Updating HPWM frequently, safe?

[T.Jackson]

Great thread this.

[mister_e]

O.K. i'll stop here... at least by respect to our members and readers, and because it's not a chit-chat area.

[skimask]

...deleted...for obvious reasons

[peu]

mmm, could be a noisy pot as well... using your code and on many low-value ADC.. they are all stable.. no noise, no frlicker.

what happen if you stick a 0.1 uF in parallel with your POT wiper?

Any 0.1uF + 10uF capacitor close to your PIC?

I just tried with a 20 turn 10k linear pot and its the same, once I reach the lower duty cycles the led flickers.
The difference is clearly visible with the multiturn pot, once you go below a certain level it flickers, move the other way a little bit and the flicker goes away

Reading in the flicker position I get this range: from $00BB to $00BF and spurious $009F , $00B4, $00C6, $00E3 at random places
Reading in the steady position I get this range: $00CD to $00CF and no spurious readings

I only read 125 word values into the device own eeprom.
Again Im using a li-ion battery for powering the circuit (3.65v)

Weird isnt it? even weirder is that sometimes when I write the values to eeprom the written values are steady (also visually) but only when writting...

Just in case here is the code I used:

'       Pic Configuration
@ device pic12f683,INTRC_OSC_NOCLKOUT , wdt_on, mclr_off, protect_off        

'       Hardware configuration

'       I/Os
TRISIO = %00010010      ' pin1 is ADC , pin4 is the calibration button

'       ADC's
ANSEL = %00000010  
DEFINE ADC_BITS 10      ' ADCIN resolution  (Bits)
DEFINE ADC_CLOCK 1      ' ADC clock source  (Fosc/8)
DEFINE ADC_SAMPLEUS 11  ' ADC sampling time (uSec)                      
ADCON0.7 = 1            ' Right justified results

'       Comparator
CMCON0 = %00000111     ' Disable comparator
VRCON  = %00000000     ' disable
'cmcon0 = %00001110      ' Multiplexed Input with Internal Reference CIS=1
'vrcon  = %10001010      ' set vref=2.25v (from 4v input) 

include "HPWM10.pbp"

' definitions
pote 		var word
Frequency   var word
calibraled  var gpio.5
calibrapin  var gpio.4
conta       var byte

gpio=0
conta=0
frequency=1000
pause 50    ' coffee break

loop:

    adcin 1, pote
    if pote>1000 then pote=1000
    
if calibrapin=1 then
if conta < 250 then
 write conta, pote.byte1
 write conta+1, pote.byte0
 conta=conta+2
endif
endif

@ HPWM10  1, _pote, _Frequency
pause 10

goto loop
end

Pablo

[mister_e]

O.K. i have to go out. I had the same code, and i didn't have any problem...

when i'll be back, i'll use a lower voltage AND use HighBrightness LEDs.

Wish i could see the same thing....

Still thinking there's an hardware problem...

[Darrel Taylor]

Pablo,

Good new, Bad news.

I finally found my 12F683 and gave it a go.
I've completely duplicated your problem. Blinking and all.

For some reason it only does it on the 683.
Every other chip I tried seems to work fine.

And the bad news is, I haven't found out why yet. (Operative word, Yet)
It appears to be with the A/D, the readings are all over the place at low voltage input.

I still have many things to try, but just wanted you to know you're not crazy.
And, I WILL find it. (unless somebody beats me to it)
<br>

[peu]

Pablo,

Good new, Bad news.

I finally found my 12F683 and gave it a go.
I've completely duplicated your problem. Blinking and all.

For some reason it only does it on the 683.
Every other chip I tried seems to work fine.

And the bad news is, I haven't found out why yet. (Operative word, Yet)
It appears to be with the A/D, the readings are all over the place at low voltage input.

I still have many things to try, but just wanted you to know you're not crazy.
And, I WILL find it. (unless somebody beats me to it)
<br>

As I see it its only good news Thanks for all your efforts!!!

I can't help with your code, is way beyond my actual knowledge, I hope to catch up with it in the future!

Now I patiently wait


Pablo

[peu]

An EE friend of mine (Thanks Doug) told me a new possibility about the cause of the problem:

What if reading the ADC so often does not give it enough time to handle the low voltage values?

