Right.  And even then, the clamp can conduct some, but I don't know if 
the 16F886 gives a specific max or not.

That current will conduct to the Vdd rail.  If it is more than the 
circuit consumes, then it may raise Vdd higher, which may or may not 
cause problems for the circuit.

Anyhow, you've gotta look at several factors:
The accuracy of the zener's forward voltage
The IV profile of the zener
The temp coefficient

A 3.3v zener with a common 5% tolerance is a 3.135v-3.465v anyways. 
Very small zeners typically have higher dynamic resistances.  It may 
have a knee voltage of 3.0v but if it needs to clamp 30mA it may rise to 
4v.  This may fail to meet your criteria for circuit protection.

And the zener does have leakage below the specified voltage.  In some 
the current is measurable even at half the zener voltage.  If your input 
impedance is high this may result in unacceptable Vh.  The point however 
is that higher zener Vf will reduce the sub-knee leakage within a 
particular product series.  You can also choose a product with a sharper 
IV curve. 

So you're gonna want to look at spec sheets and determine what protects 
your circuit in the worst-case of being on the high side of tolerance 
and seeing the highest positive clamping current.  But I'd choose the 
highest, sharpest Vf I could get to avoid leakage currents on the input 
when it doesn't need to be doing anything to the signal.

Danny

Jinx wrote:
>> so I'm expecting a 3.0V zener to be perfect for this purpose - high 
>> enough so that the voltage is over Vih, but low enough that the input 
>> will never exceed Vcc.
>>     
>
> You can exceed Vcc by 0.3V max. For 5V Vcc I'd typically use a
> 5V1 zener. Either 3V0 or 3V3 would be OK, given the 3V6 max
>
> You could calculate the maximum voltage that could be clamped by
> the series resistor and zener by using their value/voltage/wattage. And
> also the sink current limit of the PIC pin if the resistor was small enough
> to allow more than eg 20mA through after the zener had taken its share
>
>   

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist