-
1 Attachment(s)
PIC12F675 trouble
Hey all,
i am still struggling with what seems like eratic behavior with my PIC12F675.
I am trying to build an IR detection array thats smarter than a 555 timer or another hardwired 38500 oscillator and figured the small and tidy footprint of a 12F675 would be ideal.
right now, i am at the stage where i am trying to get IR LED to flash at 38500MHz for about 10ms. it seems to work fine, but when i add some "smarts" code, the whole thing stops working. here is the following code i am adding.
' ... Problem is between here
if (leftcount = duration) then
for i = 1 to 3
high ledsignal
pause 500
low ledsignal
pause 500
next i
low ledsignal
endif
' ... and here
to my mainline, the chip wont start.
the same happens if i add:
freqout gpio.3, 10, 38500
instead of the ASM delay block. I went down the ASM block path because it seems that freqout will not work with the 12F675 though i can't find that explicitly stated anywhere.
Bottom Line Issue:
~~~~~~~~~~~
It seems that when i add code that the chip does not like, it won't even start - the status LED blink i have at the top of my code won't even activate etc - nothing happens.
I am using a USB melabs programmer. i disconnect the USB from my computer between flashes to make sure i don't fry anything. i am using MicroCode studio 2.3.0.0 and PicBasic PRO compiler 2.46.
everyhing is soldered together to eliminate any potential of bad connections. i am using the following parts (http://www.hobbyengineering.com/H1765.html, http://www.hobbyengineering.com/H1992.html) without the 4Mhz resonator installed on the prototype board.
gpio.0 is a regular old red LED with a 220 ohm resistor on it.
gpio.3 is the detect from the IR Receiver (http://www.parallax.com/detail.asp?product_id=350-00014)
gpio.5 is the IR LED out with a 1K resistor (http://www.parallax.com/detail.asp?product_id=350-00017)
i have a .1uf bypass capacitor on the +5v and Ground leads of the detector
any thoughts or suggestions on what i am missing would be appreciated. probably something simple and silly.
it sure is frustrating. with that if block in there, the thing won't even start, i yank it out and it seems to work fine again.
is there a possibility that, since i am using the internal oscillator, i have overwritten some config (calibration) space or something that ensures the internal oscillator runs as it should? This question is nagging at me in the back of my mind.
here is the full code listing:
' Defines that set up the PIC
@ DEVICE pic12F675, INTRC_OSC_NOCLKOUT ' Internal Oscillator
@ DEVICE pic12F675, WDT_ON ' Watchdog Timer
@ DEVICE pic12F675, PWRT_ON ' Power-On Timer
@ DEVICE pic12F675, MCLR_OFF ' Master Clear Options set to Internal
@ DEVICE pic12F675, BOD_ON ' Brown-Out Detect
@ DEVICE pic12F675, PROTECT_OFF ' Code Protect off
@ DEVICE pic12F675, CPD_OFF ' Data EEPROM Protect off
'define xtal 4 ' don't think i need this - cant find it in the .inc
define OSC 4
' Initialise my Registers
TRISIO=%00001000 ' all output except gpio.3 is input
CMCON=%00000111 ' comparator module reg 6-1 page 35. Disable all
ANSEL=%00000000 ' A/D config reg 7-2. Set all pins to digital
'ADCON0=%00000000 ' reg 7-1 not required but shuts off A/D control
pause 1000
' Degine PINs
ledSignal var GPIO.0
leftIRLed var GPIO.5
irDetect var GPIO.3
' Constants
duration con 100
' Variables
i var word ' Genral purpose counter
dur var word
leftCount var word
' Code Mainline
' Blink the status led so we know we're ok.
high ledSignal
pause 2000
low ledSignal
pause 2000
'Mainline loop....
MainLoop:
' Check to see if we have activity on the left LED.
gosub LeftIR
' ... Problem is between here
if (leftcount = duration) then
for i = 1 to 3
high ledsignal
pause 500
low ledsignal
pause 500
next i
low ledsignal
endif
' ... and here
pause 20
Goto MainLoop
End
'******************************
'
' Subroutines ...
