PDA

View Full Version : 16F1824 "HPWM 2" woes (command seemingly forces PortC.3 from analogue to digital)


HankMcSpank
- 20th April 2011, 20:27
I have a 16F1824 PortC.3 (ie pin 9) set as an analogue ADC input pin...it works fine, until my program code gets reaches this line...


HPWM 2, duty, 32767 (where duty is a variable representing erhm...the duty cycle)

once that point is reached, my PortC.3 apparently flips to a digital pin....ie when I vary my analogue input signal into PortC.3, I observe only a '0' or a '1' as the incoming Signal level.

If I use the following as a workaround on each occurence...



HPWM 2, duty, 32767
TRISC = %11110111 ' PIN 7 RC3 HPWM OUTPUT

...then all is well again (my ADC readings are as expected....ie between 0 & 255)

Now I've never used the 2nd channel of PWM before (ie I've always just used one channel...."HPWM 1, duty, 32767"), so it's likely something simple I'm doing wrong, but I've looked at this so long now that the number of the beast starts appearing in my main loop. (which probably means it's time to get a fresh pair of eyes on the case)

Any ideas?

Here's the full test code...

REDRUM, REDRUM,REDRUM,REDRUM,REDRUM, REDRUM,REDRUM,REDRUM,REDRUM, REDRUM,REDRUM,REDRUMREDRUM, REDRUM,REDRUM,REDRUM





@ __CONFIG _CONFIG1, _FCMEN_OFF & _FOSC_INTOSC & _WDTE_OFF & _MCLRE_OFF & _CP_OFF & _IESO_OFF & _BOREN_OFF & _PWRTE_OFF & _LVP_OFF
@ __CONFIG _CONFIG2, _LVP_OFF

DEFINE OSC 4

Osccon = %01101010 'sets the internal oscillator to 4Mhz

DEFINE DEBUG_BAUD 9600
DEFINE DEBUG_MODE 2
DEFINE DEBUG_REG PORTC
DEFINE DEBUG_BIT 4

DEFINE ADC_BITS 8 ' ADCIN resolution (Bits)
DEFINE ADC_CLOCK 1 ' ADC clock source (Fosc/8)
DEFINE ADC_SAMPLEUS 11 ' ADC sampling time (uSec)

ADCON0 = %00010101 'AN5 (Pin 9) Enabled for BACKGROUND adc'ING VIA SPECIAL EVENT TRIGGER + ADC On
ADCON1 = %01000011 'set ADC +ve ref to be the internal Fixed VREF
FVRCON = %11000010 'turn fixed voltage reference on & set it to 2.048V (last two bits set the voltage ...01 =1.024V,10=2.048


CM1CON0 = 0 ' COMPARATORS OFF
CM1CON1 =0
CM2CON0 = 0 ' COMPARATORS OFF
CM2CON1 =0
CPSCON0 = 0 'CAPACTIVE SENSE MODULE OFF
RCSTA.7 = 0 'SERIAL PORT OFF

TRISA = %11111111
TRISC = %11110111 ' PIN 7 RC3 HPWM OUTPUT
ANSELA = 0
ANSELC = 0

CCP2CON = %00001100 'NOT ENTIRELY CONFIDENT ABOUT THIS ONE!

Sig_In VAR BYTE ; a variable to store to DC level coming in on pin 19 (RA0)
duty var word
latest_sample var word


'setup the AtoD & special event register to capture samples in the background
T1CON = %00000001 ' timer1 on
CCP4CON = %0001011 'set CCP4 to special event trigger.

CCPR4H = 0 ' preset for timer1 match (MSB register)
CCPR4L = 250 ' @16Mhz, 200 gives 20,000khz sample & 250 gives 16khz sample.

duty = 255
HPWM 2, duty, 32767
TRISC = %11110111 ' PIN 7 RC3 HPWM OUTPUT

TEST:
latest_sample = ADRESH
PAUSE 500
debug dec LATEST_SAMPLE,13,10
GOTO TEST
cncmachineguy
- 20th April 2011, 20:46
Hank, I am fairly certain you can't use the pin for PWM output and Analog input at the same time. Your work around is not a work around at all. It is multiplexing the pin for both functions.

There are both on C.3 - Yes?
HankMcSpank
- 20th April 2011, 21:05
Hank, I am fairly certain you can't use the pin for PWM output and Analog input at the same time. Your work around is not a work around at all. It is multiplexing the pin for both functions.

