CMUcam default code not responding

[Kevin Watson]

Originally Posted by jaustin

Yes, it worked fine with LabView using the DB9. I have all the calibration numbers for tracking an orange traffic cone to get us started (our green light is not assembled yet!). I guess that would narrow it down to the camera's 3 pin RS232 port? That does not seem very likely. I'll keep playing...

You might consider just connecting it to the programming port. If you do, you'll need to switch the port in camera.h to point to serial port one.

-Kevin

[Ryan O]

Originally Posted by Kevin Watson

If you're using a USB to serial converter, try to find a PC with a real serial port and try it again.

-Kevin

I am using a Compaq Armada 7400 with a serial port. The LEDs are fflashing when I attempt to use the code, after the .hex download is complete. Thanks for the suggestion though.

[Kevin Watson]

Originally Posted by Ryan O

I am using a Compaq Armada 7400 with a serial port. The LEDs are fflashing when I attempt to use the code, after the .hex download is complete. Thanks for the suggestion though.

You're using a 2006 controller? If so, the code is getting corrupted somehow and you should contact IFI for help diagnosing the problem. If you're using an earlier controller, you need to re-compile the code using the included instructions. Come back here if the above doesn't help out.

-Kevin

[Ryan O]

Originally Posted by Kevin Watson

You're using a 2006 controller? If so, the code is getting corrupted somehow and you should contact IFI for help diagnosing the problem. If you're using an earlier controller, you need to re-compile the code using the included instructions. Come back here if the above doesn't help out.

-Kevin

Thanks, but I just ave one question. Where are the recompile instructions? I am using last year's bot as a test platform. Are they in the work or another of the .pdf's from FIRST? I did a scan through the workbook, but I may have missed it. Thanks alot, you guys are very helpful. Good luck with your bots this year.

[Kevin Watson]

Originally Posted by Ryan O

Thanks, but I just ave one question. Where are the recompile instructions? I am using last year's bot as a test platform. Are they in the work or another of the .pdf's from FIRST? I did a scan through the workbook, but I may have missed it. Thanks alot, you guys are very helpful. Good luck with your bots this year.

"Using_Last_Years_FRC.txt" included with my code as well as IFI's.

-Kevin

[heydowns]

Originally Posted by jaustin

Hi Kevin,
We are having the same issue (failing in State 3) - something probably amiss with our camera to RC communication. The readme file for the diagnostic files says it's only for the full sized controller. We are still messing with the EDU until we get our full system running. Any tweaks I can do to get the diagnositcs to work with the EDU?

We use the camera with the Edu... We got the same error initially. We found that the baud rate was set wrong. If you are using Serial Port #2 (the TTL header is serial port #2), you need to change Kevin's serial driver to use a baud rate that matches your camera (the default is 115200).

Another thing to check - make sure your RS232<>TTL converter board is fully connected. The Edu has only 3 pins on the TTL header. The missing fourth one (as compared to full size RC) is the power pin (+5V). We took the end of a spare pwm cable and connected just one pin on the +5V pin of Digital In/Out #1, then ran that into the TTL<>RS232 board.

[Joel J]

Here is what I did to allow compatibility between years, and between IDEs. I have a "core" directory, a "modules" directory, a "2005" directory and a "visual studio" directory. I put all c/h files I personally make (or get from Kevin's website) into the modules directory, and I put all of the files supplied by IFI's 2006 default code into the "core" directory. I copied the 2006 default code's workspace files to the "2005" directory, followed the instructions to ensure compatibility with '05 controllers, and then included all my modules from the "modules" directory, and the relevant IFI code from the "core" directory into that project. I didn't trust the pic's being #define'd, so I edited the project settings in the "2005" directory to include the definition of VERSION_2005_CODE (-D_VERSION_2005_CODE), and then I test for the presence of this macro where relevant. For example, in Kevin's EEPROM code:

// Need to include a different file in the event of a compilation for the 
// '04-'05 controller.
#ifdef VERSION_2005_CODE			
	#include "ifi_picdefs.h"
#else
	#include <p18f8722.h>
#endif

#include "rkw_eeprom.h"

So now, for me, switching between the two platforms is a matter of opening the right workspace, and rebuilding all files.