'
'******************************
'1/38500 = 0.000026
LeftIR:
leftCount = 0
dur = duration
asm
cycleLeft
BSF GPIO,5 ;1uS Begin HIGH duty cycle
NOP ;2uS each nop is 1uS long
NOP ;3uS
NOP ;4uS
NOP ;5uS
NOP ;6uS
NOP ;7uS
NOP ;8uS
NOP ;9uS
NOP ;10uS
NOP ;11uS
NOP ;12uS
NOP ;13uS
BCF GPIO,5 ;1uS Bigin Low
NOP ;14uS
NOP ;15uS
NOP ;16uS
NOP ;17uS
NOP ;18uS
NOP ;19uS
NOP ;20uS
NOP ;21uS
NOP ;22uS
BTFSS _leftCount ;23,us <- this line causes a problem it seems too.
decfsz _dur, 1 ;24uS
goto cycleLeft ;26uS
endasm
ledSignal = ~GPIO.3
return
-
Hi Russ,
EDIT - on this line of code...
BTFSS _leftCount ;23,us <- this line causes a problem it seems too.
You did not specify which BIT to test - it's also a WORD variable - maybe someone here can answer if BTFSS will work on a WORD??
OOpps - I thought the 12F's WERE 12-bit core - sorry!
EndEDIT
====
I have not looked over the rest - but here's a list of unsupported commands fort the 12bit cores:
from PBP 2.44 'new.txt' file
Commands that are not supported for the 12-bit core PICmicro MCUs:
Command Reason
Adcin No internal ADCs
Data No on-chip EEPROM
Dtmfout Not enough RAM or stack
Eeprom No on-chip EEPROM
Erasecode No flash devices
Freqout Not enough RAM or stack
Hpwm No hardware PWM
Hserin No hardware serial port
Hserin2 No hardware serial port
Hserout No hardware serial port
Hserout2 No hardware serial port
On Interrupt No Interrupts
Read No on-chip EEPROM
Readcode No flash devices
Resume No Interrupts
Serin2 Not enough RAM or stack
Serout2 Not enough RAM or stack
USBIn No USB devices
USBInit No USB devices
USBOut No USB devices
Write No on-chip EEPROM
Writecode No flash devices
Xin Not enough RAM or stack
Xout Not enough RAM or stack
-
Hi Russ,
In the MainLoop, the IF (leftcount = duration) will never evaluate true.
duration is a constant(100), and leftcount never gets assigned anything but 0. Maybe you were thinking that BTFSS would do something, but that means "Bit Test File Skip if Set, so it's looking for a bit number that wasn't specified. Are you trying to test leftcount for a specific value, or trying to set the value?
Archilochus,
The 12F675's are actually 14-bit core's. I have no idea why they called them 12's. 12C508's are 12-bit, and the 12F508's are 12-bit, so you would think 12F629/675/683's are too. but, Not!
<br>
-
btfss is for testing individual "bits" in a file register.
You would use btfss _leftCount,1 to test bit position 1 in leftCount. This
seems like an unnecessary test at any rate since you're clearing leftCount
at the entry point to your asm routine, and leftCount will always be zero.
Also be aware that you're only accessing the low bytes in your leftCount
and dur word variables the way you're when doing it.
To access the high byte in dur you would use decfsz _dur+1,f. Dur is a pointer
to the low byte. Adding _dur+1 moves the pointer to the next byte location
in ram.
-
awesome information - i love it when there is a simple answer.
thx for the tips and for pointing out my issue. i will try to fix up what you guys have pointed out this evening and retry. also looks like new.txt is worth a closer look.
any idea why the PIC seems to just hang? I would expect that it would at least toggle the statusLED for me that i've set up before it gets to a piece of bad code. or, am i to expect "unexpected" behavior anytime i muck up the logic anywhere in the program?
russ
-
awesome information - i love it when there is a simple answer. i had actually meant that line to read:
btfss gpio, 3;
incfsz _leftCount, 1 ;23,us - this line causes a problem it seems too.
thx for the tips and for pointing out my issue. i will try to fix up what you guys have pointed out this evening and retry. also looks like new.txt is worth a closer look.
any idea why the PIC seems to just hang? I would expect that it would at least toggle the statusLED for me that i've set up before it gets to a piece of bad code. or, am i to expect "unexpected" behavior anytime i muck up the logic anywhere in the program?
russ
