Interrupt RPM and Taylors Elapsed time on 18F4620

**Tobias** · - 29th January 2010, 00:36

Here is what I have now. I have set it up for every rising pulse and changed the math because when I incorporate this into my other code, waiting for four pulses delays alot of events.

Its still off with the output RPM numbers by the same percentage.

Code:

DEFINE OSC 20
Prefix          con     $FE           ' needed before each command
LcdCls          CON     $51           ' clear LCD (use PAUSE 5 after)
CursorPS        con     $45           'Cursor Position
Capture         VAR     PIR1.2		    ' CCP1 capture flag CCP1IF
Overflow        VAR     PIR1.0	    	' Timer1 overflow flag TMR1IF:
RPM             var     word
period          var     Word
TotalTime       var     word      'Holds seconds and tics 

LCD             VAR     PortC.6    'LCD output
Gate1           Var     PortC.1
Gate2           var     PortD.1
Gate3           var     PortD.2
Gate4           var     PortD.3
Gate5           var     PortD.6
Gate6           var     PortD.7
Gate7           var     PortD.5
Gate8           var     PortD.4
WOT             var     PortB.0
 
ADCON0 =0       'A/D Off
ADCON1.3=1      'All AN channels digital
ADCON1.2=1      '''
ADCON1.1=1      '''
ADCON1.0=1      '''
TRISE.0=1       'EO input
TRISE.1=1       'E1 input
CCP1CON = %00000101  ; Capture mode, every rising edge 0101 
T1CON.7=1     'enable timer  16 bit `   
T1CON.6=1     'Timer1 OSC
T1CON.5=1     '1:4 prescaler
T1CON.4=0     '1:4 prescaler
T1CON.3=0     'Timer1 OSC off
T1CON.2=0     'sychro clock
T1CON.1=0     'internal clock
T1CON.0=0     'stop timer

pause 10
SEROUT2 LCD,84, [Prefix, LcdCls]

Include "modedefs.bas"	' Mode definitions for Serout
SEROUT2 LCD,84, [Prefix,CursorPS,0,"RPM test"]
pause 500

TMR1H = 0    'Clear 8-bit register                              
TMR1L = 0    'Clear 8-bit register
capture = 0  'Clear Timer

Calc_RPM_Timer:
TMR1H = 0    'Clear 8-bit register                              
TMR1L = 0    'Clear 8-bit register
capture = 0  'Clear Timer

StartRPM:
  If Capture = 0 then
     Goto StartRPM
  endif

  T1CON.0=1
  Capture = 0 
      
CaptureLoop:
  If Capture = 0 then 
    Goto CaptureLoop
  endif
T1Con.0=0  
period.LowByte=CCPR1L
period.HighByte=CCPR1H

TMR1H = 0
TMR1L = 0
Capture = 0
Overflow = 0 

RPM = 10000
RPM = RPM * RPM ' 100,000,000
RPM = DIV32 period ' 100,000,000 / RevCount
RPM = RPM * 60 ' Per minute
RPM = DIV32 1600
RPM = (RPM*5)

SEROUT2 LCD,84, [Prefix,CursorPS,20,"RPM ",dec5 RPM]
goto Calc_RPM_Timer 
return

**Bruce** · - 29th January 2010, 00:49

I've already ordered you the 4431. Sorry I placed the order before I knew you needed
or wanted 3. But I'm still holding the at cost offer open if you want this one even
if/when Darrel proves me wrong - which he does way too often...;o)

These really are some cool PICs'.

If you don't mind my asking - what boards are you using? Surely you're not de-soldering
and replacing TQFP package types!

**Tobias** · - 29th January 2010, 02:15

Yea I will take a couple of the PICS no matter the outcome here. One way or another I want to get it working right. I need to use DT's elapsed time file and also interrupt rpm at the same time. So I can't go back to the 877A unless there is another approach.

Desoldering the TQFP's is pretty easy. Not something I want to do often but when needed I have the technique down.

Thanks to both of you on helping me get this sorted out too.

**Darrel Taylor** · - 29th January 2010, 03:55

Proteus Rules.

Here's what I found.
After the capture was made, and the math completed, the result is displayed with a SEROUT2 to the LCD.

But at 9600 baud, by the time it gets around to displaying CCPR1L and CCPR1H, multiple captures have taken place without Timer1 being cleared.

If you use this line instead, you get the same numbers on both chips and they are the right ones.

