cycle length to do various tasks

[Darrel Taylor]

So you are using 2 timers to make 1 pulse?

Well, I can say that interrupt latency can be accounted for by ADDING a reload value to the accumulated timer value.
And nothing in your ISR is using PBP's system variables, so the handler "Type" can be ASM, which will reduce the latency to be accounted for.

I won't go into too much detail because I think you're probably doing it the wrong way.

What are the minimum and maximum pulse widths?
And what is the interval between them?

[Lestat]

Darrel

I am effectivly trying to produce a PWM signal to control a servo, I know there is lots of examples out there already, but I want to use the timers to control the time the pulse is high using tmr1 and tmr2 tells the PIC when to restart the high pulse, allows me to learn how the timers work and how to accurately control timer lengths.

As this is a PWM signal for a servo, TMR1 produces a high pulse between 1 and 2ms and tmr2 restarts tmr1 every 20ms.

I am still new to the whole world of programming and PIC Basic, but what variables are PBP system variables?

Thank you for the interrupt code, it really has made setting up interrupts without using oninterrupt a dream.

What I am interested in how many cycles it takes to set up the timer registers and also how long it takes for an interrupt to trigger? or alternatively a method to find out how long it takes to run the various bits of code?

Many thanks for your help.

[Darrel Taylor]

I thought it might be for a servo.

First, take a look at this program I did many years ago ...
http://www.pbpgroup.com/modules/wfse...hp?articleid=2

It uses a single timer for both the on and off times.

As for the execution times for various statements, there is no one answer.
It depends on what size the variables are, what bank the registers are in, and where in the code the statement is located.
A statement like A = B can take anywhere from 2 to 8 assembly language instructions. Which at 4mhz would be 2 to 8uS.

For getting in and out of the interrupt, the differences are even greater depending on the type of handler used.

The PWPS program accounts for those times by adding the reload values to the accumulated timer values.
It's ISR is strickly assembly language, but the same thing can be done with DT_INTS.

[Lestat]

That program was an interesting read thank you. It certainly makes me think I should consider learning some assembly code, to allow me to control the number of cycles of time critical programs and also ensure that variables are saved in certain banks so the PIC does not need to spend time jumping from one bank to the other when running a critical task.

For the time being I will stick with DT-INTS as I believe writing my own assembly level interrupt is seriously beyond me. Maybe something to try in a few years.

I think I'll spend sometime adjusting the numbers being loaded into the various timer registers and see if I can learn some assembly code.

Thanks for your help