Subject:
|
RE: System interupt problems?
|
Newsgroups:
|
lugnet.robotics.handyboard
|
Date:
|
Thu, 18 Jun 1998 12:38:12 GMT
|
Original-From:
|
Ross, Michael <michael.ross@AVOIDSPAMlmco.com>
|
Viewed:
|
1161 times
|
| |
|
|
Tim,
I just scanned your code, so I might have missed it, but don't you have to
clear (or set) a flag in your interrupt routine to reenable the TOC4
interrupt?
Mike
----------
From: Timothy B. Gold[SMTP:goldt@ee.byu.edu]
Sent: Wednesday, June 17, 1998 5:11 PM
To: handyboard@media.mit.edu
Cc: goldt@ee.byu.edu
Subject: System interupt problems?
I need some help! We have a handy board hooked up to some decoders,
which we will use to help track our robot as it moves across the
floor.
We have to do a few pokes and a peek to read the decoders, but we've
got
it all set up and it works beautifully if you do it manually.
However, because the encoders have a very high resolution, we need
to
constantly read the decoders so that their counters don't overflow
and
wrap-around. I figured that I'd write an assembly routine and have
it
triggered by the system interrupt (TOC4) so that it is constantly
reading the decoders and keeping a total of how much they have
changed.
I successfully wrote an assembly routine that works just fine if
it is
called from a C program. To test it, I have a simple loop that runs
the
motor forward and the backward (for a specified amount of time),
calling
my assembly routine to update the value and the displaying the
global
counter each time.
However, if I take the EXACT same piece of code and link it in to
the
TOC4 interrupt, it doesn't work correctly. While before I read
approx.
80 (forward) and -80 (reverse) for each turn of the motor, now it
reads
approx. -6 (forward) and -1100 (reverse) for each turn.
I'll attach the code I'm using below. Does anyone have any ideas
as
to what might be causing the problem? Are there some other things
going
on with the interrupt that I should be aware of? I have tried
slowing
it down a little (keeping a count and only reading the decoders once
every 200 TOC4 interrupts or so), but the results are identical.
Any help would be appreciated -- this one's had me stumped all day
and
I have't been making any progress with it. Thanks!
----------------------------CODE:--------------------------------
/* Assembly routine(s) use to read the decoder
and update a global counter. This piece of code
will actually only read one decoder -- reading
the other one will be added when this is debugged */
#include "/usr/local/ic/libs/6811regs.asm"
ORG MAIN_START
variable_COUNT: /* this is the total count from the decoder */
FDB 00
variable_TCOUNT: /* this is used to store the value just read
from the decoder */
FDB 00
subroutine_initialize_module:
#include "/usr/local/ic/libs/ldxibase.asm"
ldd TOC4INT,X
std interrupt_code_exit+1
ldd #interrupt_code_start
std TOC4INT,X
rts
/* This is the interrupt routine that doesn't work... */
interrupt_code_start:
ldd #$c
stab $4000 /* poke(0x4000,0x0c) -- latch internal decoder
counter
on ouput */
ldd #$1
stab $4000 /* poke (0x4000,0x01) -- pull reset for decoder
high */
ldd #$d
stab $4000 /* poke (0x4000,0x0d) -- pull reset for decoder
back down */
ldab $4000 /* peek (0x4000) -- get value from decoder */
clra /* clear upper byte of variable */
tba /* copy regb to rega */
anda #$80 /* check to see if number is negative */
cmpa #0
beq skip_pos /* skip ahead if number if positive */
ldaa #$ff /* sign extend upper byte of count if negative
*/
skip_pos:
std variable_TCOUNT /* store read value (from decoder) to
TCOUNT */
ldd variable_COUNT /* load count into double accum. */
addd variable_TCOUNT /* add TCOUNT to COUNT */
std variable_COUNT /* store COUNT to memory */
interrupt_code_exit:
jmp $0000
/* This is the stand-alone routine that DOES work...
(same piece of code, except for one changed label and
an RTS at the end instead of a JMP) */
subroutine_mytest:
ldd #$c
stab $4000 /* poke(0x4000,0x0c) -- latch internal decoder
counter
on ouput */
ldd #$1
stab $4000 /* poke (0x4000,0x01) -- pull reset for decoder
high */
ldd #$d
stab $4000 /* poke (0x4000,0x0d) -- pull reset for decoder
back down */
ldab $4000 /* peek (0x4000) -- get value from decoder */
clra /* clear upper byte of variable */
tba /* copy regb to rega */
anda #$80 /* check to see if number is negative */
cmpa #0
beq skip /* skip ahead if number if positive */
ldaa #$ff /* sign extend upper byte of count if negative
*/
skip:
std variable_TCOUNT /* store read value (from decoder) to
TCOUNT */
ldd variable_COUNT /* load count into double accum. */
addd variable_TCOUNT /* add TCOUNT to COUNT */
std variable_COUNT /* store COUNT to memory */
rts
|
|
Message has 1 Reply:
 | | RE: System interupt problems?
|
| (...) I'm pretty sure that since the Handy Board also uses TOC4 as its system interrupt, that it takes care of clearing the interrupt. My code just gets linked in and gets executed before the Handy Board code (though I don't know what the HB code (...) (26 years ago, 18-Jun-98, to lugnet.robotics.handyboard)
|
2 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