(more discussions via chat with my friend)

He points out something interesting, the test program initializes the ADC using:

DEFINE ADC_BITS 10 ' ADCIN resolution (Bits)
DEFINE ADC_CLOCK 1 ' ADC clock source (Fosc/8)
DEFINE ADC_SAMPLEUS 11 ' ADC sampling time (uSec)

But reading the pic basic manual we have:

ADC_SAMPLEUS is the number of microseconds the program waits
between setting the Channel and starting the analog to digital
conversion. This is the sampling time. The minimum number of
microseconds usable is determined by the minimum time for PAUSEUS.
See it’s section for this information.

and the PAUSEUS section says:

For 4mhz Minimum Delay is 24us

So, the initial theory appears to be right, I don't have the protoboard here now, but I will check this tomorrow.


Pablo

[Darrel Taylor]

I tried every AD oscillator / Sample time possible, and it made no difference.
Even with the A/D input tied directly to ground, it was getting crazy readings between 0 and 20.

After I tried it with a 12F675 (without the CCP) and got the same results,
I finally figured out that I had removed the big cap on the power lines on this protoboard.
Probably needed it for something else. Doh!

After putting a 100uf cap across the power supply, everything calmed down again.
It's still not perfect, but I think that has alot to do with the breadboard.

So then to really stabilize things I added an averaging routine, and I think that's about as close as we're going to get.

First make sure you have enough capacitance on your power lines (I think mister_e already suggested that).

Then add this to the bottom of the program

' -=-=-=-=-=-=  Average Analog values -=-=-=-=-=-=-=-=-=-=
Avg       VAR WORD
AvgCount  CON 16
spread    CON 20

Average:    ' Smooth data
    IF ABS (pote - Avg) > spread then
        Avg = pote
    else
        Avg = (Avg*(AvgCount-1)+pote)/AvgCount 
        pote = Avg                               
    endif
Return

And also, just after the ADCIN statement, add this

    gosub Average

HTH,

[mister_e]

it's getting interesting... i have everything on hand, will do some test here.

Sorry, i was suppose to do it yesterday

[peu]

well I hope you find something because I tried adding a 100uF capacitor in the power lines or in parallel with the wiper and my problems remain, even after adding the averaging routine provided by Darrel.

I even started looking for other pic alternatives in case this can't be solved...

Thanks again

[mister_e]

still working @ 3.x Volt?

[peu]

yes Sir, using a 3.6v li-ion batt


Pablo

[mister_e]

O.K. can you try to program you pic with the .HEX in attachment and let me know what's happening now?

Finger crossed...

[peu]

O.K. can you try to program you pic with the .HEX in attachment and let me know what's happening now?

Finger crossed...

it uses the circuit I provided right? will test it tomorrow and let you know. Thanks!!!

[mister_e]

Yup it's the same circuit. CalibraPin is also implemented.

Here, it's working, no flicker or else... 'till now

[peu]

yes it flickers, but this one instead of graduating continuously goes in steps, and the calibrapin button does nothing, at least nothing I can see or read.

Thanks!

[mister_e]

can you try something because it's working really fine here... use another ADC channel, like GP4, or GP5.

[peu]

can you try something because it's working really fine here... use another ADC channel, like GP4, or GP5.

what do you mean by using another ADC? change the my test program to read in another port ?

thanks

[mister_e]

In your example you used ADCIN 1,pote to read GP1. Change it to ADCIN 5,pote. Sure you'll need to change your hardware and some regisers settings as well.

Next step will be to use a TIMER interrupt instead.

[peu]

In your example you used ADCIN 1,pote to read GP1. Change it to ADCIN 5,pote. Sure you'll need to change your hardware and some regisers settings as well.

Next step will be to use a TIMER interrupt instead.

I will do two things, 1st etch a pcb for the circuit and test it again, second try with a 16F628A

I let you know. thanks!

[peu]

made the etched pcb and only tried it with Mister_e hex file and it worked without flickerings. Im not sure if its supposed to smooth or in steps, I see steps.

More tests on monday. Thanks!!

[mister_e]

steps are normal with my HEX. but good news! and thanks to report your results. Sure many code before my last one will now work as designed.

