On Sep 26, 2010, at 2:35 PM, RussellMc wrote: > Less completely: Less completely, to resolve N values of a resistor chain, you have to =20 have an accuracy for the chain of better than 1/2^N. 5% resistors would clearly make it impossible to do more than 20 =20 levels, right, and it's probably worse than that given the way one is =20 supposed to calculate cumulative errors for multiple components (N * =20 individual error, IIRC?) For 2^6 values with 6 resistors, this gives you the same 0.26% that RM =20 got (but the last schematic posted had N+1 resistors for N switches...) Perhaps equally obvious, for a binary ladder, the error in the largest =20 resistor has to be smaller than the smallest resistor value. This is one of the attractions of an R/2R ladder for D/A =20 applications. Not only are a particular batch of resistors "somewhat =20 likely" to match to better than their nominal tolerance, but the error =20 in any one resistor won't "overwhelm" the entire chain. Is there nothing similar to an R/2R network for a purely resistive =20 output with SPST switches? There are certainly plenty of =20 "entertaining" resistor network problems with "interesting and non-=20 obvious" solutions; aren't there any relevant USEFUL cases? (I did a =20 bit of searching, but didn't find anything.) BillW --=20 http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .