Problem using comparitor on SX28

Bandit · 2007-12-04 03:35

On an SX28 chip I set the RB.1 at 2.4 volts and RB.2 at 5 volts and expected to get a 5 volt output from RB.0.
The only output from RB.0 was a flickering 1 Volt.
Any idea what I am doing wrong?

http://forums.parallax.com/forums/emoticons/jumpin.gif
jumpin

[noparse][[/noparse]edit] I added a subject to your post... Bean

Post Edited By Moderator (Bean (Hitt Consulting)) : 12/4/2007 3:56:33 AM GMT

Bean · 2007-12-04 03:53

Have you enabled the comparitor ?
Please post your code.

Bean

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
www.iElectronicDesigns.com

·

Bandit · 2007-12-04 13:43

Here's the code; thanks for the reply!

' File Timer_RevC.SXB
' Purpose multi use timer
' November 25, 2007
'*******************************************************************************************************************
DEVICE SX28,OSCXT2, TURBO, STACKX, OPTIONX, BOR26 ' OSCHS3 HIGH SPEED OSCILLATOR BOR26 Brown out reset
'IRC_CAL IRC_SLOW ' TURBO RUNS 4X FASTER THAN COMPATIBILITY MODE.
FREQ 20_000_000 ' IRC_CAL calibrates the internal oscillator. SETS THE INTERNAL OSCILLATOR TO IT'S SLOWEST VALUE
ID "TIMER" ' SETS THE STARTING POINT OF THE PROGRAM FOR THE ASSEMBLER
' 50 kHz
'

Pin assignments

Go PIN RA.1
CLEAR PIN RA.0
AtFinish PIN RB.0 ' Comparator output bit
DigCtrl PIN RB
Segments PIN RC OUTPUT

'

' CONSTANTS
'

Yes CON 1 ' for active-high inputs
No CON 0

M_STOP CON 2 ' clock stopped
M_Run CON 1 ' clock running
M_Clear CON 0 ' clock clear and stopped

Blank CON %00000000 ' all segments off
DPoint CON %10000000 ' DP after sec display
DPDigit CON 3 ' DP after sec display

NumDigits CON 5 ' five digit display

'

Variables

ops VAR BYTE ' operational mode

ms VAR Byte 'milliseconds digit
hs VAR Byte 'hundredths digit
ts VAR Byte 'tenths digit
sec01 VAR Byte 'ones digit
sec10 VAR Byte 'tens digit

digPntr VAR Byte ' digital pointer
display VAR Byte(NumDigits) ' cuttent display segments

' THE COMPARE STATEMENT SETS RB.1 AND RB.2 AS INPUTS AND SETS RB.0 AND _PARAM1 TO 0 IF RB.1 > RB.2.
' So if the opt isolater is blocked RB.1 > RB.2. The LED lights and the timer stops due to _PARAM1 going high and
' setting to G0 = M_Stop
' Start interrupt routine no code before this routine

INTERRUPT 1000 'Rate 1000 ' CHECK ISR EVERY MILLISECOND

GOTO INT_HANDLER

'****************************************************************
PROGRAM Start
'****************************************************************
'

' Subroutine Declarations
'

UPDATE_DISPLAY SUB 0 ' Convert digits to segments
'

' Program Code
'

Start:
TRIS_B = %00000111 'CATHODE PINS SET TO OUTPUT Register B; Rb.0 Rb.1 Rb.2 are inputs Rb.3 thru Rb.7 are outputs
PLP_A = %00000011 ' pull up the unused pins on RA. Ra.0 & Ra.1 the weak pull up resistor is disable.

