Hi, I'm seeing something unexpected in the running time of a Timer2 based delay loop. With the combined prescale, period, and postscale, should be setting TMR2IF at 500Hz? I expect the polling loop below to fall out after approx 2ms. But it's taking closer to 3.8ms* and I can't think of an explanation. Part is PIC18F6680, Fosc = 40 MHz. ; THIS DOESN'T WORK... ; Timer2 input = Fosc/4 = 10,000,000 Hz ; divide by 1/16 prescale = 625,000 Hz ; divide by period (156) = ~4006 Hz ; divide by 1/8 postscale = ~500 Hz = approx 2ms! delay2ms: ; set up Timer2 movlw b'01110010' ; 1:8 postscale,timer off,1:16 prescale movwf T2CON clrf TMR2 movlw 156 ; 156 x 8 x 16 cycles = ~1996.8us = approx 2ms movwf PR2 bcf PIR1,TMR2IF bsf T2CON,TMR2ON ; start it! btfss PIR1,TMR2IF ; check for timer overflow bra $-2 return *The running time is measured using Timer0, and I have confidence in the correctness of that timer, having verified expected durations for sanity with simple cycle wasting loops.
PIC18 Timer2 weirdness
Started by ●October 29, 2004
Reply by ●October 29, 20042004-10-29
Reply by ●October 29, 20042004-10-29
On 29 Oct 2004 19:06:15 GMT, the renowned garykato@aol.com (Gary Kato) wrote:>> movlw b'01110010' ; 1:8 postscale,timer off,1:16 prescale >> > >Your postscale bits are wrong.Yup. Specifically, you're getting 1:15 rather than 1:8. Best regards, Spehro Pefhany -- "it's the network..." "The Journey is the reward" speff@interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com
Reply by ●October 30, 20042004-10-30
toby@telegraphics.com.au (Toby Thain) wrote:>Hi, > >I'm seeing something unexpected in the running time of a Timer2 based >delay loop. With the combined prescale, period, and postscale, should >be setting TMR2IF at 500Hz? I expect the polling loop below to fall >out after approx 2ms. But it's taking closer to 3.8ms* and I can't >think of an explanation. > >Part is PIC18F6680, Fosc = 40 MHz. > >; THIS DOESN'T WORK... >; Timer2 input = Fosc/4 = 10,000,000 Hz >; divide by 1/16 prescale = 625,000 Hz >; divide by period (156) = ~4006 Hz >; divide by 1/8 postscale = ~500 Hz = approx 2ms! >delay2ms: ; set up Timer2 > movlw b'01110010' ; 1:8 postscale,timer off,1:16 prescale > movwf T2CON > clrf TMR2 > movlw 156 ; 156 x 8 x 16 cycles = ~1996.8us = approx 2ms > movwf PR2 > bcf PIR1,TMR2IF > bsf T2CON,TMR2ON ; start it! > btfss PIR1,TMR2IF ; check for timer overflow > bra $-2 > return > >*The running time is measured using Timer0, and I have confidence in >the correctness of that timer, having verified expected durations for >sanity with simple cycle wasting loops.Others have pointed out the postscale problem. For future reference, here's an Excel spreadsheet for working out those pre/post-scaler issues: http://dhenry.home.sprynet.com/pic_tmr2.xls -- Dan Henry
Reply by ●October 30, 20042004-10-30
Spehro Pefhany <speffSNIP@interlogDOTyou.knowwhat> wrote in message news:<dp75o01gunm95db0s2ioo4thm06fcmu6qv@4ax.com>...> On 29 Oct 2004 19:06:15 GMT, the renowned garykato@aol.com (Gary Kato) > wrote: > > >> movlw b'01110010' ; 1:8 postscale,timer off,1:16 prescale > >> > > > >Your postscale bits are wrong. > > Yup. Specifically, you're getting 1:15 rather than 1:8.Yes, I just found this out too. What a fool. Thanks! :-) --T> > > Best regards, > Spehro Pefhany
Reply by ●October 30, 20042004-10-30
