I think Andre was writing about how to solve: >> Vref = 14.4*(R2/(R1+R2)) for R1 and R2. Obviously, it can't be done. What he apparently missed was choosing an operating current first. Then it gets easier. Andre, you need to decide how much current you want to have flowing through your resistor divider. Then R1 + R2 = 14.4V / I where I is the current. Generally, there is no reason to have high levels of current flowing (amps) unless the divider is also part of say, a discharge network to put a heavy load on a battery and provide a lower voltage for an A/D for example. The upper limit on R (or the lower limit on I) comes from at least two other considerations. How much will your A/D load the network? And are there noise sources that will influence a divider made up from megohm size resistors? If the A/D load is constant during reading, then it can be taken into account. But it is usually a better plan to either buffer the divider with a unity gain opamp or ensure the divider impedance is much lower than the A/D impedance. For example, if your A/D input impedance was 100,000kOhms, then you could use a 1kOhm network and have around a 1% error, etc. So... a) choose an operating current I b) solve for Rt (total divider resistance) = 14.4V/I c) R2= Vref/I d) R1= Rt - R2 And consider placing a capacitor in parallel with R2 to help with noise if needed. Tom Messenger (who frequently tries to break Ohm's Law) Scott Dattalo wrote concerning Andre's missive: > >I'll buy you a budweiser if this is not a sense circuit for a Lead Acid battery >charger using something like the UC3904. Btw, don't forget your temperature >compensation. Oh and don't forget the hysterisis issues too. And why you're at >it, the resistance values can be increased by an order of magnitude. > >> this is what I used to calculate resistor values >> >> Vref = 14.4*(R2/(R1+R2)) so after changing resister values I got >> R1 = 8,2 k >> R2 = 1,782 k >> > >Andre. I guess I don't understand the question. Do you need help with algebra or >am I missing something? > > Vref = Vbat* (R2/(R2+R1) > > Vref*R1 = R2*(Vbat-Vref) > R1 = R2*(Vbat-Vref)/Vref -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.