PIC asm question / Computer Science / Forums

Forums

4hv.org :: Forums :: Computer Science

« Previous topic | Next topic »

PIC asm question

Move Thread

LAN_403

AndrewM

Mon Apr 10 2006, 01:33AM

Registered Member #49 Joined: Thu Feb 09 2006, 04:05AM
Location: Bigass Pile of Penguins
Posts: 362

Easy question, just dont have a PIC on hand to lash up and resolve this for myself:

When I move binary bytes to the PIC ports, which port bit corresponds to which bit in my binary string. For example, does 00000001 have the "1" in the pin 0 or pin 7 position??

Is this true for bank 0 and bank 1?

I ask because I *think* i remember setting the port directions in bank 1 operating differently that setting the output states normally in bank 0

[mod edit] Moved from Gen. Sci

Bjørn

Mon Apr 10 2006, 02:32AM

Registered Member #27 Joined: Fri Feb 03 2006, 02:20AM
Location: Hyperborea
Posts: 2058

RA0 and RB0 both corresponds to bit 0 in the file registers, so they are identical in that sense and there are no differences in how they are set up.

Port B has programmable pull up resistors for input.
Port A may have only 5 I/O lines, depending on the model.
Port A may have a pin with open collector and may require an external pull-up resistor or a diode to protect it from overvoltage.

AndrewM

Mon Apr 10 2006, 02:33AM

Registered Member #49 Joined: Thu Feb 09 2006, 04:05AM
Location: Bigass Pile of Penguins
Posts: 362

I guess I meant which bit is bit 0?

00000000
       ^- this one?

00000000
^- or this one?

[Edit: You forgot the [code] tag.]

Bjørn

Mon Apr 10 2006, 02:57AM

Registered Member #27 Joined: Fri Feb 03 2006, 02:20AM
Location: Hyperborea
Posts: 2058

That is a matter of definition. Bit 0 is always bit 0, no matter what the actual position is.

The only way to make it completely clear, bit 0 in the file registers is labeled RA0 and RB0 on the pinout diagram in the datasheet.

movlw .1
movwf PORTB
Will cause RB0 to go high.

movlw .128
movwf PORTB
Will cause RB7 to go high.

...

Mon Apr 10 2006, 04:19AM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

I think what andrew is asking is which bit is the 0th bit in a binary number...

In that case, I am pretty sure it is the one furthest to the right (first example)

Carbon_Rod

Mon Apr 10 2006, 07:42AM

Registered Member #65 Joined: Thu Feb 09 2006, 06:43AM
Location:
Posts: 1155

1.) Multibyte variables are generally â€œLittle Endianâ€ these days (unless one runs a SPARC machine etc. which is â€œBig Endianâ€ .) (

) It refers to how multi-byte variables like â€œunsigned intâ€ are ordered when stored in memory. If ones defined a union or struct this can be a source of confusion and porting bugs. Also, if one declares signed integer math then â€œoverflow resetâ€ monitoring may make the code far more complex than needed.

2.) Simply put, the Microchip standard states RA0 as being the Least Significant Bit (

) of a memory address often referring to a single byteâ€™s bit on the right side. As Little-Endian reorders byte sequences it is a common source of bugs when doing direct bit wise operations with i/o and relying on the compiler to correctly downcast the variable. To keep it simple Microchip avoids these problems by forcing the programmer to downcast the output to 1 byte (unsigned char).
RA0 := â€˜00000001â€™b
RA1 := â€˜00000010â€™b
Etc.

3.) Most C compilers implement a C standard procedure stack order. When translated into procedure calls it is standard to pass arguments from right to left on the stack. When using precompiled libraries and extern calls to binary objects this can be an issue. This is why Pascal object code integration is rather rare as arguments would pass left to right on the stack in the compiler making it incompatible (only mentioned as there are numerous compilers for PICs out there that implement non-standard linker systems.)

4.) Microchips â€œRead Modify Writeâ€ bug is a problem that has documented procedures to avoid. See the app notes.

5.) Also the rotate ASM calls will sometimes carry in garbage from the flags if one forgets to backup, clean, and restore the core state to perform operations with interrupts active.

6.) 'xxxxxx1x'h != 'xxxxxx1x'b is a common typo

7.) Not all memory addresses are mirrored between ram banks (see TRISA and PORTA defines in the compilers pic16xxx.inc etc.)

This should cover most of the common bit related bugs,
Good luck,

Steve Conner

Mon Apr 10 2006, 10:06AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

It's the one furthest to the right.

Bjørn

Mon Apr 10 2006, 10:59AM

Registered Member #27 Joined: Fri Feb 03 2006, 02:20AM
Location: Hyperborea
Posts: 2058

If the question is about the I/O ports after all, then it is the opposite. When looking at the row of pins that is closes to you then bit 0 is to the left on chips like the 16F84 and 16F870. If you don't rotate the chip when you work on the other side then the ports on the other side will be mirrored since it is upside down.

The rule is that the significance of the bits follow the numbering of the pins so the least significant bits always have the lowest pin numbers.

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob