Manometer project continued.

[aajgss]

OK, I have the math to a point where it is calculating somewhere near the number I want.

I have connected the hardware. A MPXV4006 pressure sensor which is connected to a MCP3201 ADC which in turn goes via SPI to a 18F14K50 PIC.

Attached is the setup.

' Variables
SCK     VAR PORTB.6 'Clock pin
CS     VAR PORTC.1 'Chip Select Active low
SDI     VAR PORTB.4 'Data In pin
 
 COUNTS var word
 lbit var COUNTS.byte0
 hbit var COUNTS.byte1
'-------------------------------------------------------------------------------
Define LOADER_USED 1
INCLUDE "MODEDEFS.BAS"
DEFINE OSC 12
DEFINE SHIFT_PAUSEUS 10  'gives a 50KHZ clock pulse                                                       
TRISB.4 =1               ' set PortB 4 as an input
TRISB.6 =0               ' set PortB 6 as an output         
 
TRISC =%00000000                        
knots var long
knotsRem var long
HIGH CS
START:
 
LOW CS 
PAUSEUS 50 'delay after CS goes low for conversion to start. Probably not needed
SHIFTIN SDI,SCK,2,[counts\14]  ' using mode 2 for SPI
high cs 
 
'knots = ((SQR(counts*100))*1563)/10000
'knotsRem = (SQR(counts*100)*1563)//10000
'serout2  portb.7,396,[ #Knots,".",dec3 knotsRem,"  ",#counts,",","   Knots",13,10]
serout2  portb.7,396,[ #lbit,"   ",#hbit,13] ' Just send out ADC value to start
GOTO START:
END

I can see the data coming in on the scope. If I move the plunger of the syringe, I get the data changing, but always get a Zero when I send it out to the serial port.

I havent used the shiftin command before. It looks right so I'm not sure where I am going wrong.

Any help would be appreciated.

TIA
aajgss

[Ioannis]

If your data to see on terminal is in the counts variable, you have this line commented out.

Ioannis

[aajgss]

Thanks Ioannis,

I changed it to counts, Unfortunatly no difference, but at least if it was working I wouldn't have seen it anyway. Thanks for the tip.

I also Added

ANSELH.2 = 0

This then worked, however I must say when I read the datasheet for the 18F14K22 I interpret it that it should be digital so it wouldnt be necessary to add this statement so I am a bit confused. Maybe another 40 or 50 reads and it might sink in!!!

regards
aajgss

[aajgss]

I now have everything working and calculating air speed

I used "DEC1" to only print out to one decimal point.

Running the program gave some strange results. On occasion the count would increase, but the Knots value woud reduce.

By printing out all of the remainder, I could see what was going on.
eg for 2577 counts, the value would be 110.7 and for 2579 counts the value would be 110.1. The problem being in the true value

2577 counts is 110.457 and 110.501. The "DEC1" would give the digit either the 1 or 7 and not the 4 or the 5 as I expected (read hoped)

How can I get the first digit?

thanks
aajgss

[mackrackit]

What does your code look like now?

[aajgss]

Hi Dave,

Code is attached. Thanks for looking

Define LOADER_USED 1
INCLUDE "MODEDEFS.BAS"
DEFINE OSC 12
DEFINE SHIFT_PAUSEUS 10                                                       
TRISB.4 =1               ' set PortB 4 as an input
TRISB.6 =0               ' set PortB 6 as an output         
ANSELH.2 =0
DEFINE ADC_BITS 8
DEFINE ADC_CLOCK 3
DEFINE ADC_SAMPLEUS 50
ANSEL =%10000000      ' Enable ADC channel-AN2
ANSELH =%0000                         
OSCCON=%01100000                       
TRISA =%00001111                        ' Set ports A3-A2-A1-A0 as inputs
                       
TRISC =%01001000                        
knots var long
knotsRem var long
zero var word
zeroSw var portc.6
i var byte
avg var word
HIGH CS
START:
for i = 1 to 10 
LOW CS 
PAUSEUS 150 'delay after CS goes low for conversion to start. Probably not needed
SHIFTIN SDI,SCK,2,[counts\14]  ' using mode 2 for SPI
high cs 
avg = counts + avg
next
counts = avg/10
avg=0
if zeroSw then  zero = counts                               'Get value at zero pressure
if zeroSw then high portc.2
counts = counts -zero  'set zero point 
if counts >65000 then counts= 0                           'in case of underflow
knots = (((SQR(counts*10000))*65500)/301)/10000
knotsRem = ((SQR(counts*10000)*65500)/301)//10000
serout2  portb.7,396,[ #Knots,".",Dec1 knotsRem," Knots   ",#counts,"   ",#zero,10,13]
GOTO START:
END

Thanks
aajgss

[mackrackit]

Try DEC2 then try DEC3 to see what the command does.
Then take a look at DIG

4.17.7. DIG

DIG returns the value of a decimal digit. Simply tell it the digit number (0 - 4 with 0 being the rightmost digit) you would like the value of, and voila.

