Changing the sines frequency?

[HenrikOlsson]

Hi,
Interrupt frequency of 39060Hz?

Previously you said you wanted an output of 120Hz and with 72 steps per sin-cycle that's 120*72=8640Hz interrupt frequency. The math is then like this

65535 - (10,000,000 / 8640) + 1 = 64378

64378 is what the timer reload value is and this means that there's 65536-64378=1158 cycles between interrupts. 1158 cycles times 100ns per cycle equals 115.8us between interrupts 1/115.8us = 8635Hz so it's pretty close.

This means that there's only 1158 cycles (or instructions if you will) between interrupts. In this time the PIC has to save all the PBP system variables, reload the timer, execute your interrupt service routine and then restore all the PBP system variables.

Why are you trying to interrupt at ~39kHz? That's only about 256 cycles between interrupts which I'm pretty sure won't be enough time to do all that needs to get done in the interrupt.

Finally, this line:

If PortC.4 = 1 Then InterruptFrequency = InterruptFrequency -1 : Flag = 1

Probably doesn't work like you think. Here, Flag gets set no matter what the state of of PortC.4 actually is. So you are always running the recalc routine. If you want Flag to be set only when PortC.4 is high you need to change it to:

If PortC.4 = 1 THEN
  InterruptFrequency = InterruptFrequency - 1
  Flag = 1
ENDIF

[pxidr84]

Hi,
Interrupt frequency of 39060Hz?

Previously you said you wanted an output of 120Hz and with 72 steps per sin-cycle that's 120*72=8640Hz interrupt frequency. The math is then like this

65535 - (10,000,000 / 8640) + 1 = 64378

64378 is what the timer reload value is and this means that there's 65536-64378=1158 cycles between interrupts. 1158 cycles times 100ns per cycle equals 115.8us between interrupts 1/115.8us = 8635Hz so it's pretty close.

This means that there's only 1158 cycles (or instructions if you will) between interrupts. In this time the PIC has to save all the PBP system variables, reload the timer, execute your interrupt service routine and then restore all the PBP system variables.

Why are you trying to interrupt at ~39kHz? That's only about 256 cycles between interrupts which I'm pretty sure won't be enough time to do all that needs to get done in the interrupt.

Finally, this line:

If PortC.4 = 1 Then InterruptFrequency = InterruptFrequency -1 : Flag = 1

Probably doesn't work like you think. Here, Flag gets set no matter what the state of of PortC.4 actually is. So you are always running the recalc routine. If you want Flag to be set only when PortC.4 is high you need to change it to:

If PortC.4 = 1 THEN
  InterruptFrequency = InterruptFrequency - 1
  Flag = 1
ENDIF

Well, for 120Hz, I need an interrupt frequency of 120*90=10800Hz.
If I configure this in my program, I get output sines of only ~16Hz.

So what's wrong?

The code :

' PIC initialization
DEFINE OSC 40      
DEFINE LCD_DREG PORTC
DEFINE LCD_EREG PORTD
DEFINE LCD_RSREG PORTD
DEFINE LCD_EBIT 0
DEFINE LCD_RSBIT 1
DEFINE LCD_COMMANDUS 4000
DEFINE LCD_DATAUS 1000


' BAS includes
INCLUDE "DT_INTS-18.bas"
INCLUDE "ReEnterPBP-18.bas"
INCLUDE "Sine_table.bas"


' Port registers configuration
TRISB=%11000000   ' PWM 0,1,2,3,4,5 outputs
TRISC=%00110000   ' +/- buttons


' PCPWM registers configuration
DTCON=%110        ' Deadtime (600ns)
PTCON0=%0         ' 1:1 postscale, Fosc/4 1:1 prescale, free running mode
PTCON1=%10000000  ' PWM time base is ON, counts up, 19.45kHz/4
PWMCON0=%1000000  ' PWM 0,1,2,3,4,5 set in pair mode
PWMCON1=%1        ' PWM timer sync configuration


