AM26C32: Termination questions

Team,

I have a customer using AM26C32MWB in a design and would like to use this receiver for some +3.3V I/O signals that they have. Based on App note SNLA041A, they have come up with the following termination scheme using a resistor divider on the negative input (B input). The A input is the +3.3V I/O signal and will swing between 0V and +3.3V:

The input structure for A or B input outlined in the datasheet is shown like:

The questions:

• I don’t believe the termination scheme they have shown will work. I get the sense that they may need to change the divider for the reference voltage or take another approach. Note that they came up with the divider values prior to analyzing the input structure. I’m thinking they can take advantage of the internal 288kΩ pull down resistor on the B input and use an external pull up to set the +1.5V reference.
• Similar comments for the A input; they will need to revisit my external 10kΩ pull down. The desire is to have the default output as a ‘0’ but we are flexible.
• In light of the high input impedance, we probably should remove the 100Ω series (R703/R704) termination resistor. This will provide minimal added protection when compared with the 17k internal resistor.
• We’ve created a spice model of the input structure and took a guess at the value of the unmarked resistors. The simulation shows very little change in output voltage for a wide range of input voltages. Is this behavior expected? The voltage output difference for A input = +5V and B input = 0V respectively is showing very little difference. Is there something in the circuit that we’re missing downstream that is able to amplify the voltage difference? The input differential is specified to be +/-200mV but we’re having a hard time seeing how the chip can sense that with the given input structure.