[Acetronics2]

Hi, Darrel,

I just try tu make HPWM10 work aboard a 12F683 ...

and get the following Message ( MPASM compiler ) :

Error[108] C:\PROGRAM FILES\MICROCHIP\MPLAB IDE\PROJETS\SPEEDCOMM.ASM 185 : Illegal character (1)

the illegal character is ... the PWM Channel !!!

@ HPWM10 1, _DutyCycle, _Frequency

I used the example " straight out of the box ...

do you have any idea ???

Alain

[Acetronics2]

re,

The idea is : change "HPWM10.pbp" into "HPWM.bas", for the include command.

Alain

[Darrel Taylor]

Error[108] C:\PROGRAM FILES\MICROCHIP\MPLAB IDE\PROJETS\SPEEDCOMM.ASM 185 : Illegal character (1)

There had to be more errors than that.
The others might have had more clues, but it appears that PBP couldn't find the file.

Are you saying that you renamed it to HPWM.bas but still had INCLUDE "HPWM10.pbp"?

confused.
<br>

[Acetronics2]

There had to be more errors than that.
The others might have had more clues, but it appears that PBP couldn't find the file.

Are you saying that you renamed it to HPWM.bas but still had INCLUDE "HPWM10.pbp"?

confused.
<br>

Hi, Darrel,

Hoooo ... bad evening ! was HPWM10.pbp into HPWM10.bas.

the goal was to transform a R/C pulse into a corresponding 10 bits PWM @ 10 kHz ... and drive a "Diesel engine" sound machine.

Have a nice week end.

Alain

[Acetronics2]

Hi, Darrel

I Was a bit surprised by those lines :

 Line    Address    Opcode        Label                    Disassembly              

    359     166       00BD                    MOVWF 0x3d                             
    360     167       084C                    MOVF _HP_Channel, W                    
    361     168       3C05                    SUBLW 0x5                              
    362     169       1C03                    BTFSS STATUS, 0                        
    363     16A       2975                    GOTO L00001                            
    364     16B       3001                    MOVLW 0x1                              
    365     16C       008A                    MOVWF PCLATH                           
    366     16D       084C                    MOVF _HP_Channel, W                    
    367     16E       0782                    ADDWF PCL, F                           
    368     16F       297A                    GOTO _GetCH_2                          
    369     170       2975                    GOTO L00001                             
    370     171       297A                    GOTO _GetCH_2                          
    371     172       297A                    GOTO _GetCH_2                          
    372     173       297A                    GOTO _GetCH_2                          
    373     174       297A                    GOTO _GetCH_2                          
    374     175       0813   L00001           MOVF CCPR1L, W                         
    375     176       00C0                    MOVWF _HP_Temp                         
    376     177       0815                    MOVF CCP1CON, W                        
    377     178       00C1                    MOVWF _HP_Temp??HIGHBYTE               
    378     179       297A                    GOTO _GetCH_2                          
    379     17A       0840   _GetCH_2         MOVF _HP_Temp, W                       
    380     17B       00A0                    MOVWF BANK0_START                      
    381     17C       01A1                    CLRF 0x21                              
    382     17D       3002                    MOVLW 0x2                              
    383     17E       205A                    CALL SHIFTL

are those 6 times " GOTO _GetCH_2 " ( L368 , 370 - 373 , 378 ) something " regular " ???


the source for 12F683 @ 8 Mhz internal is :

@ __config _INTRC_OSC_NOCLKOUT &_FCMEN_OFF &_IESO_OFF & _WDT_ON &_PWRTE_ON & _MCLRE_OFF & _CP_OFF

 OSCCON = %01110111					' Internal Clock 8 Mhz
 
 CMCON0 = 7
 VRCON  = 0
 ADCON0 = 0
 ANSEL  = 0

 DEFINE OSC 8                   	'Declare 8 Mhz operation 
 DEFINE PULSIN_MAX 4500

 
' PINS DEFINITIONS

	Signal	var GPIO.3
	Drive	var GPIO.2
	
	GPIO	= 0	
	TRISIO 	= %00001000

	
' VARIABLE DEFINITIONS

  	speed     	var Word              	' assign Word variable to speed demand
  	w1			var Word				' Table	output
  	
