#use delay and OSC config

[eabir]

Hello ,

There is something I don't understand regard the following:

When I set: #use delay(clock=4000000) then the code (routines) go less faster then if I set: #use delay(clock=1000000).

How the 1Mhz is faster then 4Mhz ?

Thanks

[Ttelmah]

It isn't.

The clock statement, doesn't change the physical clock speed (exception here if using the internal oscillator, where the clock setting will change this as well). It tells the compiler _what clock speed the processor is working at_.

If you have a processor that executed 1 million instructions per second (clock=4000000), and then use 'delay_ms(1000)', the compiler will generate code to perform one million instructions, to give exactly the delay you ask for. One second.
However if instead you tell the compiler that your processor only executes 250000 instructions per second, and then use the same delay statement, the compiler will generate code to produce 250000 instructions. If you then feed this code into a processor that is really running at 1MIPS, then it'll run four times faster than it should. problem then is that if the code contains delays that _must_ be certain durations to work, they will now run faster than they should. The processor doesn't run faster, but for delays, the compiler now generates shorter code.

You should set the clock speed to the actual speed your chip is running at, otherwise it'll make working times out _very_ hard, and prevent some of the libraries (like LCD ones, or the RS232 statements, which _depend_ on this value being right), from working. If you want the code to run faster, and have delays in the code, shorten them. Delay_ms(250), with a processor running at 1MIPS, will produce the same 250000 instruction code, as using the wrong clock statement...

Best Wishes

[eabir]

hi again,

Ok, how do I know how many MIPS are in the PIC10F222 ? (I assume MIPS is fixed per chip).

Is this line is enough for declaring the frequency ? : "#use delay(clock=4000000)".

Thanks

[PCM programmer]