COMPARE 1, __PARAM1 ' STARTS THE COMPARATOR IN MODE 1; MONITORS rb.1 AND RB.2 OUTPUTS TO RB.0 to see if the optp isolator is blocked
' adjust Potentiometer to fine tune comparison. Opto isolator is = 5 volts __Param1 = 0 "two underscores before Param1.
MAIN:
DO 'ops sets the three mode of operation 0) stopped and cleared 1)running
' 2) stopped; ops is set in the interrupt routine
UPDATE_DISPLAY ' refresh segment array
IF Go = Yes THEN ' go button pressed
IF ops = M_CLEAR THEN ' were we cleared
ops = M_RUN ' yep so we can run
ENDIF
ENDIF
IF Clear = Yes THEN ' clear button pressed
IF ops = M_STOP THEN
PUT @ms, 0, 0, 0, 0, 0
ops = M_Clear ' clear the clock
ENDIF
ENDIF
LOOP
'

' Subroutine Code
'

INT_HANDLER:
IF ops <> M_RUN THEN GOTO Next_Digit 'skip update if stopped ---OPS HOLDS THE CURRENT THREE STATUS OF TIMER -
' STOPPED AND CLEAR, RUNNING AND STOPPED
Update_Clock:
INC ms ' increment millisecond display
IF ms = 10 THEN ' CHECKS TO SEE IF WE NEED TO WRAP BACK TO ZERO
ms = 0
INC hs ' increment hundredth display
IF hs = 10 THEN
hs = 0
INC ts ' increment tenths display
IF ts = 10 THEN
ts = 0
INC sec01 ' increment seconds digit
IF sec01 = 10 THEN
sec01 = 0
INC sec01 ' increment tens digit
IF sec01 = 6 THEN 'maximum time
ops = M_STOP ' stops timer
ENDIF
ENDIF
ENDIF
ENDIF
ENDIF

Next_Digit: ' Defines which LED is being cycled
INC digPntr ' point to next digit
IF digPntr = NumDigits Then ' check pointer --- number of LED (5), then resets DigPntr
digPntr = 0 ' wrap if needed
ENDIF

Update_Seqs: ' one display updated at a time
Segments = %00000000 ' clears the LED for next digit creates a crisper display.
READ Dig_Map + digPntr, DigCtrl ' update digital control - DP_Map is a table with one value READ COMMAND gets data from table offsets it digPntr and puts it in DigCtrl
Segments = display(digPntr) ' output new segments - Controls the cathode

Check_Finish:
IF AtFinish = No THEN ' CHECKS TO SEE IF OPTO-TRANSISTOR IS BLOCKED; NO = 1 At_Finish = _PARMA1 = RB.0
ops = M_STOP ' MUST PUSH THE CLEAR BUTTON TO RESET: M_STOP = 2
ENDIF

RETURNINT ' reset the RTCC value and re-enables interrupts
' End interrupt routine start program code
' Clear and Start go directly to Ra.0 and Ra.1. Ra.1 going high starts the timer, but the timer must be in mode zero defined as RAO_CLEAR.
' The timer starts by updating ops to M-Run. The timer stops when the opto-transistor goes high causing the comparator to go high the timer
' will be stop by changing the ops mode to M_STOP

UPDATE_DISPLAY: ' called only once Read (Table + location) not an addtion, Variable or word so display should be Rc
'
READ Seg_Map + ms, display(0) ' display (0) is the LED for milliseconds
READ Seg_Map + hs, display(1) ' Transfers seg_Map data table into display array.
READ Seg_Map + ts, display(2) '
READ DP_Map + sec01, display(3)
IF sec01 = 0 THEN
display(4) = %00000000 ' Blanks ms display
ELSE
READ Seg_Map + sec01, display(4)
ENDIF

RETURN

'====================================================================================
' Data tables
'====================================================================================

Seg_Map: ' segments maps without DP
' .gfedcba
DATA %00111111 ' 0
DATA %00000110 ' 1
DATA %01011011 ' 2
DATA %01001111 ' 3
DATA %01100110 ' 4
DATA %01101101 ' 5
DATA %01111101 ' 6
DATA %00000111 ' 7
DATA %01111111 ' 8
DATA %01100111 ' 9

DATA %01110111 ' A
DATA %01111100 ' B
DATA %00111001 ' C
DATA %01011110 ' D
DATA %01111001 ' E
DATA %01110001 ' F

DP_Map: ' segments maps with DP

sets the cathode
' .gfedcba
DATA %10111111 '0.
DATA %10000110 '1.
DATA %11011011 '2.
DATA %11001111 '3.
DATA %11100110 '4.
DATA %11101101 '5.
DATA %11111101 '6.
DATA %10000111 '7.
DATA %11111111 '8.
DATA %11100111 '9.

