On Friday 20 April 2007 11:26, Michael Rigby-Jones wrote:
> I think you may be over thinking this :D

If only I had a penny for every time I was told this. :-)


> 1) The datasheet recommends a maximum source impedance of 10k, you are
> significantly below this, even with the 1k series resistor (worst case
> 1.27k). Providing your aquisition times are set correctly, you should not
> experience any degradation in accuracy.

Here, I was not so worried about exceeding that max impedance, but how the 
series resistor would affect the ratio of the voltage divider.  I guess what 
I'm trying to verify is that Zad is large enough that adding RL would make a 
very insignificant change in this ratio.  I have since done some math and 
figured out that the max effect would be at the highest value of Rsen.  
Assuming that the internal impedance of the A/D is just Rss (7kohms) (and we 
know it's much more than that), then adding RL=1Kohm causes a 1.1% change in 
the ratio of the voltage divider.  Significant, but again I know that the 
internal impedance of the A/D circuit would me much more than 7K.

At this point, I think I'll still re-calibrate for the sensors though.


> 2) The absolute maximum current into the protection diodes is shown in the
> asbolute maximum ratings table as "Input clamp current" and "Output clamp
> current" which are both 20mA.

Ah.  Gotcha.  This works out to 450 ohms now for RL, but I think I may still 
use 1kohm, since there are already a few more of these on the board -- fewer 
parts during production.


> 3) I would be concerned about the effects of 14v on the input pushing Vdd
> up via the 270R resistor.  That's a current of 14-5/270 = 33mA being pumped
> into your Vdd rail.  If the circuit consumes less than this current the
> rail voltage will rise.  A zener clamp on the rails would be a good idea,
> if not already present.

Since I posted, I had changed it to 330 ohms, so that with 14V at the input, 
there's less than 1/4-Watt thru that resistor (at 27mA).  I'm balancing the 
urge to use a larger value for R1 (lower power consumption in normal use and 
better protection) vs. a lower value for R1 (better A/D range).  The circuit 
gets it power off either a 7805 or a switching regulator.  Power consumption 
for the rest of the circuit is at least 110mA.

Thanks,
-Neil.

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