I also created a visual studio solution, and stored it to the "Visual Studio" directory. I include all the c/h files into this solution, and do all my coding there. When its time to compile, I alt+tab to the MPLab editor (it senses that the files have changed at this point), and pretend as though I had been coding there all along.

c:\2006 Robot Code\
c:\2006 Robot Code\modules\
c:\2006 Robot Code\2005 Workspace\
c:\2006 Robot Code\Visual Studio\
c:\2006 Robot Code\documentation\

That might help you in the long run, though it doesnt solve your current problem.

[tribotec_ca88]

Originally Posted by Kyveck

I have spent quite a bit of time trying to get this to work, and I suspect some kind of configuration problem. I'm hoping that if I list everything, someone might be able to see what's wrong.

frc_camera_s, using compiled .hex that was in the download
CMUcam2 connected via TTL port to RC
2006 RC tethered to 2006 OI
tilt/pan servos connected to RC pwm1,2
RC pwm to CMUcam2 power
jumpers on pan reverse, oscillator connect, internal power to servos
DB-9 serial from RC program port to laptop via USB to Serial converter
12 volt and 7.2 volt batteries fully charged

The camera shows a green light, and all of RC lights are green like normal. The camera holds its default position, and does not respond to any light. Nothing comes up over the terminal, although I expected to see "Scanning...". With the camera configured stand-alone (DB-9 to USB to laptop, servos plugged directly in to camera, power still supplied by RC) I was able to use LabView and the CMUcam2 demo program to calibrate the camera well. It tracked as I believe it was supposed to without any problems.

Any assistance would be appreciated. Thanks.
Dublin Robotics, Team 1014 (with the Bad Robot t-shirts)

Unfortunately we seem to be having the same problem...we don't get any errors after compiling the program, and the results shown on MPLAB clearly state "BUILD SUCCEEDED" ... no problems loading the .hex file from IFI loader to the controller, however, the COM Terminal window is BLANK! Nothing happens or shows up on the window... any theories of what may be going on?
Any help at all would be greatly appreciated...

[Ryan O]

Originally Posted by Kevin Watson

"Using_Last_Years_FRC.txt" included with my code as well as IFI's.

-Kevin

Thanks alot, this should solve the problem. Good luck this year.

[tribotec_ca88]

Ah! yay, we got it up and running...

[Validius]

Originally Posted by Kyveck

I ran the serial port diagnostic tests. In loopback mode, terminal input was not echoed. When connected to the camera, there was no responce. I am confident that I configured everything per the ReadMe file.

When I ran debug, the following was printed to the terminal:
Camera: Initialized abnormally with code 131
Camera: Initialization state = 1
Camera: Initialization state = 2
Camera: Initialization state = 3

I hope to test the TTL to RS232 board soon using the ifi instructions.
Thank you for your help.
Kyle W

Has anyone responded to this? I'm getting the same error code.

[shobuda999]

Originally Posted by Validius

Has anyone responded to this? I'm getting the same error code.

When this happened to us, it was because we had the DEBUG flag turned on. Try turning it off and see what happens!

[Alan Anderson]

The "code 131" error says that the program did not receive a response from the camera during initialization. It could be a communication problem (try the loopback test, and make sure you don't have the PC still connected to the camera at the same time as the RC), a power problem (make sure you have a freshly charged battery and all the connections are good), or a broken camera (test it with the LabView application).

That's about all I can suggest without looking at how you have things wired up.

[Ryan O]

Ok, so I got the code scanning for the target and giving me angles and such- the only problem is that it gives me the info then keeps searching. Is it supposed to do this? and if not what can I do to get it the stop when it finds it? Your help so far has made this much easier. Thanks alot!

[Ryan O]

Ok, I have a more specific idea of the problem now: the camera is aquiring during the stage where it is moving to restart search, then doesn't do a regular scan at that height. Any suggestions?