Serin woes - qualifiers not working

Printable View

- 29th January 2010, 17:24
tekart

2 Attachment(s)

Serin woes - qualifiers not working

This is pretty basic - but it's driving me nutz.

I have a serin routine that is not responding to the qualifiers. The point of this circuit is to take the serial processing work off of another PIC, so I'm using an 12F509 to wait for a particular serial burst with 2 specific qualifiers. Then it sends a 2mS pulse to the host processor to let it know that valid data arrived. The serial is coming in via a MAX232 - so it's uninverted.

I wrote a test line that just waits for the serial data line to go low - and that works. The data line is clean and stays high between comms - see scope pic below.

The serial burst looks like this:
Tx:0F 03 00 00 00 05 84 E7

Guy
- 29th January 2010, 18:52
mackrackit

Quote:

so it's uninverted.

Depends on what it is before it hits the 232.

Please post the code from both PICs using code tags.
- 29th January 2010, 19:38
tekart

The serial data is not coming from another PIC - it's from a PC serial port DB-9, which is one reason for the MAX232. So it's pretty standardized format.
- 30th January 2010, 18:39
Bruce

I would lower the baud rate, or use DEBUGGIN, or use an external crystal.

Also change RETURN to GOTO Main.
- 30th January 2010, 20:03
ScaleRobotics

Quote:

Originally Posted by tekart

This is pretty basic - but it's driving me nutz.

I have a serin routine that is not responding to the qualifiers. The point of this circuit is to take the serial processing work off of another PIC, so I'm using an 12F509 to wait for a particular serial burst with 2 specific qualifiers. Then it sends a 2mS pulse to the host processor to let it know that valid data arrived. The serial is coming in via a MAX232 - so it's uninverted.

I wrote a test line that just waits for the serial data line to go low - and that works. The data line is clean and stays high between comms - see scope pic below.

The serial burst looks like this:
Tx:0F 03 00 00 00 05 84 E7

Guy

Hey Guy,

I got it to work for me on my 12f683, but I had to change this line, so I could use characters that I can type on my keyboard:

Code:

SCADAin,2,200,NoOp,["15","3"],X ' check f"or right qualifiers only 

Then when I enter 153x (x=any character or number) your code works, and X=x
- 1st February 2010, 22:24
tekart

2 Attachment(s)

Still stuck

Thanks for the effort I appreciate the time you spent . . . but I'm not receiving bytes as ASCII - they come in as numbers, and I can't change that.

Had a long chat with Charles at MElabs and he had me go over all the hardware and also try DEBUGIN. Still not recognizing the qualifiers.

See the screen shot from my LOGIC probe that shows that the serial data is arriving at pin6 (GP1) of the 12F508.

Also see my code with variants on several tests.

Anyone else go a take on this? I JUST need to respond to the qualifier and generate a pulse. I need a relativelty short timeout because other serial data is happening on this line at times. I just need to pick out one (or 2) qualifiers which are the first 2 bytes of the data transmission - they are $0F and $03 (15 and 3 decimal). Pretty darned simple in concept.

Guy
- 1st February 2010, 23:45
mackrackit

What is hitting the PIC?
15 or 0F
3 or 03

Code:

SERIN2 PORT.x,MODE,[WAIT("0F"),WAIT("03")]

Notice SERIN2... Never did like the single...
- 2nd February 2010, 16:37
ScaleRobotics

Quote:

Originally Posted by tekart

Had a long chat with Charles at MElabs and he had me go over all the hardware and also try DEBUGIN. Still not recognizing the qualifier

Hello Guy,

Your code works for me, even with hex. (at least on my 12f683). But, I have to add this at the top of my code. Otherwise, I get no response.

Code:

CMCON0 = 7 'TURN COMPARITORS OFF ANSEL = %00000000 'Set all pins to digital I/O ADCON0 = %00000000 'Analog converter OFF

Here's what I used for my 12f683:

Code:

define OSC 8 OSCCON = %01110000 'set for 8mhz internal CMCON0 = 7 'TURN COMPARITORS OFF ANSEL = %00000000 'Set all pins to digital I/O *wont work without this * ADCON0 = %00000000 'Analog converter OFF '* uC : PIC12F508 * '**************************************************************** ' chip setup ' oscillator = INTRC ' watchdog = enabled ' MCLR = INPUT X var byte ' dummy variable ' ************** DEBUGIN SETUP ************ define debugin_bit 1 ' incoming data bit define debug_baud 9600 ' baud define debugin_mode 0 ' 0 = true ' Pin Assignments SCADAin var gpio.1 ' data in via MAX232 Ready var gpio.2 ' signal main processor to send data 'I changed from gpio.4 to match my hardware low ready ' set low ' config TRISIO = %000010 ' set I/O directions (0 = output, 1 = input) 'OPTION_REG = %01011111 ' 12F508/9 bit 6 disables weak pull-ups OPTION_REG = %01000010 ' for my 12f683 Main: ' while scadain = 1 : wend ' test code detects presence of serial burst ' debugin 200,Noop,[wait(120)] ' wait for address and command serin scadain,2,400,noop,[$0F] ' check for right qualifiers only pause 10 ' wait for serial burst to end high Ready ' signal main processor to send current data pause 2 ' with a 2mS pulse low ready goto main NoOp: return
- 2nd February 2010, 17:35
tekart

Quote:

Originally Posted by mackrackit

What is hitting the PIC?
15 or 0F
3 or 03

Code:

SERIN2 PORT.x,MODE,[WAIT("0F"),WAIT("03")]

Notice SERIN2... Never did like the single...

Doesn't matter hex or decimal - it's the same binary number sent via serial...
I had not noticed the WAIT feature in SERIN2. But the manual warns that at 9600 baud I *may* need an osc faster than 4MHz. Have not used SERIN2 before and it is giving me compiler ERRORS with the 12F509A.
- 2nd February 2010, 18:21
Bruce

Does this work on your 12F509?

Code:

DEFINE DEBUG_BAUD 9600 DEFINE DEBUGIN_REG GPIO DEFINE DEBUGIN_BIT 1 DEFINE DEBUGIN_MODE 0 '1=inverted X var byte ' dummy variable ' Pin Assignments SCADAin var gpio.1 ' data in via MAX232 Ready var gpio.4 ' signal main processor to send data low ready ' set low ' config TRISIO = %000010 ' set I/O directions (0 = output, 1 = input) OPTION_REG = %11011111 ' 12F508/9 bit 6 disables weak pull-ups Main: debugin 200,NoOp,[wait($0F)] ' wait for address and command pause 10 ' wait for serial burst to end high Ready ' signal main processor to send current data pause 2 ' with a 2mS pulse low ready goto main NoOp: GOTO Main END
- 2nd February 2010, 18:31
tekart

Yes! :)

Bruce,
It WORKS! I had tried variations of DEBUGIN before but must have gotten the setup wrong! THANKS MILLIONS - you have no idea how much time I have put in on this (well, maybe you do...).

Still don't know why SERIN did not work.

Guy
- 2nd February 2010, 18:41
Bruce

Hi Guy,

Glad I could help.

One problem was having RETURN in NoOp. The timeout/label option for serin/serin2/debugin use a GOTO to jump to NoOp. Not good since you have no idea what's on the call stack, and no idea just where it's going RETURN to when it lands on this return after only a few mS when it times-out.

Also .. serin/serout2 don't always work as expected at 4MHz for 9600 bps. debugin does.

If you refer back to my 1st post you'll see that I hinted about all of these potential show-stoppers...;o)