There are both on C.3 - Yes?

Hi Bert.....you see how a fresh pair of eyes helps? :-) now as it goes, I'm not using the same pin for both functions...

16f1824 Pin 7 (portC.3) is CCP2 (I'm assuming that the Picbasic command "HPWM 2" uses CCP2?)

16f1824 Pin 9 (PortC.1) is my ADC input pin ...but I can see that that's been set wrong in my code cut/paste above (that pasted code was a quick rip/strip from my monster main program, and there's likely been an error as I've also been rejigging pins to make for a better PCB layout!)

correction made to my code in bold below....



@ __CONFIG _CONFIG1, _FCMEN_OFF & _FOSC_INTOSC & _WDTE_OFF & _MCLRE_OFF & _CP_OFF & _IESO_OFF & _BOREN_OFF & _PWRTE_OFF & _LVP_OFF
@ __CONFIG _CONFIG2, _LVP_OFF

DEFINE OSC 4

Osccon = %01101010 'sets the internal oscillator to 4Mhz

DEFINE DEBUG_BAUD 9600
DEFINE DEBUG_MODE 2
DEFINE DEBUG_REG PORTC
DEFINE DEBUG_BIT 4

DEFINE ADC_BITS 8 ' ADCIN resolution (Bits)
DEFINE ADC_CLOCK 1 ' ADC clock source (Fosc/8)
DEFINE ADC_SAMPLEUS 11 ' ADC sampling time (uSec)

ADCON0 = %00010101 'AN5 (Pin 9) Enabled for BACKGROUND ADC'ing VIA SPECIAL EVENT TRIGGER + ADC On
ADCON1 = %01000011 'set ADC +ve ref to be the internal Fixed VREF
FVRCON = %11000010 'turn fixed voltage reference on & set it to 2.048V (last two bits set the voltage ...01 =1.024V,10=2.048


CM1CON0 = 0 ' COMPARATORS OFF
CM1CON1 =0
CM2CON0 = 0 ' COMPARATORS OFF
CM2CON1 =0
CPSCON0 = 0 'CAPACTIVE SENSE MODULE OFF
RCSTA.7 = 0 'SERIAL PORT OFF

TRISA = %11111111
TRISC = %11110101 ' PIN 7 RC3 HPWM OUTPUT
ANSELA = 0
ANSELC = %00000010 ' AN5 Pin9 (PortC.1) ADC input ....BACKGROUND ADC'ing VIA CCP4 SPECIAL EVENT TRIGGER
CCP2CON = %00001100 'NOT ENTIRELY CONFIDENT ABOUT THIS ONE!

Sig_In VAR BYTE ; a variable to store to DC level coming in on pin 19 (RA0)
duty var word
latest_sample var word

'setup the AtoD & special event register to capture samples in the background
T1CON = %00000001 ' timer1 on
CCP4CON = %0001011 'set CCP4 to special event trigger.

CCPR4H = 0 ' preset for timer1 match (MSB register)
CCPR4L = 250 ' @16Mhz, 200 gives 20,000khz sample & 250 gives 16khz sample.

duty = 255
HPWM 2, duty, 32767
TRISC.1 =1 'Remake PIN 9 PortC.1 an input again following HPWM command

TEST:
latest_sample = ADRESH
PAUSE 500
debug dec LATEST_SAMPLE,13,10
GOTO TEST


....still the same though, when the "HPWM 2" line is reached, my ADC doesn't read properly. I have to reset PortC.1 back to an input, so it looks like the "HPWM 2" command is turning PortC.1 into an output....which for the life of me, I can't understand why that would happen?

Now there'a lot of features in this new(ish) PIC and it's likely something I'm not grasping or seeing, but like I say...way too many hours trying to nail this one.......I now need all the help I can get!
mtripoli
- 20th April 2011, 22:01
Never mind...
HankMcSpank
- 20th April 2011, 23:28
Well, for all my quick 'n dirty the workaround reverts my pin9 (PortC.1) ADC back to how it should be...