' PWM calculation variables
ustep var byte
vstep var byte
wstep var byte
uduty var word
vduty var word
wduty var word
amplitude var word
carrier VAR word
flag var bit
Dummy VAR WORD
Dummy2 VAR WORD
Dummy3 VAR WORD
InterruptFrequency VAR WORD
TMRCopy VAR WORD
TimerReloadValue VAR WORD


' Variables definition
ustep=90          ' 360 degrees phase angle
vstep=60          ' 240 degrees phase angle
wstep=30          ' 120 degrees phase angle
amplitude=65535   ' Sinewave amplitude adjust (65535=max amplitude)
carrier=1023      ' Carrier frequency adjust (1023=13kHz)
flag=0           ' Menu flag
Dummy2 = 10000
Dummy3 = 10000
InterruptFrequency = 10800


' PWM carrier frequency register configuration
PTPERL=carrier.lowbyte  
PTPERH=carrier.highbyte


' Interrupt processors 
ASM
INT_LIST macro
INT_Handler TMR1_INT,_pwmint,PBP,yes
endm
INT_CREATE       
ENDASM


' Interrupts enable
@INT_ENABLE TMR1_INT
T1CON.0=1

recalc:
Dummy = Dummy2 * Dummy3
TimerReloadValue = Div32 InterruptFrequency
TimerReloadValue = 65535 - TimerReloadValue  + 1
flag=0


' Main program loop
mainlp:

LCDOUT $FE,2,"F: ",#InterruptFrequency
lcdout $FE,$C0,"R: ",#TimerReloadValue
if PORTC.4=1 then 
InterruptFrequency=InterruptFrequency-1 
flag=1
ENDIF
if PORTC.5=1 then 
InterruptFrequency=InterruptFrequency+1
flag=1
ENDIF

if flag=1 then goto recalc

goto mainlp


' PWM calculation and update interrupt (Timer 1) 
pwmint:

T1CON.0 = 0                             ' Stop TMR1
TMRCopy.HighByte = TMR1H                ' Copy value of TMR1 registers
TMRCopy.LowByte = TMR1L 
TMRCopy = TMRCopy + TimerReloadValue    ' Add reload value (compensates for overhead)
TMR1H = TMRCOPY.HighByte                ' And back to TMR1
TMR1L = TMRCopy.LowByte
T1CON.0 = 1                             ' Restart Timer

' PWM U phase calculation
uduty=sine[ustep]
uduty=uduty<<4**amplitude

' PWM V phase calculation
vduty=sine[vstep]
vduty=vduty<<4**amplitude

' PWM W phase calculation
wduty=sine[wstep]
wduty=wduty<<4**amplitude

' PWM U, V and W update
PDC0L=uduty.lowbyte
PDC0H=uduty.highbyte
PDC1L=vduty.lowbyte
PDC1H=vduty.highbyte
PDC2L=wduty.lowbyte
PDC2H=wduty.highbyte

' Phase angle calculation
ustep=ustep-1
vstep=vstep-1
wstep=wstep-1

' Phase angle reinitialization
if ustep=0 then ustep=90
if vstep=0 then vstep=90
if wstep=0 then wstep=90

@INT_RETURN

The "flag" problem is corrected now, I can change frequency during runtime.

And there is my sine table (if needed) :

' Sine table (4 degrees/step, 90*4=360 degrees)
sine var byte[90]
sine[1]=128
sine[2]=137
sine[3]=146
sine[4]=155
sine[5]=163
sine[6]=172
sine[7]=180
sine[8]=188
sine[9]=196
sine[10]=203
sine[11]=210
sine[12]=217
sine[13]=223
sine[14]=229
sine[15]=234
sine[16]=239
sine[17]=243
sine[18]=247
sine[19]=250
sine[20]=252
sine[21]=254
sine[22]=255
sine[23]=255
sine[24]=255
sine[25]=255
sine[26]=254
sine[27]=252
sine[28]=250
sine[29]=247
sine[30]=243
sine[31]=239
sine[32]=234
sine[33]=229
sine[34]=223
sine[35]=217
sine[36]=210
sine[37]=203
sine[38]=196
sine[39]=188
sine[40]=180
sine[41]=172
sine[42]=163
sine[43]=155
sine[44]=146
sine[45]=137
sine[46]=128
sine[47]=119
sine[48]=110
sine[49]=101
sine[50]=93
sine[51]=84
sine[52]=76
sine[53]=68
sine[54]=60
sine[55]=53
sine[56]=46
sine[57]=39
sine[58]=33
sine[59]=27
sine[60]=22
sine[61]=17
sine[62]=13
sine[63]=9
sine[64]=6
sine[65]=4
sine[66]=2
sine[67]=1
sine[68]=0
sine[69]=0
sine[70]=1
sine[71]=2
sine[72]=4
sine[73]=6
sine[74]=9
sine[75]=13
sine[76]=17
sine[77]=22
sine[78]=27
sine[79]=33
sine[80]=39
sine[81]=46
sine[82]=53
sine[83]=60
sine[84]=68
sine[85]=76
sine[86]=84
sine[87]=93
sine[88]=101
sine[89]=110
sine[90]=119