Dig_Map: ' Digit select map
DATA %11110000
DATA %11101000
DATA %11011000
DATA %10111000
DATA %01111000

Bean · 2007-12-04 14:03

You need to set RB.0 as an OUTPUT.

· TRIS_B = %00000111

Should be:
· TRIS_B = %0000110

Also what is the purpose of using __PARAM1 in the "COMPARE 1, __PARAM1" line ?

Bean

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
www.iElectronicDesigns.com

·

Bandit · 2007-12-05 05:21

Thanks again - this is the second time you've helped me out.

AtFinish = RB.0
The AtFinish when set high will stop a timer; this is a Nuts and Volts program from Parallax for a pine derby timer.

The high output will light an LED indicating the end of the race.
RB.1 and RB.2 differential input is controlled via and IR emitter and detector.

This is my first project so I am treking on unfamiliar ground.

JonnyMac · 2007-12-05 15:22

Hmmm..... I wrote that program and never had any problem with it. My understanding is that the comparator output bit is written directly to the RB.0 pin, regardless of the state of the TRIS register, and the the demo program for COMPARE (in the help file) seems to bear this out. When mode 1 is used (comparator output enabled) the LED lights, even though RB.0 was not specifically set into output mode.

For example, this still causes the LED to light:

INPUT RB.0
DO
  COMPARE 1, result 
LOOP

result is a bit variable, and RB.1 and RB.2 are tied to a potentiometers.

Post Edited (JonnyMac) : 12/5/2007 3:29:03 PM GMT

Bean · 2007-12-05 15:56

Hmmm, well then maybe that wasn't the problem.

Jon,
Do you know do RB.1 and RB.2 become inputs automatically too (even if you specify them as outputs) ?

Bean

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
www.iElectronicDesigns.com

·

JonnyMac · 2007-12-05 17:12

You know, the docs (original Ubicom and current Parallax) say to make RB.0 an output, and RB.1 and RB.2 inputs; somewhere, though, I read that the comparator output is connected directly to the RB.0 pin and doesn't rely on the TRIS register (I wish I could remember where I read that).

For grins, I just did another test like this:

Start:
  COMPARE 1, result
  INPUT RB.0

Main:
  GOTO Main

And the LED still lights! -- so that continues to suggest that the comparator output is connected directly to the port pin. My concern is what happens when RB.0 is set to output low and the comparator bit is high -- is there a conflict? I can't imagine the designers would allow this.

Bean · 2007-12-07 16:09

I just did an experiment.

When you enable the comparitor RB.0 is always an output and RB.1 & RB.2 are alway inputs. Even if you try to change them AFTER the comparitor is enable they don't change.

Bean.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
www.iElectronicDesigns.com

·

JonnyMac · 2007-12-07 16:25

That makes sense of what I'm seeing. Thanks, Terry.

James Newton · 2007-12-08 00:45

I've added that to:

http://www.sxlist.com/techref/ubicom/analogcomparator.htm

along with a lot of other findings from other peoples tests...

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
---
James Newton, Host of SXList.com
james at sxlist,com 1-619-652-0593 fax:1-208-279-8767
SX FAQ / Code / Tutorials / Documentation:
http://www.sxlist.com Pick faster!

Tracy Allen · 2007-12-08 06:05

Interesting. How about CMP_OE?...

When the comparator is enabled (bit 7 of CMP_B=0), AND comparator output is enabled (bit 6 of CMP_B=0), then RB.0 is the comparator output and overrides TRIS and the RB.0 states. But when CMP_OE=1, then RB.0 and !RB.0 manage that pin.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

Problem using comparitor on SX28

Comments