HPWM 2, duty, 32767
TRISC.1 =1 'Remake PIN 9 PortC.1 an input again following HPWM command



....the actual HPWM command itself is still playing havoc with the signal....when I feed a steady sine wave into the ADC pin, it getting chopped up as viewed on a scope, as when the HPWM 2 command runs it interferes with PortC.1 - my now (naive) thinking is that it's forcing the pin to be an output, which then drags my sine wave analogue input signal down to zero, then my 'kludgey' follow on command makes the pin an input again...which restores the signal. But you can imagine the end result, it's as if a switch is being flicked rapidly thereby shunting the signal to ground .....fugly.

i'm exhausted on this one & don't have an immediate good solution...I need to sleep on it (still a high possibility I'm just doing something wrong as my head is fuzzy now!). Tomorrow, I may sit down & work out how do do HPWM longhand using the PICs registers (one whole days lost on something as simple as this....eek!)
Bruce
- 20th April 2011, 23:43
DEFINE CCP2_REG PORTC
DEFINE CCP2_BIT 3

Should take care of it. PBP defaults to PORTC.1 for CCP2 if you don't tell it otherwise.
HankMcSpank
- 21st April 2011, 00:47
Thanks Bruce ...that did it....you've just made an impossibly bad day...just bad now :)

Like I say it's my first time using the 2nd HPWM channel ....but also my first time using this newish PIC in anger - thought i was going crazy.

PWM, SHmeeWM.....
Ioannis
- 21st April 2011, 11:15
These new 16F18xx chips are small in the eye but really beasts!

I had my difficult time in setting them too, but nowI really like them.

Here is my setup for the 1827 which may help others to take into accoun the registers and not forget any of them.

Of course this is my setup and other need to make their own. It just as an example.



TRISB = %11000101
TRISA = %11111100

ADCON0 = %00000001
ADCON1 = %11110011

ANSELA = %00011100 ' all digital. A/D disabled
ANSELB = %01000000

APFCON0 = %00100000
APFCON1 = %00000001

BAUDCON = %00011000

CCP1CON = %00111100 'Control of CCPx modules
CCP2CON = %00000000
CCP3CON = %00000000
CCP4CON = %00000000

CM1CON0.7=0
CM2CON0.7=0

CPSCON0 = 0
CPSCON1 = 0

DACCON0 = 0

FVRCON = %10000011

INTCON = %11000000 ' global & peripheral ints enabled

OPTION_REG = %10000111 'Tmr0 from 256 presc and Int Clock

PIR1 = 0 ' clear TMR1 int flag
PIE1 = %00000001 ' TMR1 int enabled

T1CON = %00110000 ' TMR1 1:8 prescale, timer1 on

WPUB=%11000000

'hpwm 1,0,2000
DEFINE CCP1_REG PORTB
DEFINE CCP1_BIT 3
ccptmrs=0
pr2=249
ccp1con=$0C



HTH,

Ioannis
HankMcSpank
- 21st April 2011, 11:27
I concur...the 16F182x are a great range of chips...absolutely chocka with features.

I can't say I've had a lot of setup problems with them (certainly nothing that wasn't obvious when I complied)....other than this pesky HPWM 2 issue, the one other issue I had Darrel kindly helped sort....

I needed to modify the 16F1824.INC file (mods in red)....



;************************************************* ***************
;* 16F1824.INC *
;* *
;* By : Leonard Zerman, Jeff Schmoyer *
;* Notice : Copyright (c) 2010 microEngineering Labs, Inc. *
;* All Rights Reserved *
;* Date : 09/14/10 *
;* Version : 2.60b *
;* Notes : *
;************************************************* ***************
NOLIST
ifdef PM_USED
LIST
"Error: PM does not support this device. Use MPASM."
NOLIST
else
LIST
LIST p = 16F1824, r = dec, w = -302
INCLUDE "P16F1824.INC" ; MPASM Header
;__config _CONFIG1, _FOSC_HS
;__config _CONFIG2, _PLLEN_OFF & _LVP_OFF
NOLIST
endif
LIST
BLOCK_SIZE EQU 32
CCPTMRS0 = CCPTMRS



The first two mods (semi colons) are standard fayre, but that last one would have taken me a good few years to get to the bottom of on my own.
Ioannis
- 21st April 2011, 12:06
What version of MPLAB do you have?

On my 8.56 and in 1827.inc file of the MPLAB Suite there are two lines (BANK5 grroup):



CCPTMRS EQU H'029E'
CCPTMRS0 EQU H'029E'


But on 1824.inc there is only one:



CCPTMRS EQU H'029E'


So it may have to do with the update of MPLAB. Anyone with newer version?

Ioannis
HankMcSpank
- 21st April 2011, 14:19
I'm running with V8.63.