[HenrikOlsson]

Hi,
OK, I don't know from where I got that you had 72 steps in the lookup table. Now I know it's 90. If can live with fewer steps you'll be able to increase the output frequency.

I'm pretty sure you're simply running out of time. I made some quick tests here but in order to do so I had to change your code a little bit (don't have 4431 handy to test on). So I'm not sure exactly how much we can depend on these.

I'm using a 18F4520 @20Mhz.

The interrupt routine itself runs in about 70us, the overhead introduced by the interrupt system (DT-ints) is from previous tests I've done about 29us (again at 20Mhz) at interrupt entry and I would guess about the same on exit. This means each interrupt takes a total of ~130us.

1/0.000130 = 7700Hz

I've verified this by specifying InterruptFrequency of 7700Hz and it works, I get 7700 Hz with nice, steady pulses. If I specify 8500Hz I can see on the scope that the trace is starting to get jittery indicating that it's starting to overrun it self, missing pulses so think I'm pretty close to the absolute max at 7500-8000Hz.

Now, in theory, since you're running at 40MHz you should be able to get twice that, ~15kHz but apparently you're not. Can you set a pin high at the beginning of the interrupt service routine and set it low just before @INT_RETURN, then measure how long that output is high. That will tell you the execution time of the actual handler (where I got 70us).

Are you sure that the PIC is running at 40Mhz? Simply specifying DEFINE OSC40 doesn't make it run at 40MHZ, you need to enable the PLL thru the CONFIG bits as well.

/Henrik.

[pxidr84]

Hi,
OK, I don't know from where I got that you had 72 steps in the lookup table. Now I know it's 90. If can live with fewer steps you'll be able to increase the output frequency.

I'm pretty sure you're simply running out of time. I made some quick tests here but in order to do so I had to change your code a little bit (don't have 4431 handy to test on). So I'm not sure exactly how much we can depend on these.

I'm using a 18F4520 @20Mhz.

The interrupt routine itself runs in about 70us, the overhead introduced by the interrupt system (DT-ints) is from previous tests I've done about 29us (again at 20Mhz) at interrupt entry and I would guess about the same on exit. This means each interrupt takes a total of ~130us.

1/0.000130 = 7700Hz

I've verified this by specifying InterruptFrequency of 7700Hz and it works, I get 7700 Hz with nice, steady pulses. If I specify 8500Hz I can see on the scope that the trace is starting to get jittery indicating that it's starting to overrun it self, missing pulses so think I'm pretty close to the absolute max at 7500-8000Hz.

Now, in theory, since you're running at 40MHz you should be able to get twice that, ~15kHz but apparently you're not. Can you set a pin high at the beginning of the interrupt service routine and set it low just before @INT_RETURN, then measure how long that output is high. That will tell you the execution time of the actual handler (where I got 70us).

Are you sure that the PIC is running at 40Mhz? Simply specifying DEFINE OSC40 doesn't make it run at 40MHZ, you need to enable the PLL thru the CONFIG bits as well.

/Henrik.

