Re: New C18 3.0+ Compatible FRC Code
 

Just grabbed both 2.4 versions and opened them up with MPLAB 7.62. I only had two issues.

1) The ifi_frc_8722_25.lib file was being displayed as "Not Found" even though it was in the root directory of the project along with all the other files. I just removed the "Not Found" version and added it back in the project.

2) Error during linking saying that the CLIB.LIB wasn't found. It was looking in c:\program files\mplab\mcc18\lib instead of c:\mcc18\lib. Very likely a configuration mistake on my part here. I just went in to the project properties and added the correct library path.

Compiled and linked without any errors.

Nice work Kevin.

Sean

Re: New C18 3.0+ Compatible FRC Code
 

I just verified the 2008 KOP gyro is 12.5 mV per degree/second, which is the same as the ADXRS150. The datasheet is attached.

-Kevin

Re: New C18 3.0+ Compatible FRC Code
 

Thank you very much. I'll hopefully be able to provide gyro and encoder code feedback soon.

Re: New C18 3.0+ Compatible FRC Code
 

I've finally been able to compile the 3.1 code, so I started looking over it. It seems that it's a lot cleaner than the past few years' code, but I'm used to those, so I've got a few questions. In the past few years we just put our processing of inputs on the joysticks in Default_Routine(). (This is probably not the best way to have done it, but oh well) I'm assuming we should put all of this code such as using the buttons and stuff into Teleop()? Thanks for the help, great code!

-Arty

Re: New C18 3.0+ Compatible FRC Code
 

Kevin:

In your beta code 3 you have two definition for timer 4, one in the timers.c with its ISR in interrupts.c and a 2nd in the adc.c code again with its ISR...

this maybe confusing for user...

thanks...

ps: there are a few typo still in ADC code, ie label is timer 2, but your code is timer 4...

Re: New C18 3.0+ Compatible FRC Code
 

I just spoke with the programming team tonight. We had to decide between EasyC Pro, MPLAB and IFI default code and the newer MPLAB and compiler with Kevin's new build.
I am happy to say, 2073 will be stretching their skills and using the new code and compiler.:D

Now I have a dilemma.
The team does not have access to a laptop, plenty of computers, but no laptop. So, they will have nothing to take with them to competition. I will not be able to join them at our regional either because of a previously arranged vacation in Hawaii. :yikes:
I do have a laptop I am willing to let them borrow, but it currently has my personal copy of MPLAB 7.2 and C18 v2.4 on it that I use with my personal Vex kit. I prefer not to uninstall it if I can avoid it. So, the question is, can MPLAB's 7.2 and 8.1, and C18's v2.4 and v3.1 co-reside on the same system?

Re: New C18 3.0+ Compatible FRC Code
 

There's a method to my madness (honest). The new scheme standardizes the ISR names across different code modules to make it easier for me to write code that just drops into someone's project without them having to enter the ISR code directly into their code. The downside is that they need to enable two #defines instead of one (see comments at the top of ifi_frc.h, timers.h and interrupts.h). There is also the possibility that two ISRs with the same name will be enabled, but the compiler will complain. I have another solution that uses linked-list ISRs and self modifying code, but I decided it might become a support nightmare, so I'm not going to release it.

Thanks. I'll go code splunking and fix 'em

-Kevin

Re: New C18 3.0+ Compatible FRC Code
 

Yep. You can also call IFI's old default_routine(), which has been moved to ifi_code.c.

-Kevin

Re: New C18 3.0+ Compatible FRC Code
 

I installed 2.4 and then made a copy of the entire mcc18 directory. I then applied the 3.10 upgrade to the original 2.4 directory. Now I can use both compilers and all I have to do is make sure the correct directory is named "mcc18" at build time (thanks go to Mark at IFI for pointing this out to me).

-Kevin

Re: New C18 3.0+ Compatible FRC Code
 

Excellent! Got it running!!
Next question. You currently have 3 versions for 3.0: Basic, Basic + Gyro, Basic + encoders.
To add the encoders to the basic + gyro, do I just add the encoder.h/c files to the project? Then follow your instructions for adding your encoder code from an earlier build, like 2007's ifi_encoder.zip?

Re: New C18 3.0+ Compatible FRC Code
 

It should be a matter of:
1) Adding encoder.c/.h to the project.
2) Putting #include "encoder.h" at the top of each source file where any encoder functions will be called.
3) Calling the encoder initialization function(s) from teleop.c/Initialization().
4) Enable the individual encoder channels at the top of encoder.h.
5) Assign the encoder phase b signals to digital inputs at the top of encoder.h.
6) Enable the interrupt(s) associated with each enabled encoder channel at the top of ifi_frc.h.
7) Make sure there are no conflicting ISRs enabled at the top of interrupts.h.
8) Compile and test.

-Kevin

Re: New C18 3.0+ Compatible FRC Code
 

Kevin,

Did anything turn up on your inspection of the provided
code with respect to a possible source of the 8.3 volt bug?

With all the interrupts we will have flying this year,
we are sweating it, although we will get a good six
weeks to see if it surfaces again.

Eugene

Re: New C18 3.0+ Compatible FRC Code
 

While testing the new library code, I discovered that I could induce the 8.3 volt bug pretty reliably (I had never seen it before), and more importantly, eliminate it by initializing the entire RAM array to zero at boot time (something the C18 startup code does not do and IFI only partially does). I need to get in with a hardware debugger to see what's going on, but I haven't had time to build the custom cable I need to use the internal debugging header. I hope to get to it sometime in the next week.

If it's of any interest, I did notice the new code and compiler seem to handle high interrupt loads better than with the 2.4 compiler. I setup my ADC code to sample at 6400Hz (each sample generates an interrupt) while generating a few hundred more interrupts/sec with a encoder. The RC had no problems sending a bunch of telemetry using printf() under the load.

-Kevin

Re: New C18 3.0+ Compatible FRC Code
 

Although Kevin already responded, I felt I should say how I did it. I found a freeware diff utility for windows and compared the two folders, and I used that to see the changes and add one to the other. It worked pretty flawlessly and also helped me integrate the two test teleop() routines.

Re: New C18 3.0+ Compatible FRC Code
 

Awesome, thanks a lot Kevin!