To amplifiy Bruce's suggestion:
An INPUT appears as a high impedance (mostly). Digital
inputs look like a capacitor, so adding a series-R only
slows the rise and fall times of the signal seen at the
port pin. Depending on the speed and the ability of the
external circuit to supply current, you could easily use 1K
to 10K as an isolating resistor.
One "gotcha" is the analog input. It is also, a high
impedance, mostly. The exception is the very short time
(nanoseconds) during which the ADC sample/hold gate is
functioning. The spec sheet warns you that the source
resistance for an ADC input is max 10K to achieve the
specified accuracy.
Going the other way, it depends on the impedance seen
looking into the external circuit. But, you DON'T need
resistors here, anyway. The reason is that, diring
programming, the port pins values are defined by the
programming function. Outputs will be connected to other
circuits capable of dealing with either logic level. If the
circuit could be damaged by an extended period of "wrong"
logic level, then you need to figure out some way of
disabling the circuit.
Jim
On Mon, 19 Jun 2006 15:35:42 -0700
Bruce Parham <obparham@jpl.nasa.gov> wrote:
> Resistors.
>
> Bruce
>
> Jay Dagenais wrote:
>
> > Thanks for the response, one question that has been
> plauging my current
>
> design choice is the isolation of programming pins from
> their alternate
>
> purposes. The tiny45 is only an eight pin device, of
> which I require six
>
> I believe for the programming, and that doesn't leave
> much for else. How
>
> would one go about isolating these pins functions from
> their alternate tasks?
>
> For my project, there will be one stepper motor driven
> via a L293N, and I think
>
> that is going to require at least 4 pins of the tiny45,
> then another couple for
>
> analog inputs to figure when to start and stop driving
> the stepper...How does
>
> one go about keeping these isolated from the programming
> functions? Other than
>
> using jumper pins and having to manually switch shunts
> each time I want to
>
> program?(Although that could work for my purpose, I would
> like a more robust
>
> solution..) Thanks in advance, Jaydag71
>
>
>
---------------------------------------------------------------
The Think Different Store
http://www.thinkdifferentstore.com/
For All Your Mac Gear
---------------------------------------------------------------