DT_INTS for K42

[mpgmike]

Although a bit dated, my first recommendation is "Applying PIC18 Microcontrollers", by Barry B. Brey; Copyright 2008 by Pearson Education. ISBN-13: 978-0-13-088546-3.

Another recommendation is "PIC Microcontrollers; An Introduction to Microelectronics", by Martin Bates; Copyright 2011 by Martin Bates. ISBN: 978-0-08-096911-4.

[dtbarber]

Thanks for the suggested references.

[tumbleweed]

Same problem in asm, so if you can figure out how an isr works you can do it in pbp as well

[mpgmike]

Conventional wisdom states that ISRs should do only minimalistic functions. Never call a subroutine from an interrupt handler. Never use PAUSE type statements. In short, turn something on or off, or set a flag you can poll and service in your main loop. With that said, most ISRs can be crafted in ASM with only basic BSF/BCF and perhaps MOVLW/MOVF type commands. Leave the heavy lifting for a subroutine you write in PBP.

[dtbarber]

Thanks for the additional comments. At this point, all I want is to understand how to write two ISR for a PIC18F27K42 in which upon an overflow interrupt from Timer3 or Timer1, the appropriate ISR increments a counter, clears the interrupt flag, and returns back to the executing PBP subroutine. The PBP routine uses one overflow count to compute a time period . The other overflow count is used to timeout the subroutine . The application is a tachometer that measures RPM from 0 to 12K. My code works with a PIC18F26K22 and DT_INTS-18. I have tried polling the interrupt flags within the PBP subroutine, but at low RPMs and clocking the PIC at 64MHz the results are erratic.

Yes, I could just keep using an 18F26K22, but why not learn something new?

[richard]

the k42 series smt module in gated counter mode would be an ideal thing for a tacho and probably a simpler set of interrupts to manage.great lets go.
its at this time you realize that pbp has no support for any pic hardware modules that did not exist at least 15 years ago, the last time pbp had a new command the space shuttles still flew.
so if you want to take advantage of the newer pic's powerful built in stuff you need to learn a whole lot of asm
stuff to prop it up. sadly pbp has become a real programming cul de sac. I put my time and effort into a better method mplabx and C and the wonderful MCC [microchip code configurator]. The real beauty of C is its cross platform compatibility, it opens up a whole new world.

but why not learn something new?

that's my new

[Ioannis]

Cannot disagree with you Richard... But either I am too old or C is too twisted that I have hard time doing useful things with it... Sadly.

Ioannis

[mpgmike]

I agree with Richard in that I also have learned to work with C in MPLABX, but I was able to use PBP command structures to use newer functions that don't have PBP commands. I didn't need to resort to ASM.

As for clearing an Interrupt Flag and incrementing a variable, it would look something like this in ASM:

BANKSEL PIR1  ;or whatever holds the TIMERx IF
BCF PIR1,1  ;or whatever bit holds the TIMERxIF.  Note use of a comma versus period
BANKSEL RPM  ;or whatever your variable is called
INCF RPM  ;increment your variable by 1
RETFIE

The above does not take into account Context Saving, nor does it take into account your RPM variable is probably at least 16 bits. However, the skeletal structure is about all you need for an ASM Interrupt Service Handler.

[dtbarber]

Thanks for the framework for implementing an ISR in PBP.

I concur with the comments regarding moving to C and using Microchip's IDE. I have started down that path on numerous occasions, but overcoming the inertia of being able to use PBP to easily do what I wanted was too great of a barrier for me. While it can be laborious to use PBP to implement some of the newer PIC peripherals, it can be done. For me, PBP's problem is that the language does not effectively deal with interrupts. As a friend of mine says "You do not fix a problem. You eliminate it." Given that it appears that PBP has reached its maturity, it is time for me to eliminate my problem an learn C. Painful.