Code:

SEROUT2 LCD,84, [Prefix,CursorPS,20,"RPM ",dec5 RPM, " ", DEC5 period]

**Bruce** · - 29th January 2010, 05:08

Nope...

Lets say we had ten thousand captures after this;

Code:

Loop:
TMR1H = 0    'Clear 8-bit register                              
TMR1L = 0    'Clear 8-bit register
capture = 0  'Clear Timer
'SEROUT2 LCD,84, [Prefix,CursorPS,0,"Capture ", dec Capture," ",DEC Period]
Start:
      If Capture = 0 then
         Goto Start
      endif

      T1CON.0=1
      Capture = 0 
      
CaptureLoop:
  If Capture = 0 then 
    'SEROUT2 LCD,84, [Prefix,CursorPS,0,"Dont Capture ", dec Capture," ",DEC Period]
    Goto CaptureLoop
  endif
T1Con.0=0  
period.LowByte=CCPR1L
period.HighByte=CCPR1H

They would all be zero since TMR1 was disabled after the T1CON.0=0.
Once period = CCPR1L & CPR1H it's a done deal until the next capture,
and Timer1 is re-enabled.

Capture works wether or not Timer1 is enabled. It just returns 0 if it's off.

Yet - I may be missing something...

**Darrel Taylor** · - 29th January 2010, 05:34

Yup...

But that's a very interesting point.

In the 4620 program it does turn off the timer at that point.
But the 877A program does not.

And you can even see the difference in the simulation.
The 4620 always has the same values for CCPR1L:H and period (+1?), but the 877A doesn't.

4620

Code:

      T1CON.0=1
      Capture = 0 
      
CaptureLoop:
If Capture = 0 then 
  Goto CaptureLoop
endif
T1Con.0=0  
period.LowByte=CCPR1L
period.HighByte=CCPR1H

RPM = 10000

877A

Code:

 T1CON.0 = 1             ' Start the Timer
 capture = 0             ' Reset  the capture flag
 
CaptureLoop:
IF capture = 0 Then 
  goto CaptureLoop	' Wait here until captured
endif
; nada  
period.lowbyte = CCPR1L		' Store the captured value in
period.highbyte = CCPR1H	' period variable

RPM = 10000

And when a Timer is turned off, it retains the last value in the TMR?L:H registers.
<br>

**Tobias** · - 29th January 2010, 05:48

With all that in mind, what do I need to change from the last code I posted to get RPM correct?

**Bruce** · - 29th January 2010, 05:50

Hey Darrel,

Disabling TMR1 after the event has zero affect on the value that CCPR1l & CCPR1H record at the time of the event. The value in Timer1 is captured immediatley on event.

Can you recommend something for the OP that wiil fix it?

If you can -- he gets the free PIC, and I'll send you a six-pack of your favorite BEER...

FYI: Disabling the timer after the capture event has zero effect on the value captured by the hardware catpure module.

Remember ... It's done in hardware

**boroko** · - 29th January 2010, 10:20

Darrel, help me understand here...

I thought that Proteus dealt with many versions of compilers except PBP.

Are you just running the already compiled program and reverse troubleshooting in your head? I imagine that watching the registers would let you know where you are at and if you could transpose between the two, it would work.

Care to elaborate on how you use it to help you?
Bo

**Darrel Taylor** · - 29th January 2010, 22:13

Originally Posted by boroko

I thought that Proteus dealt with many versions of compilers except PBP.

Currently, you can only simulate the .hex files directly in Proteus.
No single stepping through the program.

But you can use it as a debugger in MPLAB where you can single step and watch variables in PBP.
It's too slow for me that way though. It simulates in "Real-Time" when not animating.

It's like real hardware is sitting in front of you, only you can't create Smoke or Fry your chips.

I'm getting hooked on it.

It's just so nice not to have to build a breadboard to test things.
20 minutes, and you can draw it on the computer.

Plus, with the Logic Analyzer, O-Scope, I2C debugger etc.
It saves thousands on test equipment.

And my cats can't play with the wires.

<br>

**Bruce** · - 30th January 2010, 16:27

Darrel,

Are you using the free or full version?

Thread: Interrupt RPM and Taylors Elapsed time on 18F4620

Thread Tools

Search Thread

Display

Hybrid View

Free beer appreciated

Members who have read this thread : 0

Bookmarks

Bookmarks

Posting Permissions