With this :

' PWM calculation and update interrupt (Timer 1) 
pwmint:

T1CON.0 = 0                             ' Stop TMR1
TMRCopy.HighByte = TMR1H                ' Copy value of TMR1 registers
TMRCopy.LowByte = TMR1L 
TMRCopy = TMRCopy + TimerReloadValue    ' Add reload value (compensates for overhead)
TMR1H = TMRCOPY.HighByte                ' And back to TMR1
TMR1L = TMRCopy.LowByte
T1CON.0 = 1                             ' Restart Timer

TOGGLE PORTD.7

@INT_RETURN

And a interrupt freq of 10800Hz, I get an output square wave of 720Hz on the oscilloscope.

Here is my config bits in MPLAB :


Note I'm using the HS oscillator with PLL enabled. My crystal (40MHz) is connected between OSC1 and OSC2 pins, with 2*47pF ceramic capacitors. On the oscilloscope, it's okay, I can see the oscillation.

[HenrikOlsson]

Hi,
Something must be wrong with your setup. I tried the same ISR as your one, with InterruptFrequency = 10800:

' ---------- Interrupt handler ------------
TimezUp:
PortB.3 = 1
  T1CON.0 = 0                             ' Stop TMR1
  TMRCopy.HighByte = TMR1H                ' Copy value of TMR1 registers
  TMRCopy.LowByte = TMR1L 
  TMRCopy = TMRCopy + TimerReloadValue    ' Add reload value (compensates for overhead)
  TMR1H = TMRCOPY.HighByte                ' And back to TMR1
  TMR1L = TMRCopy.LowByte
  T1CON.0 = 1                             ' Restart Timer
PortB.3 = 0
@ INT_RETURN

And I get ~10500Hz even when running at 20Mhz. Here you can see that I, instead of using TOGGLE set the output high at the entry point and low at the exit which allows me to measure the execution time of the code in between. This exact code executes in ~3us on this 18F4520 running at 20MHz, if you try the same on yours it should be 1.5us - if it's not then you're not running at 40MHz.

You say you have a 40MHz x-tal, is that really correct? To run at 40MHz with the PLL you should use a 10MHz x-tal.

Try the above, what's the pulsewidth on the output? If it's not 1.5us you're not running at 40MHz.

/Henrik.

[pxidr84]

Hi,
Something must be wrong with your setup. I tried the same ISR as your one, with InterruptFrequency = 10800:

' ---------- Interrupt handler ------------
TimezUp:
PortB.3 = 1
  T1CON.0 = 0                             ' Stop TMR1
  TMRCopy.HighByte = TMR1H                ' Copy value of TMR1 registers
  TMRCopy.LowByte = TMR1L 
  TMRCopy = TMRCopy + TimerReloadValue    ' Add reload value (compensates for overhead)
  TMR1H = TMRCOPY.HighByte                ' And back to TMR1
  TMR1L = TMRCopy.LowByte
  T1CON.0 = 1                             ' Restart Timer
PortB.3 = 0
@ INT_RETURN

And I get ~10500Hz even when running at 20Mhz. Here you can see that I, instead of using TOGGLE set the output high at the entry point and low at the exit which allows me to measure the execution time of the code in between. This exact code executes in ~3us on this 18F4520 running at 20MHz, if you try the same on yours it should be 1.5us - if it's not then you're not running at 40MHz.

You say you have a 40MHz x-tal, is that really correct? To run at 40MHz with the PLL you should use a 10MHz x-tal.

Try the above, what's the pulsewidth on the output? If it's not 1.5us you're not running at 40MHz.

/Henrik.

With your exact code (the same you're using) and a INTF of 10800Hz, I get 2,76ms between each pulse. Very weird, and very slow. I've disabled the PLL, the oscilator is configured as a HS oscillator.

[HenrikOlsson]

Hi,
1) What's the width of the pulses? Not the time between them but the width?
2) What's the value stamped on your x-tal?
3) The timing calculations that I provided are based on the fact that your chip is SUPPOSED to run at 40MHz, if it's not running at 40MHz then the timing will be wrong and the interrupt frequency will wrong as well.
4) Please don't quote the complete message you're responding to. It clutters the thread and it gets hard read because everything is quoted all the time - that's my personal view.