Posted by Ron Michallick on 16, 2001 at 8:14 AM:
In Reply to: K64 vs IPac: responsivity posted by Ray on 15, 2001 at 9:08 PM:
I make the MK40 and MK64 encoders.
The I-PAC has a very good polling method.
I choose not to use the same method because
it requires a timer for each input (RAM) and
my microprocessor doesn't have enough free RAM
for 64 inputs.
My method is just as fast with one exception.
When two actions (press or release) happen
less then 0.0125 seconds apart there will be
an average delay of 0.0063 seconds for the
second action. This effect is not cumulative.
That means if many actions happen within 0.0125
seconds no action will be delayed more than
0.0125 seconds. (average is still 0.0063 sec)
This effect should not be noticeable.
The timers on the IPAC and MK's are used to
prevent switch contact bounce from acting like
multiple actions. The timer value on the MK's
can be changed via computer software. I believe
the IPAC would need a chip swap the change the
timer.
There should be no reason to change the timer
value unless you had VERY bouncy switches.
:
: Folks
: Has anyone directly compared the responsivity of the K64 encoder and the I-Pac controller?
: I-Pac looks to have a tighter polling loop, but K64 seems to have some intelligent polling.
: Basically, is there a different noticable on highly timing sensitive games (fighting games, etc.)?
: Just curious how noticable different polling techniques are.
: Ray