  	DutyCycle  	VAR WORD
	Frequency 	VAR WORD
	Temp       	VAR WORD	
		

	include "HPWM10.bas"

	LOW Drive  ; Set CCP1 pin to output low

	Frequency = 10000

'******************************************************************************
'MAIN LOOP
 main:
  

   pulsin Signal,0,speed				' Read servo pulse ( Inverted !!! ) 
           
'******************************************************************************
' Throttle amount topped to 100 and reduced to 25 steps
  
   if speed > 300 then
     speed = ( speed - 300 )MIN 100
   else
     speed = ( 300 - speed )MIN 100
   endif
    
   speed = speed >> 2
   
'******************************************************************************
' Conversion logarithmique 10 bits - 800 maxi à 10 khz

	
	lookup2 speed, [0,0,161,277,378,467,543,611,670,721,766,804,836,867,893,_
	 933,950,965,978,990,1000,1009,1017,1023,1023 ],w1                     	                     

'******************************************************************************           
' !!! Passer en " HPWM 10 " !!!

    if w1 <> DutyCycle then  DutyCycle = w1
    
    @ HPWM10 1, _DutyCycle, _Frequency  
     
   goto main

END

Regards

Alain

[Darrel Taylor]

Hi, Darrel

I Was a bit surprised by those lines :
...
are those 6 times " GOTO _GetCH_2 " ( L368 , 370 - 373 , 378 ) something " regular " ???

The lines showing the same labels is the result of Disassembling.
There are actually 6 different labels, with the code in-between optimized out since the chip doesn't have 5 CCP modules.

It comes from this section ...

;-----  Get Current DutyCycle ------------------------------------------------
    Branch HP_Channel,[GetCH_Done,GetCH_1,GetCH_2,GetCH_3,GetCH_4,GetCH_5]

    GetCH_1:
            ASM
              ifdef CCP1CON  
                MOVE?BB  CCPR1L,  _HP_Temp
                MOVE?BB  CCP1CON, _HP_Temp + 1
                goto _GetCH_Done        
              endif  
            endasm
    GetCH_2:  
            ASM
              ifdef CCP2CON  
                MOVE?BB  CCPR2L,  _HP_Temp
                MOVE?BB  CCP2CON, _HP_Temp + 1
                goto _GetCH_Done        
              endif  
            endasm

The GOTO's are created by the Branch statement.
And it looks like this in the .LST file ...

016B   3001               M         movlw   (jmptbl) >> 8
016C   008A               M         movwf   PCLATH
016D   084C               M         movf    _HP_Channel, W
016E   0782               M         addwf   PCL, F
016F                      M jmptbl
016F   297A               M         goto    _GetCH_Done
0170   2975               M         goto    _GetCH_1
0171   297A               M         goto    _GetCH_2
0172   297A               M         goto    _GetCH_3
0173   297A               M         goto    _GetCH_4
0174   297A               M         goto    _GetCH_5
0175                      M L00001
0175                      M _GetCH_1
0175   0813               M         movf    CCPR1L, W
0176   00C0               M         movwf   _HP_Temp
0177   0815               M         movf    CCP1CON, W
0178   00C1               M         movwf   _HP_Temp + 1
0179   297A           00326                 goto _GetCH_Done        
017A                      M _GetCH_2
017A                      M _GetCH_3
017A                      M _GetCH_4
017A                      M _GetCH_5
017A                      M _GetCH_Done

Since the labels are all at the same address, the disassembler chooses the first label it finds at that address (_GetCH_2).

[Acetronics2]

Hi, Darrel

Tanks for your explanation

Alain

[Samoele]

Hello,

I would have two different frequency on CCP1 and CCP2 pin the first is 100HZ and second is 20Khz-40KHz. I use 1MHz quartz for clock and HPWM10 command from D.T. I use PIC16F876.
I have on both CCP pin HPWM10 command with frequency 100hz and different duty. This working ok. When changed frequency on CCP1 to 1KHz then is frequency on both pin unchanged and very unstable
When change one HPWM10 command to normal 8 bit HPWM then stay result unchanged.
what I did wrong?

here is first part of my program:




INCLUDE "ALLDIGITAL.pbp"

