LM5034 PWM controller soft start

I am looking for information regarding the shutdown of one PWM channel of the LM5034 chip.  The data sheet claims that connecting the soft start pin to an open drain connection allows the user to shut down one channel without affecting the other.  I have tested this and found that the second channel actually shuts down temporarily.  The second channel immediately recovers but is then controlled by it's soft start sequence.  This results (with my soft start capacitance = 1uF) in about 10 ms of no voltage on the supply that should have continued to regulate.  Unfortunately for me this is power to my ucontroller and as you can imagine comms (among other things) are adversely affected when power is interrupted.

The good:  I found a solution that seems to work fine which is to put 1Kohms in series between the soft start caps and the LM5034 s.s. pin.  I believe this works because the unaffected channel discharges it's charge through the logic inside the chip.  This can be seen on the scope where the voltage has an exponential decay equating to an RC value of about 2Kohms with my 1uF Css.

The bad:  the 'logic' (bottom of the block diagram on page 4 of the data sheet) is a mystery.  I contacted TI for details and they sent me here.

I would like to know what happens to my 'unaffected' Css charge.  Has anyone had a similar experience?