can be done better. We wanted to leave a lot of room for improvement.

Originally Posted by Phil Roth

I noticed K.Watson has been hammerring on this code since November.. good thing, I'd hate to write it from scratch.

I wrote part of the code. Rich Petras wrote the navigation code (Rich, BTW, is a member of the MER navigation software team and mentor to team 691).

Originally Posted by Phil Roth

By the way. Who the heck is Brian, Gail, Patricia and Patrick.????

My oldest (10) son, my wife, my mom and my youngest (6) son .

[Phil Roth]

Kevin,
Hopefully my comments were not viewed as criticism towards your efforts. I assumed the code was farmed out to some consultant and not written by a peer roboteer engineer. It may help if FIRST comes out with a simplified white paper explaining the I/O that's needed along with some tuning tips which are probably linked closely with mechanical dimensions. Naturally a lot of investigation and reverse engineering is left to students, but it may not be practical because of the new "stuff" like 'C' and only 6 short weeks to get it going along with the rest of the robot. Mount Olive Robotics Team (MORT) # 11 from New Jersey will have IR seeking as part of it's Autonomous mode.

From what I see by the code there are 4 IR sensors, each connecting to digital inputs used as Interrupts. It seems 2 sensors are mounted as pairs on the small servo. The code "hunts" for the IR signal recieved by each pair by moving the servo until a signal is locked on. The mechanical distance between the "pairs" mounted on the servo should be as large as possible to produce a large delta between them. When both pairs are "locked on", naturally there is a difference in the angle between pairs, and the delta is used as error information to steer the wheels.

We are going to start building the beacon and reciever this week and getting familiar with it. I hope you will be around for some help if needed, on the other hand if you think I can help in documenting anything for the students, making it easier to get this going, please let me know.

Thanks very much for your reply and the coding

Phil