Cifra var byte
Cifra1 var byte
Maska1 var byte
Maska2 var byte
W var byte
i var byte
time var byte
PR var byte
VU var word
VD var word
Duty VAR word
Duty1 VAR word
FREQUENCY var word
FREQUENCY1 var word

LED var PORTC.4
LEDDis1 var PORTA.0
LEDDis2 var PORTA.1
tipka1 var porta.2
tipka2 var porta.3
tipka3 var porta.4
BUZ var porta.5
enable1 var portc.0
b0 var byte
b1 var byte
b2 var byte
b3 var byte
impulz var WORD
minuta var byte
key var byte
pavza var word

bp var byte

TRISA=%11111100
TRISB=%00000000
TRISC=%00000000


clear
include "HPWM10.pbp"
LOW PORTC.1
LOW PORTC.2
PWM_ON CON %00001100
PWM_OFF CON 0
time=6
FREQUENCY=93 '
FREQUENCY1=1000
low led
gosub zvok
eeprom 0,[31]
eeprom 2,[55]
high enable1
high led
pause 100
low led




INTCON = %00100000
OPTION_REG = %10001000

On Interrupt Goto ISR
INTCON = %10100000
w=0


Main:

gosub tipka
gosub timing
read 0,b2
gosub tpwm
@ HPWM10 2,_IMPULZ,_FREQUENCY
read 2,b0
gosub ppwm
@ HPWM10 1,_duty,_FREQUENCY1
Gosub Priprava
goto Main

ppwm: b1=b0-50
lookup2 b1,[100,200,300,400,500,600,700,800,900,1000],duty
return

tpwm: b3=b2-30
lookup b3,[0,31,16,6,21,46],duty1
return

[Darrel Taylor]

With 16F876's, both CCP modules use the same Timer in PWM mode (TMR2).
If you change the frequency of one, the other one changes too.

With a 4 Mhz crystal, the minimum HPWM frequency is 245 Hz.
Since you are using a 1 Mhz crystal, you can get down to 62 Hz, but you have to specify the frequency as if it were running at 4 Mhz.

If you want 100 Hz, then specify 400 Hz in the HPWM10 command.
93 is way below the minimum.
<br>

[Samoele]

Frequency of 100 Hz is not a problem. In the case that I use PIC16F886 asking whether the command HPWM10 set timer TMR1 and TMR2 for each channel?
Can you write the registry settings for the PIC16F886?

tnx
Samo

[Darrel Taylor]

Even on a PIC16F887, the CCP modules can only use Timer2 in PWM mode.
So both PWM channels will operate at the same frequency.

So I guess the answer to your question is ...
HPWM10 only changes Timer2, not Timer1.
<br>

[jamie_s]

Darrel,

Whats required to have 'glitch free' operation but with 8 bit PWM, can the include file be easily modified for 8 bit, or should i just multiply my desired 8 bit dutycycle by 4?

[jamie_s]

Also, just testing on 16F887, at 4mhz,

@ HPWM10 1,_OUTDUTY,300 results in a PWM frequency of 10khz according to my o'scope ???
I see the duty cycle changing, but the freq is around 9-10khz?

[Darrel Taylor]

The frequency must be a variable.
You can't pass a constant as the parameter.

And just shift left 2 (<<2) to use 8-bit values.

hth,

[Christopher4187]

Old thread but very relevant to my issue.

Can I use HPWM10:

1. On an 18F87J50?
2. And have it glitch free?
3. To make a pulse train that's 140mS long and have varying duty cycles within the pulse train?

In other words: 300uS on, 200uS off, 300uS on on, 200uS off, 130uS on, 370 uS off, 300uS on, 200uS off, 130uS on, 370uS off, 170uS on, 330uS off, 130uS on, 370uS off, etc.

Each bit is 500uS and I don't need to modify anything while the packet is being sent; I can modify right before the next one is sent.

[Art]

If by “glitch free” you mean free of any resolution error, not every frequency or duty cycle you want to generate will be.
for any particular frequency, you need the pic’s clock frequency to tell you that.
This online PWM calculator can tell you:
http://www.micro-examples.com/public...alculator.html

whether or not you can change it for arbitrary duration for every pulse like that would be a tough one,
I don’t think that’s possible, but might be wrong.

Depending what else the chip has to do, it might be possible to use timer1 overflow that way.