Math question relating to AD input on PIC12F675

[rfetech]

OK.... that was a long night!!

I've finished the code and tested the circuit on a breadboard. After getting my head on straight, I was able to revise the main code quite easily, even for me!!

The headache and long hours occurred when I just couldn't get the A/D working correctly. I spent hours going over the data sheet and dozens of forum messages, tried different combination's of Defines and register settings but never got it working... UNTIL... I saw that the PAUSE statement was expecting a 16 bit argument. At least that is the way that I understand it. So, based upon that, I changed my code from being setup for 8 bit resolution to 10 bit and everything purred like a kitten. Got to bed at 6am with a smile on my face!!

Bruce, if you happen to read this, I strongly believe that your proposed book would have saved me many hours of head-scratching by translating the data sheet into easier to understand terminology. So, don't give up on it.

Thanks to all who post on this forum for giving me the use of your knowledge and helping us all learn from your own trials and your successes.

My extremely complex, yet simple code is available for scrutiny.

p.s. Special thanks to Louie and Al for responding to my original post and helping me much more than they probably realize. THANKS!!!

'****************************************************************
'*  Name    : Train_Auto-Reverser.pbp                           *
'*  Author  :                                                   *
'*  Notice  : Copyright (c) 2010                                *
'*          : All Rights Reserved                               *
'*  Date    : 10/27/2010                                        *
'*  Version : 1.0                                               *
'*  Notes   : Model train auto-reversing circuit using          *
'*          : a PIC12F675 and reed switch sensors.              *
'****************************************************************


'   ---------Configuration--------

' PICBasic Pro v2.6a default configuration for PIC12F675
'   __config _INTRC_OSC_NOCLKOUT & _WDT_ON & _MCLRE_ON & _CP_OFF
' Changed MCLR function to Input Pin during programming of PIC.


'   ---------Defines/Constants-------

DEFINE OSCCAL_1K 1              ' Calibrate internal oscillator

DEFINE ADC_BITS 10              ' Set number of bits in A/D result
DEFINE ADC_CLOCK 3              ' Set clock source (3=rc)
DEFINE ADC_SAMPLEUS 50          ' Set sampling time in uS

RY1 CON 4                       ' Relay 1 (Power) to GPIO.4
RY2 CON 5                       ' Relay 2 (Voltage reversing) to GPIO.5


'   ---------Variables---------

SW1 VAR GPIO.0                  ' Reed switch 1 to GPIO.0
SW2 VAR GPIO.1                  ' Reed switch 2 to GPIO.1
adval VAR WORD                  ' Store result of each A/D conversion
delay VAR WORD                  ' Store result of math conversion


'   ---------Initialization--------

ANSEL = %00000100               ' Set AN2 analog, all others digital
CMCON = 7                       ' Turn off analog comparators
ADCON0.7 = 1                    ' Right justify A/D conversion output


'   ---------Main Code---------

PAUSE 2000                      ' Let everything stabalize for 2 seconds.
LOW RY2                         ' Make sure RY2 is off for forward movement.

                            
mainloop:

HIGH RY1                        ' Turn on Power relay to start forward movement.

  WHILE SW1 = 1                 ' Keep RY1 active until reed switch SW1
    HIGH RY1                    ' is closed.
  WEND
                      
LOW RY1                         ' Removes power and stops movement.

GOSUB delayloop                 ' Keep stopped for x number of seconds.(0-63)
    
HIGH RY2                        ' Prepare RY2 for reverse movement.
PAUSE 500                       ' Let RY2 settle before activate RY1.
HIGH RY1                        ' Turn on power relay to start reverse movement.
    
  WHILE SW2 = 1                 ' Keep RY1 active until reed switch SW2
    HIGH RY1                    ' is closed.
  WEND

LOW RY1                         ' Removes power and stops movement.
LOW RY2                         ' Prepare RY2 for next forward movement.
    
GOSUB delayloop                 ' Keep stopped for x number of seconds.(0-63)
  
GOTO mainloop                   ' Keep repeating until user terminated.
    
END
  
    
delayloop:
ADCIN 2, adval                  ' Read A/D channel 2 into adval variable.        
delay = adval * 60              ' Creates values for PAUSE command that                     
PAUSE delay                     ' allow delays from approx. 0 to 61 seconds.
RETURN

[aratti]

Hi rfetech, glad you got it working as you wanted!

If I may suggest you an improvement to increase the stability of your delay, just take more ADC reading instead of one. In the example given I take 10 reading and than multiply for 6 instead of 60, this will reduce by a factor of ten any difference in the ADC reading.

Cheers

Al.

A0  Var Byte

delayloop:
Delay = 0
For A0 = 1 to 10
ADCIN 2, adval                  ' Read A/D channel 2 into adval variable.        
Delay = Delay + Adval
next A0
delay = delay * 6              ' Creates values for PAUSE command that                     
PAUSE delay                     ' allow delays from approx. 0 to 61 seconds.
RETURN

[rfetech]

Thanks Al. I'll add that to my code. The delay timing isn't critical but this extra code is good learning experience for me and I do appreciate it. FYI, this controller is for a small train that sits on a 12 ft shelf in the barber shop of a friend. He wants it to simply go back and forth to entertain the kids while they wait their turn to get a haircut. Who knows... some kid might be more curious about the controller, than the train, and become a PIC programmer.

Thanks again!

Mitch

[rfetech]

I thought this might be a good starter project for someone so I have attached the schematic.

Mitch

[aratti]

Mitch, it is a good practice never leave inputs pins or reset floating. So I suggest you to pullup pin 4 with a 10K resistor. Apart from this well done.

Al.

[tenaja]

The biggest problem was not your 8 or 10 bit ad reading, or even the type of variable used for the pause command. (Pause will work with a byte.) In fact, the problem was the use of a byte var instead of a word in your calculation.... 255/4 *1000 = over 63,000... and that obviously does not fit within a byte

[rfetech]

Thanks for the input, Al. I build a PCB for my project today and that is when I discovered that I had left out resistors R6 and R7 from my schematic. I also realized that I had not done anything with pin 4. I went to the data sheet and, after reading, I decided to use pin 4 as MCLR. I also decided that I wanted POR active so I added the RC network that was suggested in the data sheet and then turned on those options in the PIC. I don't know the pros and cons of that decision but the MCLR and POR could be useful, in my opinion. I've updated my schematic and included it in the following attachment.

Tenaja, thank you for looking over the code and pointing out that PAUSE will work with a byte variable. Actually, I did realize that a BYTE was not large enough for my original calculation after I had struggled with it for a long while. Thanks again!

OK, here's the latest iteration of my circuit. This is the way I have built it onto a PCB today and it appears to be working great. Just have to experiment with the magnets under the train in order to get the spacing correct for the reed switches. I will post the final version of the code once I get a chance to update it and maybe try a couple little changes.

Regards,
Mitch