simple opamp circuit ?

i need this for a Due input so 0-3.3

my signal is ~1-4v so i need to subtract 1v from the input to get it in range
it has to be fast (signal changes every 2us) & not AC coupled ..

IF i had a negitive supply i could sum the inputs ... but i don't have negitive supplys at hand (ment to be in a portable device )....

does anyone have a solution ?

IF i had a negitive supply i could sum the inputs

Most op-amps are going to require a positive & negative supply if the output needs to go all the way to zero (or negative). Otherwise a [u]differential amplifier[/u] (difference amplifier) with a gain of one, and +1V on the negative input would do the trick.

.75-3 & doing in software doesn't cut it - you loose 1/4 the range ... no point in having 12 bits (4096 ) if you toss out 1/4 of them

plenty of rail to rail single supply opamps /

Most (all?) "rail to rail" op-amps, aren't. Close maybe, but still asymptotic. Only way to get 100% of a voltage range is for the supplies to exceed it.

. but i don't have negitive supplys at hand

How about using a voltage mirror chip for the negitave rail?

Here’s a circuit that should work with any 5V rail-to-rail output op amp.

SirNickity:
Most (all?) "rail to rail" op-amps, aren't. Close maybe, but still asymptotic. Only way to get 100% of a voltage range is for the supplies to exceed it.

understood / but i don't want to lose 25% of my width (5-10% if i really have to ..)
and at some point don't mind if 0V is hard to get to ...

dc42:
Here’s a circuit that should work with any 5V rail-to-rail output op amp.

i’ll give that one a try
i had tried today with a similar circuit / but to find out that i have to wait for smd holders for the proper opamp - 8dip ones i was trying with didn’t have the bandwidth …

ralphnev:
.75-3 & doing in software doesn't cut it - you loose 1/4 the range ... no point in having 12 bits (4096 ) if you toss out 1/4 of them

plenty of rail to rail single supply opamps /

Losing 1/4 of the range still leaves 11.59 bits, and that might be more than the effective bits anyway.

ralphnev:
i had tried today with a similar circuit / but to find out that i have to wait for smd holders for the proper opamp - 8dip ones i was trying with didn't have the bandwidth ...

This could be my own ignorance, but what about that circuit requires higher bandwidth than you can get in a DIP-8 package? Was it oscillating?