(Note: This was recently posted to the Microchip PIC WebBoard.  Sorry if =
you've read
it there as well.)

OK, I'm looking to write several bytes of data to the EEPROM quickly. =
Here's the
code:

; Initialize data fields needed
        movlw   data
        movwf   FSR
        movlw   5
        movwf   scratch
        clrf    EEADDR  ; Start with addr 1

wrt_et  incf    EEADDR,F
        movf    INDF,W
        movwf   EEDATA
wrt_wt =20
        bsf     GPIO,LINE1
        bcf     GPIO,LINE1
        call    WRITE_BYTE
        btfss   PC_OFFSET,7
        goto    wrt_wt

        bsf     GPIO,LINE2
        bcf     GPIO,LINE2

        incf    FSR,F
        decf    scratch,F
        btfss   STATUS,Z
        goto    wrt_et

Quick summary: FSR is pointing to the beginning of five bytes of data. =
LINE1 and
LINE2 are two I/O pins just for debugging access time. All data locations =
are
properly declared, and TRIS bits are correct for LINE1/2. Vcc is 4.0 V =
(+/- 0.1V).

OK. The original example of the Flash51x code for a byte write is as =
follows:


Example 1:  Write a byte to EE Data memory
        MOVLW   0x05
        MOVWF   EEADDR  ; Set EE Data Address
        MOVLW   'A'
        MOVWF   EEDATA  ; Set EE Data
        CALL    WRITE_BYTE ; initiate write
        BTFSS   PCOFFSET,7  ; was command accepted?
        GOTO    NOWRITE    ; Command rejected...


Now, they don't give any examples for what NOWRITE is, or where it goes. =
From
what I figured, if a write command failed, I want to try it again until it =
does.

OK, here's the problem: It's taking long than expected to write a byte. =
WRITE_BYTE
takes approx 2.5 ms at our clock rate. The data sheet (DS40139E) states =
that it
can take up to 4 ms for the write to finish. So, from my estimate, 6.5 =
ms/byte is the
max time, and I'm looking at 9 to 10 ms of write time.

Here's some timing info:

BYTE1:
  LINE1 toggles.
  LINE2 toggles approx. 7.5 ms later.

BYTE2 through 5:
  LINE1 toggles right after previous LINE2.
  LINE1 toggles again approx. 2.5 ms later.
  LINE2 toggles approx. 7.5 ms later.

Does everyone else call the WRITE_BYTE routine immediately again after a =
write
failure? Right now, I could wait 4 ms default after each call, but I'd =
hate to waste
time if I don't need to wait.  Another question is: what is the code doing =
during
this 7.5 ms?  I'd expect the code to try to write a byte every 2.5 ms, =
constantly,
until it was taken.  I didn't see any while loops in the MChip flash51x.asm=
 code.

Anyone have any idea, or am I the only one on this Conference that needs =
the
minimum write times possible and the only one to have timed the writes?

John
--
John Patrick
Software Engineer - Job
Competitive Robotics - Hobby

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu