I need to drive a resistor divider bridge with relatively high current from a digital output of a microcontroller. I have some control over the fixed resistor value, but the variable resistor value (sensor) could cause the current in any event to be as much as 10mA+. I'm measuring the output voltage at the bridge point with an ADC channel. It works with an AVR microntroller but as the variable resistor decreases, the current increases and the voltage from the digital output decreases. This creates a non-linear effect which is messing with my calibration. Its like a third variable resistor. The sensor resistance varies from ~100K to ~100ohm and I'm using a fixed resistor of 330ohms. So thats about 12mA max at 5V.

It occurs to me I could use an open collector/drain buffer. I'm using 2 pins to create a pseudo square wave output to excite the sensor. When one is low the other is high. So although the high voltage will now not drop the low voltage drop is important.

This buffer looks like it would be close to what I need:

https://www.ti.com/lit/ds/symlink/sn75452b.pdf

If I'm reading the Vce(sat) vs Ic graph correctly it looks like the voltage drop would be about 100mV at 12mA. Actually more importantly for less current the voltage drop would seem be constant. This is what I really want. I'm assuming there would be some leakage current through the buffer that is in the high state but that should be very small and could be ignored? Thanks.

 

Don't really understand what your are trying to do.
Since a picture is worth a thousand words, please post a schematic of what your circuit and what you want it to do.

If you are tying to linearize the output of the bridge (make the voltage directly proportional to the sensor resistance), there may be better ways to do it.