Dan Henry <dhenry@sprynet.com> wrote:>toby@telegraphics.com.au (Toby Thain) wrote: > >>Hi, >> >>I'm seeing something unexpected in the running time of a Timer2 based >>delay loop. With the combined prescale, period, and postscale, should >>be setting TMR2IF at 500Hz? I expect the polling loop below to fall >>out after approx 2ms. But it's taking closer to 3.8ms* and I can't >>think of an explanation. >> >>Part is PIC18F6680, Fosc = 40 MHz. >> >>; THIS DOESN'T WORK... >>; Timer2 input = Fosc/4 = 10,000,000 Hz >>; divide by 1/16 prescale = 625,000 Hz >>; divide by period (156) = ~4006 Hz >>; divide by 1/8 postscale = ~500 Hz = approx 2ms! >>delay2ms: ; set up Timer2 >> movlw b'01110010' ; 1:8 postscale,timer off,1:16 prescale >> movwf T2CON >> clrf TMR2 >> movlw 156 ; 156 x 8 x 16 cycles = ~1996.8us = approx 2ms >> movwf PR2 >> bcf PIR1,TMR2IF >> bsf T2CON,TMR2ON ; start it! >> btfss PIR1,TMR2IF ; check for timer overflow >> bra $-2 >> return >> >>*The running time is measured using Timer0, and I have confidence in >>the correctness of that timer, having verified expected durations for >>sanity with simple cycle wasting loops. > >Others have pointed out the postscale problem. For future reference, >here's an Excel spreadsheet for working out those pre/post-scaler >issues: > >http://dhenry.home.sprynet.com/pic_tmr2.xlsPlus using that spreadsheet, you'll quickly discover that 1:16 pre/1:8 post with PR2 = 156 only gets you close to 500Hz and that 1:16/1:10 PR2 = 125 is spot-on. -- Dan Henry
Reply by ●October 30, 20042004-10-30
Dan Henry <dhenry@sprynet.com> wrote in message news:<tmb6o0d6h8a3se80be3jrvdq4imtmhos3j@4ax.com>...> Dan Henry <dhenry@sprynet.com> wrote: > > >... > >Others have pointed out the postscale problem. For future reference, > >here's an Excel spreadsheet for working out those pre/post-scaler > >issues: > > > >http://dhenry.home.sprynet.com/pic_tmr2.xls > > Plus using that spreadsheet, you'll quickly discover that 1:16 pre/1:8 > post with PR2 = 156 only gets you close to 500Hz and that 1:16/1:10 > PR2 = 125 is spot-on.Yes, that imprecision is also due to my misreading of the manual; I overlooked that Timer2 actually allows all divisors 2-16 (more familiar with the other timer modules that only have powers-of-2 scales). You know how it is when you stare at something too long! :) --T
Reply by ●October 30, 20042004-10-30
On 30 Oct 2004 09:58:33 -0700, the renowned toby@telegraphics.com.au (Toby Thain) wrote:>Dan Henry <dhenry@sprynet.com> wrote in message news:<tmb6o0d6h8a3se80be3jrvdq4imtmhos3j@4ax.com>... >> Dan Henry <dhenry@sprynet.com> wrote: >> >> >... >> >Others have pointed out the postscale problem. For future reference, >> >here's an Excel spreadsheet for working out those pre/post-scaler >> >issues: >> > >> >http://dhenry.home.sprynet.com/pic_tmr2.xls >> >> Plus using that spreadsheet, you'll quickly discover that 1:16 pre/1:8 >> post with PR2 = 156 only gets you close to 500Hz and that 1:16/1:10 >> PR2 = 125 is spot-on. > >Yes, that imprecision is also due to my misreading of the manual; I >overlooked that Timer2 actually allows all divisors 2-16 (more >familiar with the other timer modules that only have powers-of-2 >scales). You know how it is when you stare at something too long! :) >--TThe comments that indicate what you *wanted* the code to do can impede troubleshooting. Maybe they should all left off 'til the end. ;-) ;-) Best regards, Spehro Pefhany -- "it's the network..." "The Journey is the reward" speff@interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com
Reply by ●October 30, 20042004-10-30
On 30 Oct, in article
<d6ce4a6c.0410300858.4a756b75@posting.google.com>
toby@telegraphics.com.au "Toby Thain" wrote:
....
>Yes, that imprecision is also due to my misreading of the manual; I
>overlooked that Timer2 actually allows all divisors 2-16 (more
>familiar with the other timer modules that only have powers-of-2
>scales). You know how it is when you stare at something too long! :)
It stares back and makes silly faces at you :-^
--
Paul Carpenter | paul@pcserviceselectronics.co.uk
<http://www.pcserviceselectronics.co.uk/> PC Services
<http://www.gnuh8.org.uk/> GNU H8 & mailing list info
<http://www.badweb.org.uk/> For those web sites you hate