B0 = 123 ' Set B0 to 123
B1 = B0 DIG 1 ' Sets B1 to 2 (digit 1 of 123)

Add a DIG statement to your code to "pull" the DIGit you want.

[aajgss]

Thanks Dave,

I looked at both as you suggested. These are the results

Number 40.8914

DEC2 returns the value 1
DEC4 RETURNS
Number 50.143

[aajgss]

Hit the end button before finished.

Results

DEC 1
47.6 Knots   686   
46.8 Knots   9598   
46.7 Knots   8727  
46.2 Knots   9162  
47.6 Knots   686   
46.4 Knots   8074   
46.2 Knots   9162   
46.7 Knots   8727   
46.2 Knots   9162   
46.8 Knots   9598   
DEC 2
46.28 Knots   5028   
46.69 Knots   3069  
46.28 Knots   5028   
46.81 Knots   5681   
46.93 Knots   893   
46.39 Knots   7639   
46.62 Knots   9162  
46.16 Knots   6116   
46.81 Knots   5681   
46.04 Knots   7204  
DIG 1
43.1 Knots   3910   
42.4 Knots   6946   
43.7 Knots   3475   
43.8 Knots   6086   
44.3 Knots   438  
43.6 Knots   4563   
43.5 Knots   5651  
43.5 Knots   5651   
43.8 Knots   6086  
43.8 Knots   6086   
DIG 2
49.0 Knots   7016  
49.5 Knots   8539   
49.0 Knots   5058   
49.9 Knots   8975
50.4 Knots   498   
50.4 Knots   498   
49.0 Knots   5058   
49.4 Knots   7451   
49.6 Knots   4622   
49.0 Knots   5058

The value before Knots is what is the result of whether I have used DEC 1or 2 and DIG 1 or 2. The value fi=ollowing is the knotsRem variable.

The DEC1 and 2 seem to make sense, although not displaying the first digit but the DIG 1 and 2 look to return the 2nd digit for DIG1 and 3rd digit for DIG2

Before I was confused - now I'm way past that!!
Looking at the manual there doesn't seem to be anywhere that left justifies a number.

Thanks
aajgss

[mackrackit]

    NUM VAR WORD
    NUM = 12345
    NUM_DIG VAR BYTE
    
    BOOT:
    SEROUT2 TX,BAUD,[$d,$a]
    'RETURNS 12345
    SEROUT2 TX,BAUD,["NUM= ",DEC NUM,$d ,$a]
    'RETURNS 345
    SEROUT2 TX,BAUD,["NUM= ",DEC3 NUM,$d ,$a]
    NUM_DIG = NUM DIG 4 'RETURNS 1
    SEROUT2 TX,BAUD,["NUM_DIG4= ",DEC NUM_DIG,$d ,$a]
    NUM_DIG = NUM DIG 3 'RETURNS 2
    SEROUT2 TX,BAUD,["NUM_DIG3= ",DEC NUM_DIG,$d ,$a]
    NUM_DIG = NUM DIG 2 'RETURNS 3
    SEROUT2 TX,BAUD,["NUM_DIG2= ",DEC NUM_DIG,$d ,$a]
    GOTO BOOT

[aajgss]

Hi Dave,

I think I understand where your heading with this.

Where I can see a problem with what I am doing is I dont have a constant number of digits in the remainder. It could be 1,2,3 or 4 characters long, So DIG2 would pick the right number if the remainder was in the hundreds, but the wrong number if the result was in the thousands.

If I remember back to Q Basic days there was a Left$ or Right$ and number of characters command, but I cant see anything like that in PBP.

Am I seeing this right or not?

aajgss

[mackrackit]

If I remember back to Q Basic days there was a Left$ or Right$ and number of characters command, but I cant see anything like that in PBP.

Nope...

Try something like this.
IF NUM_REM >= 9999 and <= 1000 THEN
DIG3 NUM_REM

IF NUM_REM >= 999 and <= 100 THEN
DIG2 NUM_REM

[aajgss]

Ah Yes,

A good solution, thanks. Using that method I could also now look at the second digit and if >= 5 round up to the next value.

aajgss