LM2903-Q1: Saturation voltage

Hidekazu-san,

We do not have distribution data for 8mA, just the 4mA datasheet production spec.

The char data that was taken was used to create the Io vs Vo graph in the datasheet.

The problem is that 8mA is on the knee of the current limit area - so that will add some variability, so we cannot extrapolate from the 4mA values.

Since this is an "uncontrolled" (untrimmed) parameter, the variation can be as high as 30% between lots.

The over temp spec is 700mV at 4mA, so it would be much worse at 8mA.

Have they considered the new & improved LM2903B-Q1 version? It has a slightly higher current limit, so the 8mA is further away from the current limit knee, and lower overall saturation.

Hi Hidekazu,

I would decrease Iol to 4mA, if possible.

Kai

Kai's suggestion is reasonable.  Unless you especially concerned with the rise and fall time of the output transition, increasing your pullup resistor value will improve the noise margin of your output logic levels.

Chuck