Timer in autoload mode

void delay()
{
  TMOD = 0X20; /*Timer 1 operation in mode 2*/
  TH1 = -184;
  TR1 = 1;
  while(TF1==0);/*Polling mode is used instead of interupt mode*/
  TR1 = 0;
  TF1 = 0;
}

It works in autoload mode and the TL1 starts intially from 00 and later gets

loaded by the value in TH1 register.The problem is that it doesnt leave the delay routine

to go back to main program again

how can it 'work' when "The problem is "

for the first delay you depend on whatever is in TL1 for your delay no autoload takes place before the timer roll over. for the next delay you will get the autoload value plus whetever counts happened before TR1=0. You are confusing the whole issue by using autoload. With a lot of feniggeling I have no doubt this could be done with autoload but why use auotoload when only one run in the routine. Using illogical modes does nothing but make the issue confusing, complex. less maintainable, less understandable, and error prone.

why not just straight code

void delay()
{
  TMOD = 0X10; /*Timer 1 operation in mode 1*/
  TH1 =  0xff;
  TL1 = -184;
  TR1 = 1;
  while(TF1==0);/*Polling mode is used instead of interupt mode*/
  TR1 = 0;
  TF1 = 0;
}

Erik

void delay()
{
  TMOD = 0X20; /*Timer 1 operation in mode 2*/
  TH1 = -184;
  TR1 = 1;
  while(TF1==0);/*Polling mode is used instead of interupt mode*/
  TR1 = 0;
  TF1 = 0;
}

It works in autoload mode and the TL1 starts intially from 00 and later gets

loaded by the value in TH1 register.The problem is that it doesnt leave the delay routine

to go back to main program again

how can it 'work' when "The problem is "

for the first delay you depend on whatever is in TL1 for your delay no autoload takes place before the timer roll over. for the next delay you will get the autoload value plus whetever counts happened before TR1=0. You are confusing the whole issue by using autoload. With a lot of feniggeling I have no doubt this could be done with autoload but why use auotoload when only one run in the routine. Using illogical modes does nothing but make the issue confusing, complex. less maintainable, less understandable, and error prone.

why not just straight code

void delay()
{
  TMOD = 0X10; /*Timer 1 operation in mode 1*/
  TH1 =  0xff;
  TL1 = -184;
  TR1 = 1;
  while(TF1==0);/*Polling mode is used instead of interupt mode*/
  TR1 = 0;
  TF1 = 0;
}

Erik

"why not just straight code"

it doesn't fundamentally solve his problem - that the code doesn't return (for some extremely strange reason, in my view) to main().

your code will have the exact same "problem" - if the code didn't return from his "while (TF1==0);", it wouldn't return from your "while (TF1==0);". thus why i called your line of thoughts "secondary".

at this point, I am inclined to think this is a user (observational) problem.

Erik,

I am beginning to suspect that Ashley is Jack Sprat's mother in law...

your code will have the exact same "problem" - if the code didn't return from his "while (TF1==0);", it wouldn't return from your "while (TF1==0);". thus why i called your line of thoughts "secondary".

TR1 = 1;
  while(TF1==0);/*Polling mode is used instead of interupt mode*/

will continue, a running timer (0 or 1) will (eventually) set TF to 1.

Erik

Are we sure that we have the full source code?

It was quite some time since I used an 8051 but isn't TF1 auto-cleared if interrupts are used with the timer, while manual-clear without interrupts? So a program that does enable an interrupt handler for this timer but just have an empty ISR would get an interesting competition between the interrupt logic of the chip and the TF1 polling code.

'will continue, a running timer (0 or 1) will (eventually) set TF to 1.'

that's the point of the entire discussion, erik: his code, as well as yours, SHOULD work. but he reported it didn't work.

both of you used exactly the same return mechanism "while (TFn==0);". so if his code, for whatever reason, didn't return from that, your code wouldn't return from that either. using mode 1 vs. 2 does NOT matter at all from that perspective.

jesus! is it really that unreasonable to expect people to have some minimum reading comprehension?

"I am beginning to suspect that Ashley is Jack Sprat's mother in law..."

tammy, as long as i am not YOUR mother or mother in law, i'm ok, :)

Tamir;
Ashley is a guy or a shim. He may be Jack Sprat.
Bradford

"He may be Jack Sprat."

Or maybe even a Zeusti.

But he/she definitely also hides behind the names Qili, Millwood, and fdan00 on other forums to name a few.

'will continue, a running timer (0 or 1) will (eventually) set TF to 1.'

that's the point of the entire discussion, erik: his code, as well as yours, SHOULD work. but he reported it didn't work.

it does work, I have a sneaky suspicion that autoload mode confuses the simulator. Not the first time the simulator has 'simulated' instead of 'replicated'. If it confuses the chip the chip is defective.

why on earth not just see what happens with non-autoload the OP has it right at his fingertips and trying it instead relying on bile from the homewrecker would take, at most 5 minutes.

Erik