----- Original Message ----- 
From: "Jinx" <joecolquitt@clear.net.nz>
To: "Microcontroller discussion list - Public." <piclist@mit.edu>
Sent: Wednesday, August 13, 2008 8:59 PM
Subject: Re: [PIC] A/D impedances


>> The problem with this is that is does nothing to help in the case of a
>> constant bias current on the input.
>>
>> If your actual input impedence is 910K || 10K (call it 10K), then a
>> 1ua bias current results in an error of 10K*1ua = 0.01V
>>
>> This is not helped by any amount of capacitance.
>
> Taking a specific measurement example -
>
> 48V (nominal) SLA back-up battery
>
> 10000 ohm : 910 ohm divider (ie 11 : 1 or /12) to present a voltage
> of 4V to the PIC pin
>
> With a reference of 5.120V this equates to an FSD of (5.12/4 * 48)
> = 61.44V, or 60mV (of actual battery voltage) per step in the A/D
> result
>
> So, what I think you're saying is that because of the higher-than-spec
> impedance, the 4V has been reduced to 3.99V due to leakage in the
> A/D cap. Like the droop in a PSU o/p V with too great a load

Not leakage in the cap, but rather unpredictable currents in to (or out of!) 
the PIC input pin.

And as for accuracy vs. input impedence: that is exactly what the data sheet 
requirement is all about in regard to requiring a max source impedence less 
than some limit.
There are many factors that affect the accuracy of the A/D input, many of 
them intrinsic to the chip. This results in an error budget allocating some 
'slop' to various components of the system in order to allow mChip to 
guarantee a specific accuracy. One of the items in the budget is the offset 
voltage due to stray input current (and sample cap charge current, etc.).

-- Bob Ammerman
RAm Systems


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