Olin Lathrop <olin_piclist@EMBEDINC.COM> wrote:
> I wrote:
> > Note that the 5% values of 3K and 18K give an exact match,
>
> Yeah, but in name only.  Such a voltage divider will have a 1/7 ratio with
> up to +8.9% to -8.3% error.

Well, yes. (Duh!)

If you need precision, use precision resistors.

If the distribution of individual resistor errors is assumed to be uniform
(not really true in practice), the error of the divider has a triangular
PDF. 75% of the time, the error will be less than half the peak value you
calculated. For all practical purposes, you can assume that the error of
the divider is no worse than the error of the individual resistors.

> I recommend that anyone seriously using this Perl script modify it to
> show the worst case error for each combination, taking into account the
> resistor tolerances.

Which is pointless unless the nominal error value is within reason.
My script handles the grunt work of finding candidate resistor pairs.
It isn't going to design your circuit for you.

-- Dave Tweed

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads